System

Abstract

A system is provided.SOLUTION: A system comprising means for collecting weather data, means for collecting soil data, means for collecting crop growth data, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting an optimal harvest time and harvest method using the trained generative AI model, means for notifying an agricultural operator of the optimal harvest time and harvest method based on the prediction, and means for collecting feedback data and retraining to improve the generative AI model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In traditional agriculture, a lack of information to determine the appropriate harvest time and method is a challenge. In particular, uncertainties exist in weather fluctuations, soil conditions, and crop growth, forcing farmers to rely on experience and intuition. As a result, yields and crop quality are often not optimized, leading to unstable profits. [Means for solving the problem]

[0005] The present invention provides a means for collecting weather data, soil data, and crop growth data, preprocessing the data, and training a generative AI model. The system includes a means for using the generative AI model to predict the optimal harvest time and harvesting method, and notifying farmers of the optimal harvest time and harvesting method based on the prediction results. Furthermore, the accuracy of the generative AI model can be improved by collecting feedback data and retraining it. In this way, it is possible to streamline agricultural harvesting operations, maximize yield and quality, and achieve sustainable agriculture.

[0006] "Weather data" refers to information about the weather, specifically data such as temperature, precipitation, humidity, wind speed, and wind direction.

[0007] "Soil data" refers to information about the condition of agricultural soil, specifically data such as soil humidity, pH, water content, and nutrient content.

[0008] "Crop growth status" refers to information about the growing conditions of the crop, specifically data such as crop height, leaf color, leaf area, and fruit size.

[0009] "Preprocessing" refers to a series of steps to convert collected raw data into an analyzable form, specifically including filling in missing data, correcting outliers, and removing noise.

[0010] A "generative AI model" is a predictive model trained using machine learning techniques based on collected data. Specifically, it is an algorithmic model used to predict the optimal time and method for harvesting.

[0011] "Means for training" refers to the training process that uses collected pre-processed data to improve the performance of the generative AI model.

[0012] "Optimal harvest time and harvesting method" refers to the optimal timing and method of harvesting predicted by the generative AI model to maximize crop yield and quality.

[0013] "Means of notification" refers to mechanisms for informing farmers of the optimal harvest time and method predicted by the generative AI model, including push notifications, email notifications, alerts on applications, etc.

[0014] "Feedback Data" refers to user input regarding actual harvest results and the effectiveness of suggestions.

[0015] "Retraining" refers to an additional training process to refine the generative AI model based on collected feedback data to improve its accuracy.

[0016] "System" refers to the overall configuration of multiple means that work in conjunction with each other to effectively optimize agricultural operations. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7]FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0018] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0021] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0022] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0028] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0037] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0038] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system can be implemented as follows.

[0039] Data collection

[0040] The server collects weather, soil, and crop growth data via the internet and sensors. Weather data is automatically obtained daily through a weather API, and soil data is received in real time from soil sensors installed in the farmland. Crop growth status is collected from drone photography data and periodically captured image data.

[0041] Data Preprocessing

[0042] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces a single outlier, it is imputed with the average of the other normal values.

[0043] Training generative AI models

[0044] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0045] Predicting optimal harvest times and methods

[0046] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0047] Proposal Notification

[0048] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0049] Gathering feedback and retraining

[0050] The user harvests based on the suggestions and provides feedback on the results. This sends data about the actual results of the harvest and the effectiveness of the suggestions to the server. The server then uses this feedback data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0051] Specific examples

[0052] 1. Data collection example: The server retrieves one-week weather forecast data for Tokyo from the weather API at 5:00 AM every day. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from sensors installed in farmland.

[0053] 2. Preprocessing: The server interpolates the 70% of humidity data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0054] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0055] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0056] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0057] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0058] The above is a specific embodiment for carrying out the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing the yield and quality of their crops.

[0059] The processing flow will be explained below.

[0060] Step 1: Data collection

[0061] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0062] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0063] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0064] Step 2: Data Preprocessing

[0065] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format.

[0066] The server imputes missing data with the mean or median, detects outliers and replaces them with reasonable values.

[0067] The server performs noise removal and creates a preprocessed dataset, for example, by interpolating outliers from a sensor with the average of normal values ​​from other sensors.

[0068] Step 3: Model generation

[0069] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0070] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0071] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0072] Step 4: Optimization

[0073] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0074] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0075] Step 5: Notification

[0076] The server then notifies the farmer of the results of the predictions via push notification, email, or an application.

[0077] The device will then display the received notification to the user, for example, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected."

[0078] Step 6: Feedback

[0079] The user carries out harvesting work based on the suggestions and provides feedback on the results through the application.

[0080] The server collects feedback data on harvest results and the effectiveness of the proposals, specifically on harvest volume, quality, and work efficiency.

[0081] The server retrains the generative AI model based on the collected feedback data to improve the accuracy of the next prediction, including harvest result feedback data.

[0082] The above is the processing procedure of the program of the present invention.

[0083] Example 1

[0084] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0085] Agricultural harvesting depends on complex factors such as weather conditions, soil conditions, and crop growth, making it difficult to properly evaluate these and determine the optimal harvest time and method. Furthermore, methods that rely on experience and intuition are prone to variations in harvest volume and quality, making efficient agricultural management difficult. Furthermore, conventional systems lack the ability to utilize feedback for improvements, making it difficult to improve prediction accuracy.

[0086] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0087] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying farmers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data from farmers and retraining the generative AI model to improve it, means for detecting and imputing outliers and imputing missing data with mean or median values, and means for iteratively training based on past data to evaluate and improve model performance. This makes it possible to improve the accuracy of harvest time and method predictions, maximize harvest volume and quality, and streamline agricultural management.

[0088] "Weather data" refers to data collected based on meteorological observations that indicates atmospheric conditions such as temperature, humidity, precipitation, wind speed, and sunshine hours.

[0089] "Soil data" refers to data measuring the condition of soil in agricultural or cultivated areas, such as humidity, pH, temperature, and nutrient content.

[0090] "Crop growth status" refers to data indicating the growth and health status of the crop under cultivation, such as its height, leaf area, color, and the presence or absence of pests and diseases.

[0091] "Preprocessing" refers to a series of processes that supplement missing values ​​and correct outliers in collected data to prepare it in a format suitable for analysis and model training.

[0092] A "generative AI model" is an artificial intelligence model trained from pre-processed data using machine learning algorithms to automatically perform specific tasks or predictions.

[0093] The "optimal harvest time and harvest method" refers to the most suitable time and method for harvesting a crop, determined by a comprehensive assessment of weather conditions, soil conditions, and the growth state of the crop.

[0094] "Notification" refers to the act or means of transmitting information to inform users of predictions or suggestions.

[0095] "Feedback data" refers to data on the results and evaluation of harvesting work carried out by farmers based on the suggestions.

[0096] "Retraining" is the training process of updating and improving existing generative AI models using new collected data.

[0097] "Outlier detection and imputation" is the process of detecting values ​​in a dataset that fall outside the normal range and, if necessary, replacing them with other reasonable values.

[0098] "Iterative training" is a technique in which a generative AI model is trained multiple times using the same or extended dataset to improve its performance.

[0099] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system is implemented using the following hardware and software.

[0100] Hardware and software used

[0101] Weather data collection method: Daily weather data is collected using a weather API (e.g., OpenWeatherMap API).

[0102] Soil data collection methods: Soil sensors (e.g., soil moisture sensors, pH sensors) are used to collect real-time soil data.

[0103] Methods for collecting crop growth status: Drones and ground cameras are used to photograph crop growth status and collect image data.

[0104] Generative AI models: Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models to predict the optimal time and method of harvesting.

[0105] Data preprocessing: Programs for imputing missing data and detecting and correcting outliers (e.g., Python's Pandas library).

[0106] Explanation of the program processing flow

[0107] Data collection

[0108] The server collects data in the following ways:

[0109] Weather data is automatically collected via a weather API at 5:00 AM every day.

[0110] Soil data is received in real time from soil sensors installed on farmland.

[0111] The growth status of the crops is collected from image data taken by drones on a regular basis and from image data taken by ground cameras.

[0112] Data Preprocessing

[0113] The server pre-processes the collected raw data:

[0114] If there are missing data, they are imputed with the mean or median.

[0115] Anomalies are detected and interpolated with the average of other normal values. For example, if 70% of the data obtained from a soil moisture sensor is abnormal, it is interpolated with the average of other normal values.

[0116] Training generative AI models

[0117] The server trains a generative AI model using the preprocessed dataset:

[0118] Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models based on historical weather, soil, and crop growth data.

[0119] Cross-validation is performed to evaluate the model's performance and hyperparameters are adjusted as needed.

[0120] Predicting optimal harvest times and methods

[0121] The server inputs the latest data into the generative AI model and makes a prediction:

[0122] For example, new weather forecast data and soil moisture data are input to generate a prediction that "the best time to harvest is next Tuesday at 8:00 a.m."

[0123] Proposal Notification

[0124] The server then uses the prediction results to provide farmers with optimal harvesting suggestions:

[0125] Notifications are sent to farmers' devices (smartphones or computers) via push notifications or email.

[0126] For example, make a specific suggestion like, "Please start harvesting next Tuesday at 8:00 a.m. The current soil moisture and weather forecast will result in the best possible harvest."

[0127] Gathering feedback and retraining

[0128] The user harvests based on the suggestions and provides feedback on the results:

[0129] Users input data on harvest yield and quality through the terminal.

[0130] The server uses the collected feedback data to retrain the generative AI model and improve the accuracy of its predictions.

[0131] Specific examples

[0132] Data collection example: Every day at 5:00 AM, the server retrieves one-week weather forecast data from the weather API. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from the soil sensor.

[0133] Preprocessing: The 70% of humidity data detected as abnormal values ​​are interpolated with the average normal value (60%).

[0134] Model generation: Train a generative AI model using weather data, soil data, and crop results from the past five years.

[0135] Optimization: Input the latest data and predict that "8:00 a.m. next Tuesday is the best time to harvest."

[0136] Notification: The prediction results will be sent to the user's smartphone via push notification.

[0137] Feedback: Users harvest and input their results into the app, and the server uses that data to retrain the generative AI model.

[0138] As described above, farmers can use the system to know the optimal harvesting time and method in advance, maximizing harvest volume and quality.

[0139] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0140] Step 1: Data collection

[0141] The server collects weather data, soil data, and crop growth status data. Inputs are data from weather APIs, soil sensors, drones, and ground cameras. Weather data is automatically acquired at 5:00 AM every day, and soil data is received in real time. Crop growth status data is captured periodically. The output of the data collection step is a raw dataset of weather data, soil data, and crop growth status data.

[0142] Step 2: Data Preprocessing

[0143] The server preprocesses the collected raw data. The input is the raw data collected in step 1. First, it imputes missing data with the mean or median, and detects and corrects outliers. For example, if the data from a soil moisture sensor contains an outlier of 70%, it imputes it with the mean of the other normal values. The output is a preprocessed dataset.

[0144] Step 3: Training the generative AI model

[0145] The server trains a generative AI model using the preprocessed dataset. The input is the dataset preprocessed in step 2. Using a machine learning framework (e.g., TensorFlow, PyTorch), the model is trained based on historical weather data, soil data, and crop growth data. Cross-validation is performed during the training process to evaluate and improve the model's performance. The output is a trained generative AI model.

[0146] Step 4: Predict the optimal harvest time and method

[0147] The server inputs the latest data into the generative AI model to predict the optimal harvest time and method. The inputs are the latest weather data, soil data, and growth data. For example, by combining weather data and soil moisture data, a prediction is generated that "the optimal harvest time is 8:00 a.m. next Tuesday." The output is a prediction of the specific harvest time and method.

[0148] Step 5: Proposal Notification

[0149] The server notifies the farmer of the optimal harvesting suggestion based on the prediction results. The input is the prediction result from step 4. The notification is sent to the farmer's device (smartphone or PC) via push notification or email. For example, a specific suggestion may be made such as, "Please start harvesting at 8:00 a.m. next Tuesday. Based on the current soil moisture and weather forecast, the best harvest is expected." The output is the notified suggestion.

[0150] Step 6: Gather feedback and retrain

[0151] The user harvests based on the suggestions and provides feedback on the results. The input is the harvest result data. The user inputs data on the harvest quantity and quality through their terminal, and the server uses the feedback data to retrain the generative AI model. The output is an improved generative AI model.

[0152] (Application example 1)

[0153] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0154] Optimizing harvest times in agriculture is a crucial issue for improving yield and quality. Traditionally, farmers decide when to harvest based on their own experience and intuition, which can lead to errors in judgment and oversights. Furthermore, there are limited concrete methods for improving the efficiency and automating harvesting work. Therefore, there is a need for a system that automatically optimizes harvest times and methods and enables autonomous vehicles to carry out harvesting work based on the results.

[0155] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0156] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining to improve the generative AI model, means for generating a driving plan for an autonomous vehicle based on the prediction and for the autonomous vehicle to automatically start harvesting work based on the driving plan, and means for collecting driving results of the autonomous vehicle as feedback and retraining to improve the generative AI model. This automatically optimizes the harvest time and harvesting method, realizes automation and efficiency of harvesting work, and enables improved harvest volume and quality.

[0157] "Weather data" is information indicating weather conditions such as weather, temperature, precipitation, humidity, and wind speed.

[0158] "Soil data" refers to information about agricultural soil, including soil moisture, pH value, nutrient content, etc.

[0159] "Crop growth status" is information indicating the growth stage and health status of the crop, and includes, for example, germination, growth, flowering, and harvestable status.

[0160] "Preprocessing" refers to the process of preparing raw data in an analyzable format, and includes filling in missing data and correcting outliers.

[0161] A "generative AI model" is an artificial intelligence model trained to use collected data to predict optimal harvest times and methods.

[0162] "Training" refers to the process by which an AI model learns from collected data so that it can make accurate predictions.

[0163] "Prediction" means that a generative AI model infers future situations based on input data.

[0164] "Notification" refers to the method of informing farmers of the prediction results, and includes push notifications and emails.

[0165] An "autonomous vehicle" is a vehicle that operates automatically based on programmed instructions to perform agricultural tasks.

[0166] An "operation plan" is a plan that shows the route and work procedures required for an autonomous vehicle to carry out harvesting work.

[0167] "Feedback" refers to collecting information based on the results of harvesting operations and actual conditions, and using it to improve the system and retrain the AI ​​model.

[0168] "Retraining" is the process of re-learning a generative AI model to improve its performance using newly collected data.

[0169] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. It also has the function of generating operation plans for autonomous vehicles based on the prediction results and automatically carrying out harvesting operations.

[0170] Data collection

[0171] The server uses a weather API to collect weather data. This automatically obtains the latest weather, temperature, precipitation, humidity, wind speed, and other weather conditions every day. It also collects soil data in real time using soil sensors installed in the farmland. This soil data includes soil moisture, pH, and nutrient content. Furthermore, drones are used to capture images of the growing conditions of the crops and collect the data. This allows the growth stage and health of the crops to be understood.

[0172] Data Preprocessing

[0173] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces an abnormal value, it is imputed with the average of the other normal values.

[0174] Training generative AI models

[0175] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0176] Predicting optimal harvest times and methods

[0177] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0178] Proposal Notification

[0179] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0180] Autonomous vehicles perform harvesting operations

[0181] The server generates a driving plan for the autonomous vehicle based on the predictions. Based on this driving plan, the autonomous vehicle automatically starts harvesting work. For example, instructions such as "Harvesting work will automatically start at 8:00 a.m. next Tuesday" are sent to the autonomous vehicle.

[0182] Gathering feedback and retraining

[0183] Users harvest based on the suggestions and provide feedback on the results. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and uses this data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0184] Specific examples

[0185] 1. Data Collection Example: The server retrieves a week's worth of weather data for a city from a weather API at 5:00 AM every day. At the same time, it receives soil moisture data from sensors installed in farmland.

[0186] 2. Data preprocessing: The server interpolates the humidity data detected as abnormal values ​​with the average value of other normal values.

[0187] 3. Model generation: The server trains a generative AI model that uses historical weather data, soil data, and harvest results to predict optimal harvest times.

[0188] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0189] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0190] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0191] Prompt Sentence Examples

[0192] Enter the weather, soil and crop growth data as follows:

[0193] Weather data: Temperature=30, Humidity=70, Precipitation=0

[0194] Soil data: soil moisture=60, pH=6.5

[0195] Crop growth data: Growth stage=vegetative, Image data=base64_encoded_image_string

[0196] By inputting this data into a generative AI model, the optimal harvest time and method can be predicted.

[0197] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0198] Step 1: Data collection

[0199] The server obtains weather data from the weather API and collects soil and crop growth data from soil sensors and drones installed in the farmland. The server sends requests to the weather API to obtain information such as temperature, humidity, and precipitation, and obtains soil humidity and pH data from the soil sensors. It also operates a drone to take images of the crops and collect data on their growth status. The input data consists of weather data, soil sensor data, and crop image data, and these are used to proceed to the next step.

[0200] Step 2: Data Preprocessing

[0201] The server preprocesses the collected weather data, soil data, and crop growth data. If there are missing data or outliers, it fills them in with the mean or median and corrects the outliers. For example, if data from a soil moisture sensor contains an outlier, the server fills it in using the mean value of other normal data. The input data is the collected raw data, which is then processed to output preprocessed data.

[0202] Step 3: Training the generative AI model

[0203] The server uses the preprocessed dataset to train a generative AI model. It uses historical weather, soil, and crop growth data to build a multi-layer neural network and train a model to predict optimal harvest times and methods. The server uses the preprocessed dataset as input data and outputs a trained generative AI model.

[0204] Step 4: Predict harvest time and method

[0205] The server inputs the latest data into a trained generative AI model to predict the optimal harvest time and method. For example, it uses next week's weather data and current soil moisture data to obtain a specific prediction result, such as "The optimal time to harvest is next Tuesday morning." The input data is the latest weather data, soil data, and crop growth data, and the output data is the prediction result of the optimal harvest time and method.

[0206] Step 5: Proposal Notification

[0207] Based on the prediction results, the server notifies the farmer of the optimal harvest time and harvesting method. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, a message could be sent saying, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The input data is the prediction result, and the output data is the notification message.

[0208] Step 6: Generate a driving plan for the autonomous vehicle

[0209] The server generates a driving plan for the autonomous vehicle based on the predictions. The autonomous vehicle then automatically starts harvesting according to the generated driving plan. For example, an instruction such as "Harvesting will begin automatically at 8:00 AM next Tuesday" is sent to the autonomous vehicle. The input data is the predicted results of the optimal harvesting time and method, and the output data is the driving plan and specific instructions for the autonomous vehicle.

[0210] Step 7: Gather feedback

[0211] The user harvests based on the suggestions and sends the results as feedback to the server. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and retrains the generative AI model to improve future prediction accuracy. The input data are the harvest results and feedback data, and the output data is an improved generative AI model.

[0212] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0213] This invention is a system for optimizing agricultural harvesting operations, combining a generative AI model and an emotion engine to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. This system recognizes the user's emotions and uses them as feedback, enabling even more accurate suggestions.

[0214] Data collection

[0215] The server obtains weather data daily through a weather API. The obtained data includes temperature, precipitation, humidity, wind speed, and wind direction. The server also receives soil data in real time from soil sensors. The soil data includes soil humidity, pH, moisture content, and nutrient content. The server also collects information on the growth status of crops from drone photography data and periodically captured image data. Specifically, the image data includes crop height, leaf color, leaf area, and fruit size.

[0216] Data Preprocessing

[0217] The server converts the collected raw data into an analyzable form. For example, it standardizes the acquired data format and fills in missing data with the mean or median. Outliers are detected and replaced with reasonable values. It also performs noise removal and creates a preprocessed data set. For example, it fills in outliers obtained from a soil moisture sensor with the mean of other normal values.

[0218] Training generative AI models

[0219] The server splits the preprocessed dataset into a training dataset and a test dataset and trains a generative AI model that uses historical weather, soil, and crop growth data to predict the optimal time and method of harvesting. Machine learning techniques are used for training, and iterative training is performed to evaluate and improve the model's performance.

[0220] Predicting optimal harvest times and methods

[0221] The server inputs the latest weather, soil, and crop growth data into the generative AI model to predict the optimal harvest time and method, generating specific predictions such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0222] Collecting and analyzing user emotion data

[0223] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine how the user feels about the harvesting suggestions (satisfaction, dissatisfaction, stress, etc.).

[0224] Proposal Notification

[0225] The server then uses the prediction results and emotion data to provide farmers with optimal harvest suggestions. These suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected." The tone and details of the suggestion may also be adjusted based on the user's emotion data.

[0226] Gathering feedback and retraining

[0227] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and uses it to retrain the generative AI model, further improving the accuracy of the next prediction.

[0228] Specific examples

[0229] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0230] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0231] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0232] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0233] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0234] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0235] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0236] The above is a specific example of how to implement the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotions also increase the likelihood of suggestions being accepted.

[0237] The processing flow will be explained below.

[0238] Program processing

[0239] Step 1: Data collection

[0240] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0241] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0242] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0243] Step 2: Data Preprocessing

[0244] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format and filling in missing data with the mean or median.

[0245] The server detects outliers and replaces them with reasonable values, for example, interpolating outliers from a soil moisture sensor with the average of other normal values.

[0246] The server performs noise removal and creates a preprocessed dataset.

[0247] Step 3: Model generation

[0248] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0249] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0250] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0251] Step 4: Optimization

[0252] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0253] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0254] Step 5: Collecting sentiment data

[0255] The server collects user emotion data using an emotion engine, which uses voice and face recognition technology to determine emotions from the user's tone of voice and facial expressions.

[0256] The device analyzes the acquired emotional data and evaluates the user's stress level and satisfaction.

[0257] Step 6: Generate and send notifications

[0258] The server then generates optimal harvesting suggestions for the user based on the prediction results and emotion data, such as "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best harvest is expected."

[0259] The server sends the generated suggestions to the user's device via push notification, email, or application.

[0260] Step 7: Gather feedback and retrain

[0261] The user harvests based on the suggestions and provides feedback on the results through the application, with data on the harvest results and the effectiveness of the suggestions being sent to the server.

[0262] The server retrains the generative AI model based on the collected feedback and emotion data to improve the accuracy of the next prediction.

[0263] Specific examples

[0264] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0265] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0266] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0267] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0268] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0269] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0270] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0271] The above is a specific embodiment for carrying out the present invention.

[0272] Example 2

[0273] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0274] In agriculture, determining the optimal harvest time and method is extremely important, but it is difficult to make accurate predictions due to the influence of numerous factors, such as weather conditions, soil conditions, and crop growth status. Furthermore, adjusting suggestions based on the farmer's emotions and stress levels is also essential to improving work efficiency. There is a need for a system that can comprehensively analyze these complex factors and make optimal harvest suggestions for farmers.

[0275] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0276] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvest method using the trained generative AI model, means for collecting and analyzing user emotion data, means for notifying farmers of the optimal harvest time and harvest method based on the prediction and the user emotion data, and means for collecting feedback data and emotion data and retraining to improve the generative AI model. This enables accurate harvest predictions based on the collected data and individual responses according to farmers' emotions, thereby maximizing harvest volume and quality and improving work efficiency.

[0277] "Weather data" refers to various information about weather conditions, such as temperature, precipitation, humidity, wind speed, and wind direction.

[0278] "Soil data" refers to information that indicates the condition of the soil, such as soil humidity, pH, water content, and nutrient content.

[0279] "Crop growth status" refers to information about the physical growth of the crop, such as crop height, leaf color, leaf area, and fruit size.

[0280] "Preprocessing" refers to converting collected data into an analyzable form, standardizing data formats, filling in missing data, detecting and replacing outliers, and removing noise.

[0281] A "generative AI model" is a model trained using machine learning techniques and is an algorithm that predicts the optimal harvest time and method based on collected data.

[0282] "User emotion data" refers to data related to emotions obtained from the user's tone of voice and facial expressions determined using voice recognition and facial recognition technology.

[0283] "Feedback data" refers to data on the results of actual harvesting work and the effectiveness of suggestions.

[0284] This invention is a system for optimizing agricultural harvesting operations. The system combines a generative AI model and an emotion engine to analyze weather, soil, and crop growth data to suggest optimal harvest times and methods to farmers.

[0285] Data collection

[0286] The server obtains weather data daily through a weather API. This data includes temperature, precipitation, humidity, wind speed, and wind direction. For example, the temperature might be 20°C, the probability of precipitation 40%, and humidity 70%. The server also receives real-time soil data from soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. For example, the server verifies that the soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. The server also collects crop growth status data from drone photography. Specifically, this data includes image data such as crop height, leaf color, leaf area, and fruit size. For example, the server verifies that the crop height is 30 cm, leaf color is a healthy green, leaf area is 15 square centimeters, and fruit size is 5 cm.

[0287] Data Preprocessing

[0288] The server converts the collected raw data into an analyzable form. First, it standardizes the data format and fills in missing data with the average or median. For example, missing temperature data for one day is filled in with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, it fills in the 70% soil humidity data detected as an outlier with the average value (60%) of the other normal values. It also removes noise to create a preprocessed dataset.

[0289] Training generative AI models

[0290] The server splits the preprocessed dataset into training data and test data and trains a generative AI model. This model is used to predict the optimal harvest time and method based on historical weather, soil, and crop growth data. Machine learning techniques are used for training, and iterative training is performed to achieve high accuracy. For example, the model is trained using weather data, soil data, and harvest results from the past five years.

[0291] Predicting optimal harvest times and methods

[0292] The server inputs the latest weather, soil, and crop growth data into the generative AI model, which predicts the optimal harvest time and method. The generative AI model then provides specific predictions, such as "The optimal harvest time is next Tuesday at 8:00 AM."

[0293] Collecting and analyzing user emotion data

[0294] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine the user's emotion (satisfaction, dissatisfaction, stress, etc.) in response to the harvesting suggestion.

[0295] Proposal Notification

[0296] The server then sends farmers optimal harvesting suggestions based on the predictions of the optimal harvesting time and method, as well as the user's emotional data. The suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best possible harvest is expected." The tone and detailed information of the suggestions may also be adjusted based on the user's emotional data.

[0297] Gathering feedback and retraining

[0298] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and retrains the generative AI model to further improve prediction accuracy next time.

[0299] This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on user emotion data also increase the likelihood of suggestions being accepted.

[0300] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0301] Step 1:

[0302] The server collects weather data through the weather API at 5:00 AM every day. Specifically, it obtains data such as temperature, precipitation, humidity, wind speed, and wind direction. It receives this weather data as input and stores it in a database. For example, it obtains data such as a temperature of 20 degrees, a 40% chance of precipitation, and a 70% humidity. As output, the weather data is recorded in the database.

[0303] Step 2:

[0304] The server collects soil data in real time from soil sensors installed in farmland. Specifically, data such as soil humidity, pH, moisture content, and nutrient content are acquired. The server receives the soil data as input and stores it in a database. For example, data is collected showing that soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. As output, the soil data is recorded in the database.

[0305] Step 3:

[0306] The server uses drone image data to collect information on the growth status of crops. Specifically, it analyzes image data such as crop height, leaf color, leaf area, and fruit size. It receives this crop growth data as input and stores it in a database. For example, it obtains data such as a crop height of 30 cm, a healthy green leaf color, a leaf area of ​​15 square centimeters, and a fruit size of 5 cm. As output, the crop growth data is recorded in the database.

[0307] Step 4:

[0308] The server preprocesses the collected weather, soil, and crop growth data. First, it standardizes the data format and imputes missing data with the mean or median. For example, missing temperature data for one day is imputed with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, soil moisture data detected as an outlier at 70% is imputed with 60%, the average of the other normal values. This preprocessed data is received as input and noise is removed. A preprocessed dataset is created as output.

[0309] Step 5:

[0310] The server splits the preprocessed dataset into training data and test data and trains the generative AI model. For example, it uses weather data, soil data, and harvest results from the past five years. It receives the training data as input and trains the AI ​​model. It then evaluates the model's performance using the test data. The output is a highly accurate generative AI model.

[0311] Step 6:

[0312] The server inputs the latest weather, soil, and crop growth data into the AI ​​model to predict the optimal harvest time and method. For example, it generates a specific prediction result, such as "8:00 AM next Tuesday is the optimal harvest time." It receives this prediction result as input and generates information to notify farmers. The output is the predicted harvest time and method.

[0313] Step 7:

[0314] The server uses an emotion engine to collect and analyze the user's emotional data. Specifically, it uses voice recognition and face recognition technology to determine the user's emotions from their tone of voice and facial expressions. It receives the emotional data as input and analyzes the user's emotional state. The output is the user's emotional data.

[0315] Step 8:

[0316] The server notifies the farmer of optimal harvesting suggestions based on the prediction results of the optimal harvesting time and method and the user's emotional data. Specifically, the suggestion is sent via push notification or email to the farmer's device (smartphone, PC, etc.). For example, the suggestion might be, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The suggestion notification is sent to the farmer as output.

[0317] Step 9:

[0318] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The feedback data is received as input and used to improve the AI ​​model. The output is a retrained generative AI model.

[0319] The above is a specific example of how to implement the invention. Using this system, farmers can predict the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotional data also increase the likelihood of their suggestions being accepted.

[0320] (Application example 2)

[0321] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0322] In conventional agricultural harvesting work, determining the optimal harvesting time and method is difficult, and in many cases workers have had to rely on experience and intuition. As a result, harvest volume and quality are unstable, making it difficult for agricultural workers to harvest at the appropriate time. Furthermore, when harvesting suggestions are simply notified mechanically, they do not take into account the emotions and reactions of agricultural workers, making them less likely to accept the suggestions. This has led to a demand for a system that can provide optimal harvesting suggestions and reduce the burden on agricultural workers.

[0323] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth status, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining the generative AI model to improve it, means for recognizing the emotions of users who receive the prediction notification and collecting emotion data, means for adjusting the notification content based on the emotion data, and means for using the collected emotion data to improve the generative AI model. This makes it possible to simultaneously optimize harvesting operations and increase the likelihood that agricultural workers will accept suggestions based on their emotions.

[0324] "Weather data" refers to information related to weather, including data such as temperature, precipitation, humidity, wind speed, and wind direction.

[0325] "Soil data" refers to information about the condition of the soil, including data on soil humidity, pH, water content, nutrient content, etc.

[0326] "Crop growth status" is information that indicates how the crop is growing, and includes data such as the height of the crop, leaf color, leaf area, and fruit size.

[0327] A "generative AI model" is an artificial intelligence model that is trained on past data to predict specific outcomes.

[0328] "Emotional data" refers to information about a user's emotional state, including data derived from vocal tone and facial expressions.

[0329] The "means for adjusting the notification content" is a means for changing the notification content regarding the optimal harvesting time and harvesting method based on the user's emotional data.

[0330] "Feedback Data" refers to information regarding actual harvest results and the effectiveness of recommendations, including data collected from users.

[0331] The "training means" is the process of training a generative AI model using preprocessed data.

[0332] "Preprocessing" refers to the process of converting the raw data acquired into an analyzable form.

[0333] "Retraining" is a training process that uses collected feedback data to improve the accuracy of an existing generative AI model.

[0334] The present invention is a system for highly optimizing agricultural harvesting operations, and is realized by including the following means.

[0335] First, the server collects weather data (temperature, precipitation, humidity, wind speed, wind direction, etc.) using a weather API at a specific time each day. Next, soil data is collected in real time using soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. Additionally, cameras and drones installed in the farmland are used to collect information on the growth status of the crops (plant height, leaf color, leaf area, fruit size, etc.).

[0336] The server preprocesses the collected data. Specifically, it uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, unify multiple data formats, and remove noise.

[0337] The preprocessed data is then used to train a generative AI model using machine learning frameworks such as TensorFlow and PyTorch, based on historical weather, soil, and crop growth data, which is then used to predict the optimal harvest time and method.

[0338] Using the trained generative AI model, the server inputs the latest data and predicts the best time and method for harvesting, generating a specific prediction result, for example, "The best time to harvest is next Tuesday at 8:00 AM."

[0339] In addition, the server uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to collect the user's emotional data. The server analyzes the user's tone of voice and facial expressions when receiving the notification and collects emotional data. Based on this emotional data, the server determines the user's emotional state, such as "comfortable," "anxious," or "confused," and adjusts the content of the notification accordingly.

[0340] Finally, the server notifies the farmer of the best time and method to harvest. The notification is sent via push notification or email. The content of the notification is adjusted based on the sentiment data, so it is presented in a way that is easy for the farmer to accept and understand.

[0341] Users harvest based on the suggestions and provide feedback on the results via their smartphone or computer. This feedback data is sent to the server and used to retrain the existing generative AI model, allowing the system to further improve its prediction accuracy in the future.

[0342] For example, a weather API retrieves the weather forecast for the coming week at 8:00 a.m. every morning, and a soil sensor collects soil moisture data every hour. The user's reaction to the harvest suggestion can be collected and the next notification can be tailored based on that reaction. A prompt such as "Please retrieve the weather forecast for next week and collect data to predict the optimal harvest time" can be used to create input data for a generative AI model.

[0343] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0344] Step 1:

[0345] The server collects weather data. Specifically, it uses a weather API to obtain data such as temperature, precipitation, humidity, wind speed, and wind direction at a specific time each day. This data is stored on the server in JSON format. The input is raw data obtained from the weather API, and the output is weather data in a unified JSON format.

[0346] Step 2:

[0347] The server collects soil data. Soil sensors acquire data such as soil humidity, pH, moisture content, and nutrient content in real time and send it to the server. This data is stored in the server in RAW format. The input is raw data from the soil sensors, and the output is formatted soil data.

[0348] Step 3:

[0349] The server collects information on the growth status of crops. It analyzes image data taken using cameras and drones installed on farmland to extract information such as crop height, leaf color, leaf area, and fruit size. It preprocesses the acquired image data to generate the necessary growth data. The input is image data from the cameras and drones, and the output is analyzed growth data.

[0350] Step 4:

[0351] The server integrates preprocessed weather data, soil data, and crop growth data. It uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, and unify multiple data formats. The input is individual data before preprocessing, and the output is the integrated, preprocessed data.

[0352] Step 5:

[0353] The server trains a generative AI model based on the preprocessed data. Using TensorFlow and PyTorch, the model is trained based on historical weather, soil, and crop growth data to generate a model that can predict the optimal harvest time and method. The input is the preprocessed data, and the output is the trained generative AI model.

[0354] Step 6:

[0355] The server uses a generative AI model to predict the optimal harvest time and method. The latest data is input into the model to generate a specific prediction. For example, the output may be "The optimal harvest time is 8:00 a.m. next Tuesday." The input is the latest data and a trained generative AI model, and the output is a prediction of the harvest time and method.

[0356] Step 7:

[0357] The server collects the user's emotional data. It uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to analyze the user's tone of voice and facial expression when they receive the suggestion notification. This emotional data is stored on the server. The input is the user's voice and facial image data, and the output is the analyzed emotional data.

[0358] Step 8:

[0359] The server adjusts the notification content based on the emotional data. Based on the collected emotional data, it modifies the suggestions to make them more acceptable to the user, providing the optimal tone and detailed information. The input is the emotional data and prediction results, and the output is the adjusted notification content.

[0360] Step 9:

[0361] The server notifies the farmer of the best time and method to harvest, sending specific suggestions via push notification or email. For example, it might say, "Start harvesting next Tuesday morning. Current soil moisture and forecasted weather suggest the best possible harvest." The input is the tailored notification content, and the output is the notification message.

[0362] Step 10:

[0363] The user harvests based on the suggestions and provides feedback on the results. Feedback data is sent to the server via smartphone or PC. The input is the user's harvest results and information on the validity of the suggestions, and the output is new feedback data.

[0364] Step 11:

[0365] The server uses the collected feedback data to retrain the generative AI model. By adding new data, the accuracy of the model improves. The input is the feedback data, and the output is an improved generative AI model.

[0366] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0367] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0368] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0369] [Second embodiment]

[0370] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0371] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0372] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0373] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0374] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0375] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0376] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0377] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0378] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0379] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0380] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0381] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0382] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system can be implemented as follows.

[0383] Data collection

[0384] The server collects weather, soil, and crop growth data via the internet and sensors. Weather data is automatically obtained daily through a weather API, and soil data is received in real time from soil sensors installed in the farmland. Crop growth status is collected from drone photography data and periodically captured image data.

[0385] Data Preprocessing

[0386] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces a single outlier, it is imputed with the average of the other normal values.

[0387] Training generative AI models

[0388] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0389] Predicting optimal harvest times and methods

[0390] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0391] Proposal Notification

[0392] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0393] Gathering feedback and retraining

[0394] The user harvests based on the suggestions and provides feedback on the results. This sends data about the actual results of the harvest and the effectiveness of the suggestions to the server. The server then uses this feedback data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0395] Specific examples

[0396] 1. Data collection example: The server retrieves one-week weather forecast data for Tokyo from the weather API at 5:00 AM every day. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from sensors installed in farmland.

[0397] 2. Preprocessing: The server interpolates the 70% of humidity data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0398] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0399] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0400] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0401] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0402] The above is a specific embodiment for carrying out the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing the yield and quality of their crops.

[0403] The processing flow will be explained below.

[0404] Step 1: Data collection

[0405] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0406] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0407] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0408] Step 2: Data Preprocessing

[0409] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format.

[0410] The server imputes missing data with the mean or median, detects outliers and replaces them with reasonable values.

[0411] The server performs noise removal and creates a preprocessed dataset, for example, by interpolating outliers from a sensor with the average of normal values ​​from other sensors.

[0412] Step 3: Model generation

[0413] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0414] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0415] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0416] Step 4: Optimization

[0417] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0418] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0419] Step 5: Notification

[0420] The server then notifies the farmer of the results of the predictions via push notification, email, or an application.

[0421] The device will then display the received notification to the user, for example, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected."

[0422] Step 6: Feedback

[0423] The user carries out harvesting work based on the suggestions and provides feedback on the results through the application.

[0424] The server collects feedback data on harvest results and the effectiveness of the proposals, specifically on harvest volume, quality, and work efficiency.

[0425] The server retrains the generative AI model based on the collected feedback data to improve the accuracy of the next prediction, including harvest result feedback data.

[0426] The above is the processing procedure of the program of the present invention.

[0427] Example 1

[0428] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0429] Agricultural harvesting depends on complex factors such as weather conditions, soil conditions, and crop growth, making it difficult to properly evaluate these and determine the optimal harvest time and method. Furthermore, methods that rely on experience and intuition are prone to variations in harvest volume and quality, making efficient agricultural management difficult. Furthermore, conventional systems lack the ability to utilize feedback for improvements, making it difficult to improve prediction accuracy.

[0430] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0431] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying farmers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data from farmers and retraining the generative AI model to improve it, means for detecting and imputing outliers and imputing missing data with mean or median values, and means for iteratively training based on past data to evaluate and improve model performance. This makes it possible to improve the accuracy of harvest time and method predictions, maximize harvest volume and quality, and streamline agricultural management.

[0432] "Weather data" refers to data collected based on meteorological observations that indicates atmospheric conditions such as temperature, humidity, precipitation, wind speed, and sunshine hours.

[0433] "Soil data" refers to data measuring the condition of soil in agricultural or cultivated areas, such as humidity, pH, temperature, and nutrient content.

[0434] "Crop growth status" refers to data indicating the growth and health status of the crop under cultivation, such as its height, leaf area, color, and the presence or absence of pests and diseases.

[0435] "Preprocessing" refers to a series of processes that supplement missing values ​​and correct outliers in collected data to prepare it in a format suitable for analysis and model training.

[0436] A "generative AI model" is an artificial intelligence model trained from pre-processed data using machine learning algorithms to automatically perform specific tasks or predictions.

[0437] The "optimal harvest time and harvest method" refers to the most suitable time and method for harvesting a crop, determined by a comprehensive assessment of weather conditions, soil conditions, and the growth state of the crop.

[0438] "Notification" refers to the act or means of transmitting information to inform users of predictions or suggestions.

[0439] "Feedback data" refers to data on the results and evaluation of harvesting work carried out by farmers based on the suggestions.

[0440] "Retraining" is the training process of updating and improving existing generative AI models using new collected data.

[0441] "Outlier detection and imputation" is the process of detecting values ​​in a dataset that fall outside the normal range and, if necessary, replacing them with other reasonable values.

[0442] "Iterative training" is a technique in which a generative AI model is trained multiple times using the same or extended dataset to improve its performance.

[0443] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system is implemented using the following hardware and software.

[0444] Hardware and software used

[0445] Weather data collection method: Daily weather data is collected using a weather API (e.g., OpenWeatherMap API).

[0446] Soil data collection methods: Soil sensors (e.g., soil moisture sensors, pH sensors) are used to collect real-time soil data.

[0447] Methods for collecting crop growth status: Drones and ground cameras are used to photograph crop growth status and collect image data.

[0448] Generative AI models: Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models to predict the optimal time and method of harvesting.

[0449] Data preprocessing: Programs for imputing missing data and detecting and correcting outliers (e.g., Python's Pandas library).

[0450] Explanation of the program processing flow

[0451] Data collection

[0452] The server collects data in the following ways:

[0453] Weather data is automatically collected via a weather API at 5:00 AM every day.

[0454] Soil data is received in real time from soil sensors installed on farmland.

[0455] The growth status of the crops is collected from image data taken by drones on a regular basis and from image data taken by ground cameras.

[0456] Data Preprocessing

[0457] The server pre-processes the collected raw data:

[0458] If there are missing data, they are imputed with the mean or median.

[0459] Anomalies are detected and interpolated with the average of other normal values. For example, if 70% of the data obtained from a soil moisture sensor is abnormal, it is interpolated with the average of other normal values.

[0460] Training generative AI models

[0461] The server trains a generative AI model using the preprocessed dataset:

[0462] Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models based on historical weather, soil, and crop growth data.

[0463] Cross-validation is performed to evaluate the model's performance and hyperparameters are adjusted as needed.

[0464] Predicting optimal harvest times and methods

[0465] The server inputs the latest data into the generative AI model and makes a prediction:

[0466] For example, new weather forecast data and soil moisture data are input to generate a prediction that "the best time to harvest is next Tuesday at 8:00 a.m."

[0467] Proposal Notification

[0468] The server then uses the prediction results to provide farmers with optimal harvesting suggestions:

[0469] Notifications are sent to farmers' devices (smartphones or computers) via push notifications or email.

[0470] For example, make a specific suggestion like, "Please start harvesting next Tuesday at 8:00 a.m. The current soil moisture and weather forecast will result in the best possible harvest."

[0471] Gathering feedback and retraining

[0472] The user harvests based on the suggestions and provides feedback on the results:

[0473] Users input data on harvest yield and quality through the terminal.

[0474] The server uses the collected feedback data to retrain the generative AI model and improve the accuracy of its predictions.

[0475] Specific examples

[0476] Data collection example: Every day at 5:00 AM, the server retrieves one-week weather forecast data from the weather API. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from the soil sensor.

[0477] Preprocessing: The 70% of humidity data detected as abnormal values ​​are interpolated with the average normal value (60%).

[0478] Model generation: Train a generative AI model using weather data, soil data, and crop results from the past five years.

[0479] Optimization: Input the latest data and predict that "8:00 a.m. next Tuesday is the best time to harvest."

[0480] Notification: The prediction results will be sent to the user's smartphone via push notification.

[0481] Feedback: Users harvest and input their results into the app, and the server uses that data to retrain the generative AI model.

[0482] As described above, farmers can use the system to know the optimal harvesting time and method in advance, maximizing harvest volume and quality.

[0483] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0484] Step 1: Data collection

[0485] The server collects weather data, soil data, and crop growth status data. Inputs are data from weather APIs, soil sensors, drones, and ground cameras. Weather data is automatically acquired at 5:00 AM every day, and soil data is received in real time. Crop growth status data is captured periodically. The output of the data collection step is a raw dataset of weather data, soil data, and crop growth status data.

[0486] Step 2: Data Preprocessing

[0487] The server preprocesses the collected raw data. The input is the raw data collected in step 1. First, it imputes missing data with the mean or median, and detects and corrects outliers. For example, if the data from a soil moisture sensor contains an outlier of 70%, it imputes it with the mean of the other normal values. The output is a preprocessed dataset.

[0488] Step 3: Training the generative AI model

[0489] The server trains a generative AI model using the preprocessed dataset. The input is the dataset preprocessed in step 2. Using a machine learning framework (e.g., TensorFlow, PyTorch), the model is trained based on historical weather data, soil data, and crop growth data. Cross-validation is performed during the training process to evaluate and improve the model's performance. The output is a trained generative AI model.

[0490] Step 4: Predict the optimal harvest time and method

[0491] The server inputs the latest data into the generative AI model to predict the optimal harvest time and method. The inputs are the latest weather data, soil data, and growth data. For example, by combining weather data and soil moisture data, a prediction is generated that "the optimal harvest time is 8:00 a.m. next Tuesday." The output is a prediction of the specific harvest time and method.

[0492] Step 5: Proposal Notification

[0493] The server notifies the farmer of the optimal harvesting suggestion based on the prediction results. The input is the prediction result from step 4. The notification is sent to the farmer's device (smartphone or PC) via push notification or email. For example, a specific suggestion may be made such as, "Please start harvesting at 8:00 a.m. next Tuesday. Based on the current soil moisture and weather forecast, the best harvest is expected." The output is the notified suggestion.

[0494] Step 6: Gather feedback and retrain

[0495] The user harvests based on the suggestions and provides feedback on the results. The input is the harvest result data. The user inputs data on the harvest quantity and quality through their terminal, and the server uses the feedback data to retrain the generative AI model. The output is an improved generative AI model.

[0496] (Application example 1)

[0497] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0498] Optimizing harvest times in agriculture is a crucial issue for improving yield and quality. Traditionally, farmers decide when to harvest based on their own experience and intuition, which can lead to errors in judgment and oversights. Furthermore, there are limited concrete methods for improving the efficiency and automating harvesting work. Therefore, there is a need for a system that automatically optimizes harvest times and methods and enables autonomous vehicles to carry out harvesting work based on the results.

[0499] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0500] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining to improve the generative AI model, means for generating a driving plan for an autonomous vehicle based on the prediction and for the autonomous vehicle to automatically start harvesting work based on the driving plan, and means for collecting driving results of the autonomous vehicle as feedback and retraining to improve the generative AI model. This automatically optimizes the harvest time and harvesting method, realizes automation and efficiency of harvesting work, and enables improved harvest volume and quality.

[0501] "Weather data" is information indicating weather conditions such as weather, temperature, precipitation, humidity, and wind speed.

[0502] "Soil data" refers to information about agricultural soil, including soil moisture, pH value, nutrient content, etc.

[0503] "Crop growth status" is information indicating the growth stage and health status of the crop, and includes, for example, germination, growth, flowering, and harvestable status.

[0504] "Preprocessing" refers to the process of preparing raw data in an analyzable format, and includes filling in missing data and correcting outliers.

[0505] A "generative AI model" is an artificial intelligence model trained to use collected data to predict optimal harvest times and methods.

[0506] "Training" refers to the process by which an AI model learns from collected data so that it can make accurate predictions.

[0507] "Prediction" means that a generative AI model infers future situations based on input data.

[0508] "Notification" refers to the method of informing farmers of the prediction results, and includes push notifications and emails.

[0509] An "autonomous vehicle" is a vehicle that operates automatically based on programmed instructions to perform agricultural tasks.

[0510] An "operation plan" is a plan that shows the route and work procedures required for an autonomous vehicle to carry out harvesting work.

[0511] "Feedback" refers to collecting information based on the results of harvesting operations and actual conditions, and using it to improve the system and retrain the AI ​​model.

[0512] "Retraining" is the process of re-learning a generative AI model to improve its performance using newly collected data.

[0513] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. It also has the function of generating operation plans for autonomous vehicles based on the prediction results and automatically carrying out harvesting operations.

[0514] Data collection

[0515] The server uses a weather API to collect weather data. This automatically obtains the latest weather, temperature, precipitation, humidity, wind speed, and other weather conditions every day. It also collects soil data in real time using soil sensors installed in the farmland. This soil data includes soil moisture, pH, and nutrient content. Furthermore, drones are used to capture images of the growing conditions of the crops and collect the data. This allows the growth stage and health of the crops to be understood.

[0516] Data Preprocessing

[0517] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces an abnormal value, it is imputed with the average of the other normal values.

[0518] Training generative AI models

[0519] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0520] Predicting optimal harvest times and methods

[0521] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0522] Proposal Notification

[0523] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0524] Autonomous vehicles perform harvesting operations

[0525] The server generates a driving plan for the autonomous vehicle based on the predictions. Based on this driving plan, the autonomous vehicle automatically starts harvesting work. For example, instructions such as "Harvesting work will automatically start at 8:00 a.m. next Tuesday" are sent to the autonomous vehicle.

[0526] Gathering feedback and retraining

[0527] Users harvest based on the suggestions and provide feedback on the results. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and uses this data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0528] Specific examples

[0529] 1. Data Collection Example: The server retrieves a week's worth of weather data for a city from a weather API at 5:00 AM every day. At the same time, it receives soil moisture data from sensors installed in farmland.

[0530] 2. Data preprocessing: The server interpolates the humidity data detected as abnormal values ​​with the average value of other normal values.

[0531] 3. Model generation: The server trains a generative AI model that uses historical weather data, soil data, and harvest results to predict optimal harvest times.

[0532] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0533] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0534] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0535] Prompt Sentence Examples

[0536] Enter the weather, soil and crop growth data as follows:

[0537] Weather data: Temperature=30, Humidity=70, Precipitation=0

[0538] Soil data: soil moisture=60, pH=6.5

[0539] Crop growth data: Growth stage=vegetative, Image data=base64_encoded_image_string

[0540] By inputting this data into a generative AI model, the optimal harvest time and method can be predicted.

[0541] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0542] Step 1: Data collection

[0543] The server obtains weather data from the weather API and collects soil and crop growth data from soil sensors and drones installed in the farmland. The server sends requests to the weather API to obtain information such as temperature, humidity, and precipitation, and obtains soil humidity and pH data from the soil sensors. It also operates a drone to take images of the crops and collect data on their growth status. The input data consists of weather data, soil sensor data, and crop image data, and these are used to proceed to the next step.

[0544] Step 2: Data Preprocessing

[0545] The server preprocesses the collected weather data, soil data, and crop growth data. If there are missing data or outliers, it fills them in with the mean or median and corrects the outliers. For example, if data from a soil moisture sensor contains an outlier, the server fills it in using the mean value of other normal data. The input data is the collected raw data, which is then processed to output preprocessed data.

[0546] Step 3: Training the generative AI model

[0547] The server uses the preprocessed dataset to train a generative AI model. It uses historical weather, soil, and crop growth data to build a multi-layer neural network and train a model to predict optimal harvest times and methods. The server uses the preprocessed dataset as input data and outputs a trained generative AI model.

[0548] Step 4: Predict harvest time and method

[0549] The server inputs the latest data into a trained generative AI model to predict the optimal harvest time and method. For example, it uses next week's weather data and current soil moisture data to obtain a specific prediction result, such as "The optimal time to harvest is next Tuesday morning." The input data is the latest weather data, soil data, and crop growth data, and the output data is the prediction result of the optimal harvest time and method.

[0550] Step 5: Proposal Notification

[0551] Based on the prediction results, the server notifies the farmer of the optimal harvest time and harvesting method. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, a message could be sent saying, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The input data is the prediction result, and the output data is the notification message.

[0552] Step 6: Generate a driving plan for the autonomous vehicle

[0553] The server generates a driving plan for the autonomous vehicle based on the predictions. The autonomous vehicle then automatically starts harvesting according to the generated driving plan. For example, an instruction such as "Harvesting will begin automatically at 8:00 AM next Tuesday" is sent to the autonomous vehicle. The input data is the predicted results of the optimal harvesting time and method, and the output data is the driving plan and specific instructions for the autonomous vehicle.

[0554] Step 7: Gather feedback

[0555] The user harvests based on the suggestions and sends the results as feedback to the server. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and retrains the generative AI model to improve future prediction accuracy. The input data are the harvest results and feedback data, and the output data is an improved generative AI model.

[0556] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0557] This invention is a system for optimizing agricultural harvesting operations, combining a generative AI model and an emotion engine to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. This system recognizes the user's emotions and uses them as feedback, enabling even more accurate suggestions.

[0558] Data collection

[0559] The server obtains weather data daily through a weather API. The obtained data includes temperature, precipitation, humidity, wind speed, and wind direction. The server also receives soil data in real time from soil sensors. The soil data includes soil humidity, pH, moisture content, and nutrient content. The server also collects information on the growth status of crops from drone photography data and periodically captured image data. Specifically, the image data includes crop height, leaf color, leaf area, and fruit size.

[0560] Data Preprocessing

[0561] The server converts the collected raw data into an analyzable form. For example, it standardizes the acquired data format and fills in missing data with the mean or median. Outliers are detected and replaced with reasonable values. It also performs noise removal and creates a preprocessed data set. For example, it fills in outliers obtained from a soil moisture sensor with the mean of other normal values.

[0562] Training generative AI models

[0563] The server splits the preprocessed dataset into a training dataset and a test dataset and trains a generative AI model that uses historical weather, soil, and crop growth data to predict the optimal time and method of harvesting. Machine learning techniques are used for training, and iterative training is performed to evaluate and improve the model's performance.

[0564] Predicting optimal harvest times and methods

[0565] The server inputs the latest weather, soil, and crop growth data into the generative AI model to predict the optimal harvest time and method, generating specific predictions such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0566] Collecting and analyzing user emotion data

[0567] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine how the user feels about the harvesting suggestions (satisfaction, dissatisfaction, stress, etc.).

[0568] Proposal Notification

[0569] The server then uses the prediction results and emotion data to provide farmers with optimal harvest suggestions. These suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected." The tone and details of the suggestion may also be adjusted based on the user's emotion data.

[0570] Gathering feedback and retraining

[0571] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and uses it to retrain the generative AI model, further improving the accuracy of the next prediction.

[0572] Specific examples

[0573] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0574] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0575] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0576] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0577] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0578] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0579] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0580] The above is a specific example of how to implement the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotions also increase the likelihood of suggestions being accepted.

[0581] The processing flow will be explained below.

[0582] Program processing

[0583] Step 1: Data collection

[0584] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0585] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0586] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0587] Step 2: Data Preprocessing

[0588] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format and filling in missing data with the mean or median.

[0589] The server detects outliers and replaces them with reasonable values, for example, interpolating outliers from a soil moisture sensor with the average of other normal values.

[0590] The server performs noise removal and creates a preprocessed dataset.

[0591] Step 3: Model generation

[0592] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0593] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0594] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0595] Step 4: Optimization

[0596] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0597] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0598] Step 5: Collecting sentiment data

[0599] The server collects user emotion data using an emotion engine, which uses voice and face recognition technology to determine emotions from the user's tone of voice and facial expressions.

[0600] The device analyzes the acquired emotional data and evaluates the user's stress level and satisfaction.

[0601] Step 6: Generate and send notifications

[0602] The server then generates optimal harvesting suggestions for the user based on the prediction results and emotion data, such as "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best harvest is expected."

[0603] The server sends the generated suggestions to the user's device via push notification, email, or application.

[0604] Step 7: Gather feedback and retrain

[0605] The user harvests based on the suggestions and provides feedback on the results through the application, with data on the harvest results and the effectiveness of the suggestions being sent to the server.

[0606] The server retrains the generative AI model based on the collected feedback and emotion data to improve the accuracy of the next prediction.

[0607] Specific examples

[0608] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0609] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0610] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0611] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0612] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0613] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0614] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0615] The above is a specific embodiment for carrying out the present invention.

[0616] Example 2

[0617] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0618] In agriculture, determining the optimal harvest time and method is extremely important, but it is difficult to make accurate predictions due to the influence of numerous factors, such as weather conditions, soil conditions, and crop growth status. Furthermore, adjusting suggestions based on the farmer's emotions and stress levels is also essential to improving work efficiency. There is a need for a system that can comprehensively analyze these complex factors and make optimal harvest suggestions for farmers.

[0619] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0620] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvest method using the trained generative AI model, means for collecting and analyzing user emotion data, means for notifying farmers of the optimal harvest time and harvest method based on the prediction and the user emotion data, and means for collecting feedback data and emotion data and retraining to improve the generative AI model. This enables accurate harvest predictions based on the collected data and individual responses according to farmers' emotions, thereby maximizing harvest volume and quality and improving work efficiency.

[0621] "Weather data" refers to various information about weather conditions, such as temperature, precipitation, humidity, wind speed, and wind direction.

[0622] "Soil data" refers to information that indicates the condition of the soil, such as soil humidity, pH, water content, and nutrient content.

[0623] "Crop growth status" refers to information about the physical growth of the crop, such as crop height, leaf color, leaf area, and fruit size.

[0624] "Preprocessing" refers to converting collected data into an analyzable form, standardizing data formats, filling in missing data, detecting and replacing outliers, and removing noise.

[0625] A "generative AI model" is a model trained using machine learning techniques and is an algorithm that predicts the optimal harvest time and method based on collected data.

[0626] "User emotion data" refers to data related to emotions obtained from the user's tone of voice and facial expressions determined using voice recognition and facial recognition technology.

[0627] "Feedback data" refers to data on the results of actual harvesting work and the effectiveness of suggestions.

[0628] This invention is a system for optimizing agricultural harvesting operations. The system combines a generative AI model and an emotion engine to analyze weather, soil, and crop growth data to suggest optimal harvest times and methods to farmers.

[0629] Data collection

[0630] The server obtains weather data daily through a weather API. This data includes temperature, precipitation, humidity, wind speed, and wind direction. For example, the temperature might be 20°C, the probability of precipitation 40%, and humidity 70%. The server also receives real-time soil data from soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. For example, the server verifies that the soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. The server also collects crop growth status data from drone photography. Specifically, this data includes image data such as crop height, leaf color, leaf area, and fruit size. For example, the server verifies that the crop height is 30 cm, leaf color is a healthy green, leaf area is 15 square centimeters, and fruit size is 5 cm.

[0631] Data Preprocessing

[0632] The server converts the collected raw data into an analyzable form. First, it standardizes the data format and fills in missing data with the average or median. For example, missing temperature data for one day is filled in with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, it fills in the 70% soil humidity data detected as an outlier with the average value (60%) of the other normal values. It also removes noise to create a preprocessed dataset.

[0633] Training generative AI models

[0634] The server splits the preprocessed dataset into training data and test data and trains a generative AI model. This model is used to predict the optimal harvest time and method based on historical weather, soil, and crop growth data. Machine learning techniques are used for training, and iterative training is performed to achieve high accuracy. For example, the model is trained using weather data, soil data, and harvest results from the past five years.

[0635] Predicting optimal harvest times and methods

[0636] The server inputs the latest weather, soil, and crop growth data into the generative AI model, which predicts the optimal harvest time and method. The generative AI model then provides specific predictions, such as "The optimal harvest time is next Tuesday at 8:00 AM."

[0637] Collecting and analyzing user emotion data

[0638] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine the user's emotion (satisfaction, dissatisfaction, stress, etc.) in response to the harvesting suggestion.

[0639] Proposal Notification

[0640] The server then sends farmers optimal harvesting suggestions based on the predictions of the optimal harvesting time and method, as well as the user's emotional data. The suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best possible harvest is expected." The tone and detailed information of the suggestions may also be adjusted based on the user's emotional data.

[0641] Gathering feedback and retraining

[0642] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and retrains the generative AI model to further improve prediction accuracy next time.

[0643] This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on user emotion data also increase the likelihood of suggestions being accepted.

[0644] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0645] Step 1:

[0646] The server collects weather data through the weather API at 5:00 AM every day. Specifically, it obtains data such as temperature, precipitation, humidity, wind speed, and wind direction. It receives this weather data as input and stores it in a database. For example, it obtains data such as a temperature of 20 degrees, a 40% chance of precipitation, and a 70% humidity. As output, the weather data is recorded in the database.

[0647] Step 2:

[0648] The server collects soil data in real time from soil sensors installed in farmland. Specifically, data such as soil humidity, pH, moisture content, and nutrient content are acquired. The server receives the soil data as input and stores it in a database. For example, data is collected showing that soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. As output, the soil data is recorded in the database.

[0649] Step 3:

[0650] The server uses drone image data to collect information on the growth status of crops. Specifically, it analyzes image data such as crop height, leaf color, leaf area, and fruit size. It receives this crop growth data as input and stores it in a database. For example, it obtains data such as a crop height of 30 cm, a healthy green leaf color, a leaf area of ​​15 square centimeters, and a fruit size of 5 cm. As output, the crop growth data is recorded in the database.

[0651] Step 4:

[0652] The server preprocesses the collected weather, soil, and crop growth data. First, it standardizes the data format and imputes missing data with the mean or median. For example, missing temperature data for one day is imputed with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, soil moisture data detected as an outlier at 70% is imputed with 60%, the average of the other normal values. This preprocessed data is received as input and noise is removed. A preprocessed dataset is created as output.

[0653] Step 5:

[0654] The server splits the preprocessed dataset into training data and test data and trains the generative AI model. For example, it uses weather data, soil data, and harvest results from the past five years. It receives the training data as input and trains the AI ​​model. It then evaluates the model's performance using the test data. The output is a highly accurate generative AI model.

[0655] Step 6:

[0656] The server inputs the latest weather, soil, and crop growth data into the AI ​​model to predict the optimal harvest time and method. For example, it generates a specific prediction result, such as "8:00 AM next Tuesday is the optimal harvest time." It receives this prediction result as input and generates information to notify farmers. The output is the predicted harvest time and method.

[0657] Step 7:

[0658] The server uses an emotion engine to collect and analyze the user's emotional data. Specifically, it uses voice recognition and face recognition technology to determine the user's emotions from their tone of voice and facial expressions. It receives the emotional data as input and analyzes the user's emotional state. The output is the user's emotional data.

[0659] Step 8:

[0660] The server notifies the farmer of optimal harvesting suggestions based on the prediction results of the optimal harvesting time and method and the user's emotional data. Specifically, the suggestion is sent via push notification or email to the farmer's device (smartphone, PC, etc.). For example, the suggestion might be, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The suggestion notification is sent to the farmer as output.

[0661] Step 9:

[0662] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The feedback data is received as input and used to improve the AI ​​model. The output is a retrained generative AI model.

[0663] The above is a specific example of how to implement the invention. Using this system, farmers can predict the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotional data also increase the likelihood of their suggestions being accepted.

[0664] (Application example 2)

[0665] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0666] In conventional agricultural harvesting work, determining the optimal harvesting time and method is difficult, and in many cases workers have had to rely on experience and intuition. As a result, harvest volume and quality are unstable, making it difficult for agricultural workers to harvest at the appropriate time. Furthermore, when harvesting suggestions are simply notified mechanically, they do not take into account the emotions and reactions of agricultural workers, making them less likely to accept the suggestions. This has led to a demand for a system that can provide optimal harvesting suggestions and reduce the burden on agricultural workers.

[0667] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth status, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining the generative AI model to improve it, means for recognizing the emotions of users who receive the prediction notification and collecting emotion data, means for adjusting the notification content based on the emotion data, and means for using the collected emotion data to improve the generative AI model. This makes it possible to simultaneously optimize harvesting operations and increase the likelihood that agricultural workers will accept suggestions based on their emotions.

[0668] "Weather data" refers to information related to weather, including data such as temperature, precipitation, humidity, wind speed, and wind direction.

[0669] "Soil data" refers to information about the condition of the soil, including data on soil humidity, pH, water content, nutrient content, etc.

[0670] "Crop growth status" is information that indicates how the crop is growing, and includes data such as the height of the crop, leaf color, leaf area, and fruit size.

[0671] A "generative AI model" is an artificial intelligence model that is trained on past data to predict specific outcomes.

[0672] "Emotional data" refers to information about a user's emotional state, including data derived from vocal tone and facial expressions.

[0673] The "means for adjusting the notification content" is a means for changing the notification content regarding the optimal harvesting time and harvesting method based on the user's emotional data.

[0674] "Feedback Data" refers to information regarding actual harvest results and the effectiveness of recommendations, including data collected from users.

[0675] The "training means" is the process of training a generative AI model using preprocessed data.

[0676] "Preprocessing" refers to the process of converting the raw data acquired into an analyzable form.

[0677] "Retraining" is a training process that uses collected feedback data to improve the accuracy of an existing generative AI model.

[0678] The present invention is a system for highly optimizing agricultural harvesting operations, and is realized by including the following means.

[0679] First, the server collects weather data (temperature, precipitation, humidity, wind speed, wind direction, etc.) using a weather API at a specific time each day. Next, soil data is collected in real time using soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. Additionally, cameras and drones installed in the farmland are used to collect information on the growth status of the crops (plant height, leaf color, leaf area, fruit size, etc.).

[0680] The server preprocesses the collected data. Specifically, it uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, unify multiple data formats, and remove noise.

[0681] The preprocessed data is then used to train a generative AI model using machine learning frameworks such as TensorFlow and PyTorch, based on historical weather, soil, and crop growth data, which is then used to predict the optimal harvest time and method.

[0682] Using the trained generative AI model, the server inputs the latest data and predicts the best time and method for harvesting, generating a specific prediction result, for example, "The best time to harvest is next Tuesday at 8:00 AM."

[0683] In addition, the server uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to collect the user's emotional data. The server analyzes the user's tone of voice and facial expressions when receiving the notification and collects emotional data. Based on this emotional data, the server determines the user's emotional state, such as "comfortable," "anxious," or "confused," and adjusts the content of the notification accordingly.

[0684] Finally, the server notifies the farmer of the best time and method to harvest. The notification is sent via push notification or email. The content of the notification is adjusted based on the sentiment data, so it is presented in a way that is easy for the farmer to accept and understand.

[0685] Users harvest based on the suggestions and provide feedback on the results via their smartphone or computer. This feedback data is sent to the server and used to retrain the existing generative AI model, allowing the system to further improve its prediction accuracy in the future.

[0686] For example, a weather API retrieves the weather forecast for the coming week at 8:00 a.m. every morning, and a soil sensor collects soil moisture data every hour. The user's reaction to the harvest suggestion can be collected and the next notification can be tailored based on that reaction. A prompt such as "Please retrieve the weather forecast for next week and collect data to predict the optimal harvest time" can be used to create input data for a generative AI model.

[0687] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0688] Step 1:

[0689] The server collects weather data. Specifically, it uses a weather API to obtain data such as temperature, precipitation, humidity, wind speed, and wind direction at a specific time each day. This data is stored on the server in JSON format. The input is raw data obtained from the weather API, and the output is weather data in a unified JSON format.

[0690] Step 2:

[0691] The server collects soil data. Soil sensors acquire data such as soil humidity, pH, moisture content, and nutrient content in real time and send it to the server. This data is stored in the server in RAW format. The input is raw data from the soil sensors, and the output is formatted soil data.

[0692] Step 3:

[0693] The server collects information on the growth status of crops. It analyzes image data taken using cameras and drones installed on farmland to extract information such as crop height, leaf color, leaf area, and fruit size. It preprocesses the acquired image data to generate the necessary growth data. The input is image data from the cameras and drones, and the output is analyzed growth data.

[0694] Step 4:

[0695] The server integrates preprocessed weather data, soil data, and crop growth data. It uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, and unify multiple data formats. The input is individual data before preprocessing, and the output is the integrated, preprocessed data.

[0696] Step 5:

[0697] The server trains a generative AI model based on the preprocessed data. Using TensorFlow and PyTorch, the model is trained based on historical weather, soil, and crop growth data to generate a model that can predict the optimal harvest time and method. The input is the preprocessed data, and the output is the trained generative AI model.

[0698] Step 6:

[0699] The server uses a generative AI model to predict the optimal harvest time and method. The latest data is input into the model to generate a specific prediction. For example, the output may be "The optimal harvest time is 8:00 a.m. next Tuesday." The input is the latest data and a trained generative AI model, and the output is a prediction of the harvest time and method.

[0700] Step 7:

[0701] The server collects the user's emotional data. It uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to analyze the user's tone of voice and facial expression when they receive the suggestion notification. This emotional data is stored on the server. The input is the user's voice and facial image data, and the output is the analyzed emotional data.

[0702] Step 8:

[0703] The server adjusts the notification content based on the emotional data. Based on the collected emotional data, it modifies the suggestions to make them more acceptable to the user, providing the optimal tone and detailed information. The input is the emotional data and prediction results, and the output is the adjusted notification content.

[0704] Step 9:

[0705] The server notifies the farmer of the best time and method to harvest, sending specific suggestions via push notification or email. For example, it might say, "Start harvesting next Tuesday morning. Current soil moisture and forecasted weather suggest the best possible harvest." The input is the tailored notification content, and the output is the notification message.

[0706] Step 10:

[0707] The user harvests based on the suggestions and provides feedback on the results. Feedback data is sent to the server via smartphone or PC. The input is the user's harvest results and information on the validity of the suggestions, and the output is new feedback data.

[0708] Step 11:

[0709] The server uses the collected feedback data to retrain the generative AI model. By adding new data, the accuracy of the model improves. The input is the feedback data, and the output is an improved generative AI model.

[0710] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0711] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0712] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0713] [Third embodiment]

[0714] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0715] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0716] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0717] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0718] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0719] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0720] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0721] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0722] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0723] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0724] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0725] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0726] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system can be implemented as follows.

[0727] Data collection

[0728] The server collects weather, soil, and crop growth data via the internet and sensors. Weather data is automatically obtained daily through a weather API, and soil data is received in real time from soil sensors installed in the farmland. Crop growth status is collected from drone photography data and periodically captured image data.

[0729] Data Preprocessing

[0730] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces a single outlier, it is imputed with the average of the other normal values.

[0731] Training generative AI models

[0732] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0733] Predicting optimal harvest times and methods

[0734] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0735] Proposal Notification

[0736] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0737] Gathering feedback and retraining

[0738] The user harvests based on the suggestions and provides feedback on the results. This sends data about the actual results of the harvest and the effectiveness of the suggestions to the server. The server then uses this feedback data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0739] Specific examples

[0740] 1. Data collection example: The server retrieves one-week weather forecast data for Tokyo from the weather API at 5:00 AM every day. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from sensors installed in farmland.

[0741] 2. Preprocessing: The server interpolates the 70% of humidity data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0742] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0743] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0744] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0745] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0746] The above is a specific embodiment for carrying out the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing the yield and quality of their crops.

[0747] The processing flow will be explained below.

[0748] Step 1: Data collection

[0749] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0750] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0751] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0752] Step 2: Data Preprocessing

[0753] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format.

[0754] The server imputes missing data with the mean or median, detects outliers and replaces them with reasonable values.

[0755] The server performs noise removal and creates a preprocessed dataset, for example, by interpolating outliers from a sensor with the average of normal values ​​from other sensors.

[0756] Step 3: Model generation

[0757] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0758] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0759] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0760] Step 4: Optimization

[0761] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0762] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0763] Step 5: Notification

[0764] The server then notifies the farmer of the results of the predictions via push notification, email, or an application.

[0765] The device will then display the received notification to the user, for example, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected."

[0766] Step 6: Feedback

[0767] The user carries out harvesting work based on the suggestions and provides feedback on the results through the application.

[0768] The server collects feedback data on harvest results and the effectiveness of the proposals, specifically on harvest volume, quality, and work efficiency.

[0769] The server retrains the generative AI model based on the collected feedback data to improve the accuracy of the next prediction, including harvest result feedback data.

[0770] The above is the processing procedure of the program of the present invention.

[0771] Example 1

[0772] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0773] Agricultural harvesting depends on complex factors such as weather conditions, soil conditions, and crop growth, making it difficult to properly evaluate these and determine the optimal harvest time and method. Furthermore, methods that rely on experience and intuition are prone to variations in harvest volume and quality, making efficient agricultural management difficult. Furthermore, conventional systems lack the ability to utilize feedback for improvements, making it difficult to improve prediction accuracy.

[0774] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0775] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying farmers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data from farmers and retraining the generative AI model to improve it, means for detecting and imputing outliers and imputing missing data with mean or median values, and means for iteratively training based on past data to evaluate and improve model performance. This makes it possible to improve the accuracy of harvest time and method predictions, maximize harvest volume and quality, and streamline agricultural management.

[0776] "Weather data" refers to data collected based on meteorological observations that indicates atmospheric conditions such as temperature, humidity, precipitation, wind speed, and sunshine hours.

[0777] "Soil data" refers to data measuring the condition of soil in agricultural or cultivated areas, such as humidity, pH, temperature, and nutrient content.

[0778] "Crop growth status" refers to data indicating the growth and health status of the crop under cultivation, such as its height, leaf area, color, and the presence or absence of pests and diseases.

[0779] "Preprocessing" refers to a series of processes that supplement missing values ​​and correct outliers in collected data to prepare it in a format suitable for analysis and model training.

[0780] A "generative AI model" is an artificial intelligence model trained from pre-processed data using machine learning algorithms to automatically perform specific tasks or predictions.

[0781] The "optimal harvest time and harvest method" refers to the most suitable time and method for harvesting a crop, determined by a comprehensive assessment of weather conditions, soil conditions, and the growth state of the crop.

[0782] "Notification" refers to the act or means of transmitting information to inform users of predictions or suggestions.

[0783] "Feedback data" refers to data on the results and evaluation of harvesting work carried out by farmers based on the suggestions.

[0784] "Retraining" is the training process of updating and improving existing generative AI models using new collected data.

[0785] "Outlier detection and imputation" is the process of detecting values ​​in a dataset that fall outside the normal range and, if necessary, replacing them with other reasonable values.

[0786] "Iterative training" is a technique in which a generative AI model is trained multiple times using the same or extended dataset to improve its performance.

[0787] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system is implemented using the following hardware and software.

[0788] Hardware and software used

[0789] Weather data collection method: Daily weather data is collected using a weather API (e.g., OpenWeatherMap API).

[0790] Soil data collection methods: Soil sensors (e.g., soil moisture sensors, pH sensors) are used to collect real-time soil data.

[0791] Methods for collecting crop growth status: Drones and ground cameras are used to photograph crop growth status and collect image data.

[0792] Generative AI models: Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models to predict the optimal time and method of harvesting.

[0793] Data preprocessing: Programs for imputing missing data and detecting and correcting outliers (e.g., Python's Pandas library).

[0794] Explanation of the program processing flow

[0795] Data collection

[0796] The server collects data in the following ways:

[0797] Weather data is automatically collected via a weather API at 5:00 AM every day.

[0798] Soil data is received in real time from soil sensors installed on farmland.

[0799] The growth status of the crops is collected from image data taken by drones on a regular basis and from image data taken by ground cameras.

[0800] Data Preprocessing

[0801] The server pre-processes the collected raw data:

[0802] If there are missing data, they are imputed with the mean or median.

[0803] Anomalies are detected and interpolated with the average of other normal values. For example, if 70% of the data obtained from a soil moisture sensor is abnormal, it is interpolated with the average of other normal values.

[0804] Training generative AI models

[0805] The server trains a generative AI model using the preprocessed dataset:

[0806] Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models based on historical weather, soil, and crop growth data.

[0807] Cross-validation is performed to evaluate the model's performance and hyperparameters are adjusted as needed.

[0808] Predicting optimal harvest times and methods

[0809] The server inputs the latest data into the generative AI model and makes a prediction:

[0810] For example, new weather forecast data and soil moisture data are input to generate a prediction that "the best time to harvest is next Tuesday at 8:00 a.m."

[0811] Proposal Notification

[0812] The server then uses the prediction results to provide farmers with optimal harvesting suggestions:

[0813] Notifications are sent to farmers' devices (smartphones or computers) via push notifications or email.

[0814] For example, make a specific suggestion like, "Please start harvesting next Tuesday at 8:00 a.m. The current soil moisture and weather forecast will result in the best possible harvest."

[0815] Gathering feedback and retraining

[0816] The user harvests based on the suggestions and provides feedback on the results:

[0817] Users input data on harvest yield and quality through the terminal.

[0818] The server uses the collected feedback data to retrain the generative AI model and improve the accuracy of its predictions.

[0819] Specific examples

[0820] Data collection example: Every day at 5:00 AM, the server retrieves one-week weather forecast data from the weather API. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from the soil sensor.

[0821] Preprocessing: The 70% of humidity data detected as abnormal values ​​are interpolated with the average normal value (60%).

[0822] Model generation: Train a generative AI model using weather data, soil data, and crop results from the past five years.

[0823] Optimization: Input the latest data and predict that "8:00 a.m. next Tuesday is the best time to harvest."

[0824] Notification: The prediction results will be sent to the user's smartphone via push notification.

[0825] Feedback: Users harvest and input their results into the app, and the server uses that data to retrain the generative AI model.

[0826] As described above, farmers can use the system to know the optimal harvesting time and method in advance, maximizing harvest volume and quality.

[0827] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0828] Step 1: Data collection

[0829] The server collects weather data, soil data, and crop growth status data. Inputs are data from weather APIs, soil sensors, drones, and ground cameras. Weather data is automatically acquired at 5:00 AM every day, and soil data is received in real time. Crop growth status data is captured periodically. The output of the data collection step is a raw dataset of weather data, soil data, and crop growth status data.

[0830] Step 2: Data Preprocessing

[0831] The server preprocesses the collected raw data. The input is the raw data collected in step 1. First, it imputes missing data with the mean or median, and detects and corrects outliers. For example, if the data from a soil moisture sensor contains an outlier of 70%, it imputes it with the mean of the other normal values. The output is a preprocessed dataset.

[0832] Step 3: Training the generative AI model

[0833] The server trains a generative AI model using the preprocessed dataset. The input is the dataset preprocessed in step 2. Using a machine learning framework (e.g., TensorFlow, PyTorch), the model is trained based on historical weather data, soil data, and crop growth data. Cross-validation is performed during the training process to evaluate and improve the model's performance. The output is a trained generative AI model.

[0834] Step 4: Predict the optimal harvest time and method

[0835] The server inputs the latest data into the generative AI model to predict the optimal harvest time and method. The inputs are the latest weather data, soil data, and growth data. For example, by combining weather data and soil moisture data, a prediction is generated that "the optimal harvest time is 8:00 a.m. next Tuesday." The output is a prediction of the specific harvest time and method.

[0836] Step 5: Proposal Notification

[0837] The server notifies the farmer of the optimal harvesting suggestion based on the prediction results. The input is the prediction result from step 4. The notification is sent to the farmer's device (smartphone or PC) via push notification or email. For example, a specific suggestion may be made such as, "Please start harvesting at 8:00 a.m. next Tuesday. Based on the current soil moisture and weather forecast, the best harvest is expected." The output is the notified suggestion.

[0838] Step 6: Gather feedback and retrain

[0839] The user harvests based on the suggestions and provides feedback on the results. The input is the harvest result data. The user inputs data on the harvest quantity and quality through their terminal, and the server uses the feedback data to retrain the generative AI model. The output is an improved generative AI model.

[0840] (Application example 1)

[0841] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0842] Optimizing harvest times in agriculture is a crucial issue for improving yield and quality. Traditionally, farmers decide when to harvest based on their own experience and intuition, which can lead to errors in judgment and oversights. Furthermore, there are limited concrete methods for improving the efficiency and automating harvesting work. Therefore, there is a need for a system that automatically optimizes harvest times and methods and enables autonomous vehicles to carry out harvesting work based on the results.

[0843] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0844] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining to improve the generative AI model, means for generating a driving plan for an autonomous vehicle based on the prediction and for the autonomous vehicle to automatically start harvesting work based on the driving plan, and means for collecting driving results of the autonomous vehicle as feedback and retraining to improve the generative AI model. This automatically optimizes the harvest time and harvesting method, realizes automation and efficiency of harvesting work, and enables improved harvest volume and quality.

[0845] "Weather data" is information indicating weather conditions such as weather, temperature, precipitation, humidity, and wind speed.

[0846] "Soil data" refers to information about agricultural soil, including soil moisture, pH value, nutrient content, etc.

[0847] "Crop growth status" is information indicating the growth stage and health status of the crop, and includes, for example, germination, growth, flowering, and harvestable status.

[0848] "Preprocessing" refers to the process of preparing raw data in an analyzable format, and includes filling in missing data and correcting outliers.

[0849] A "generative AI model" is an artificial intelligence model trained to use collected data to predict optimal harvest times and methods.

[0850] "Training" refers to the process by which an AI model learns from collected data so that it can make accurate predictions.

[0851] "Prediction" means that a generative AI model infers future situations based on input data.

[0852] "Notification" refers to the method of informing farmers of the prediction results, and includes push notifications and emails.

[0853] An "autonomous vehicle" is a vehicle that operates automatically based on programmed instructions to perform agricultural tasks.

[0854] An "operation plan" is a plan that shows the route and work procedures required for an autonomous vehicle to carry out harvesting work.

[0855] "Feedback" refers to collecting information based on the results of harvesting operations and actual conditions, and using it to improve the system and retrain the AI ​​model.

[0856] "Retraining" is the process of re-learning a generative AI model to improve its performance using newly collected data.

[0857] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. It also has the function of generating operation plans for autonomous vehicles based on the prediction results and automatically carrying out harvesting operations.

[0858] Data collection

[0859] The server uses a weather API to collect weather data. This automatically obtains the latest weather, temperature, precipitation, humidity, wind speed, and other weather conditions every day. It also collects soil data in real time using soil sensors installed in the farmland. This soil data includes soil moisture, pH, and nutrient content. Furthermore, drones are used to capture images of the growing conditions of the crops and collect the data. This allows the growth stage and health of the crops to be understood.

[0860] Data Preprocessing

[0861] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces an abnormal value, it is imputed with the average of the other normal values.

[0862] Training generative AI models

[0863] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[0864] Predicting optimal harvest times and methods

[0865] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[0866] Proposal Notification

[0867] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[0868] Autonomous vehicles perform harvesting operations

[0869] The server generates a driving plan for the autonomous vehicle based on the predictions. Based on this driving plan, the autonomous vehicle automatically starts harvesting work. For example, instructions such as "Harvesting work will automatically start at 8:00 a.m. next Tuesday" are sent to the autonomous vehicle.

[0870] Gathering feedback and retraining

[0871] Users harvest based on the suggestions and provide feedback on the results. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and uses this data to retrain the generative AI model, further improving the accuracy of the next prediction.

[0872] Specific examples

[0873] 1. Data Collection Example: The server retrieves a week's worth of weather data for a city from a weather API at 5:00 AM every day. At the same time, it receives soil moisture data from sensors installed in farmland.

[0874] 2. Data preprocessing: The server interpolates the humidity data detected as abnormal values ​​with the average value of other normal values.

[0875] 3. Model generation: The server trains a generative AI model that uses historical weather data, soil data, and harvest results to predict optimal harvest times.

[0876] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0877] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[0878] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[0879] Prompt Sentence Examples

[0880] Enter the weather, soil and crop growth data as follows:

[0881] Weather data: Temperature=30, Humidity=70, Precipitation=0

[0882] Soil data: soil moisture=60, pH=6.5

[0883] Crop growth data: Growth stage=vegetative, Image data=base64_encoded_image_string

[0884] By inputting this data into a generative AI model, the optimal harvest time and method can be predicted.

[0885] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0886] Step 1: Data collection

[0887] The server obtains weather data from the weather API and collects soil and crop growth data from soil sensors and drones installed in the farmland. The server sends requests to the weather API to obtain information such as temperature, humidity, and precipitation, and obtains soil humidity and pH data from the soil sensors. It also operates a drone to take images of the crops and collect data on their growth status. The input data consists of weather data, soil sensor data, and crop image data, and these are used to proceed to the next step.

[0888] Step 2: Data Preprocessing

[0889] The server preprocesses the collected weather data, soil data, and crop growth data. If there are missing data or outliers, it fills them in with the mean or median and corrects the outliers. For example, if data from a soil moisture sensor contains an outlier, the server fills it in using the mean value of other normal data. The input data is the collected raw data, which is then processed to output preprocessed data.

[0890] Step 3: Training the generative AI model

[0891] The server uses the preprocessed dataset to train a generative AI model. It uses historical weather, soil, and crop growth data to build a multi-layer neural network and train a model to predict optimal harvest times and methods. The server uses the preprocessed dataset as input data and outputs a trained generative AI model.

[0892] Step 4: Predict harvest time and method

[0893] The server inputs the latest data into a trained generative AI model to predict the optimal harvest time and method. For example, it uses next week's weather data and current soil moisture data to obtain a specific prediction result, such as "The optimal time to harvest is next Tuesday morning." The input data is the latest weather data, soil data, and crop growth data, and the output data is the prediction result of the optimal harvest time and method.

[0894] Step 5: Proposal Notification

[0895] Based on the prediction results, the server notifies the farmer of the optimal harvest time and harvesting method. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, a message could be sent saying, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The input data is the prediction result, and the output data is the notification message.

[0896] Step 6: Generate a driving plan for the autonomous vehicle

[0897] The server generates a driving plan for the autonomous vehicle based on the predictions. The autonomous vehicle then automatically starts harvesting according to the generated driving plan. For example, an instruction such as "Harvesting will begin automatically at 8:00 AM next Tuesday" is sent to the autonomous vehicle. The input data is the predicted results of the optimal harvesting time and method, and the output data is the driving plan and specific instructions for the autonomous vehicle.

[0898] Step 7: Gather feedback

[0899] The user harvests based on the suggestions and sends the results as feedback to the server. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and retrains the generative AI model to improve future prediction accuracy. The input data are the harvest results and feedback data, and the output data is an improved generative AI model.

[0900] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0901] This invention is a system for optimizing agricultural harvesting operations, combining a generative AI model and an emotion engine to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. This system recognizes the user's emotions and uses them as feedback, enabling even more accurate suggestions.

[0902] Data collection

[0903] The server obtains weather data daily through a weather API. The obtained data includes temperature, precipitation, humidity, wind speed, and wind direction. The server also receives soil data in real time from soil sensors. The soil data includes soil humidity, pH, moisture content, and nutrient content. The server also collects information on the growth status of crops from drone photography data and periodically captured image data. Specifically, the image data includes crop height, leaf color, leaf area, and fruit size.

[0904] Data Preprocessing

[0905] The server converts the collected raw data into an analyzable form. For example, it standardizes the acquired data format and fills in missing data with the mean or median. Outliers are detected and replaced with reasonable values. It also performs noise removal and creates a preprocessed data set. For example, it fills in outliers obtained from a soil moisture sensor with the mean of other normal values.

[0906] Training generative AI models

[0907] The server splits the preprocessed dataset into a training dataset and a test dataset and trains a generative AI model that uses historical weather, soil, and crop growth data to predict the optimal time and method of harvesting. Machine learning techniques are used for training, and iterative training is performed to evaluate and improve the model's performance.

[0908] Predicting optimal harvest times and methods

[0909] The server inputs the latest weather, soil, and crop growth data into the generative AI model to predict the optimal harvest time and method, generating specific predictions such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0910] Collecting and analyzing user emotion data

[0911] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine how the user feels about the harvesting suggestions (satisfaction, dissatisfaction, stress, etc.).

[0912] Proposal Notification

[0913] The server then uses the prediction results and emotion data to provide farmers with optimal harvest suggestions. These suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected." The tone and details of the suggestion may also be adjusted based on the user's emotion data.

[0914] Gathering feedback and retraining

[0915] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and uses it to retrain the generative AI model, further improving the accuracy of the next prediction.

[0916] Specific examples

[0917] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0918] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0919] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0920] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0921] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0922] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0923] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0924] The above is a specific example of how to implement the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotions also increase the likelihood of suggestions being accepted.

[0925] The processing flow will be explained below.

[0926] Program processing

[0927] Step 1: Data collection

[0928] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[0929] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[0930] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[0931] Step 2: Data Preprocessing

[0932] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format and filling in missing data with the mean or median.

[0933] The server detects outliers and replaces them with reasonable values, for example, interpolating outliers from a soil moisture sensor with the average of other normal values.

[0934] The server performs noise removal and creates a preprocessed dataset.

[0935] Step 3: Model generation

[0936] The server splits the preprocessed dataset into a training dataset and a test dataset.

[0937] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[0938] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[0939] Step 4: Optimization

[0940] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[0941] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[0942] Step 5: Collecting sentiment data

[0943] The server collects user emotion data using an emotion engine, which uses voice and face recognition technology to determine emotions from the user's tone of voice and facial expressions.

[0944] The device analyzes the acquired emotional data and evaluates the user's stress level and satisfaction.

[0945] Step 6: Generate and send notifications

[0946] The server then generates optimal harvesting suggestions for the user based on the prediction results and emotion data, such as "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best harvest is expected."

[0947] The server sends the generated suggestions to the user's device via push notification, email, or application.

[0948] Step 7: Gather feedback and retrain

[0949] The user harvests based on the suggestions and provides feedback on the results through the application, with data on the harvest results and the effectiveness of the suggestions being sent to the server.

[0950] The server retrains the generative AI model based on the collected feedback and emotion data to improve the accuracy of the next prediction.

[0951] Specific examples

[0952] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[0953] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[0954] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[0955] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[0956] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[0957] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[0958] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[0959] The above is a specific embodiment for carrying out the present invention.

[0960] Example 2

[0961] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0962] In agriculture, determining the optimal harvest time and method is extremely important, but it is difficult to make accurate predictions due to the influence of numerous factors, such as weather conditions, soil conditions, and crop growth status. Furthermore, adjusting suggestions based on the farmer's emotions and stress levels is also essential to improving work efficiency. There is a need for a system that can comprehensively analyze these complex factors and make optimal harvest suggestions for farmers.

[0963] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0964] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvest method using the trained generative AI model, means for collecting and analyzing user emotion data, means for notifying farmers of the optimal harvest time and harvest method based on the prediction and the user emotion data, and means for collecting feedback data and emotion data and retraining to improve the generative AI model. This enables accurate harvest predictions based on the collected data and individual responses according to farmers' emotions, thereby maximizing harvest volume and quality and improving work efficiency.

[0965] "Weather data" refers to various information about weather conditions, such as temperature, precipitation, humidity, wind speed, and wind direction.

[0966] "Soil data" refers to information that indicates the condition of the soil, such as soil humidity, pH, water content, and nutrient content.

[0967] "Crop growth status" refers to information about the physical growth of the crop, such as crop height, leaf color, leaf area, and fruit size.

[0968] "Preprocessing" refers to converting collected data into an analyzable form, standardizing data formats, filling in missing data, detecting and replacing outliers, and removing noise.

[0969] A "generative AI model" is a model trained using machine learning techniques and is an algorithm that predicts the optimal harvest time and method based on collected data.

[0970] "User emotion data" refers to data related to emotions obtained from the user's tone of voice and facial expressions determined using voice recognition and facial recognition technology.

[0971] "Feedback data" refers to data on the results of actual harvesting work and the effectiveness of suggestions.

[0972] This invention is a system for optimizing agricultural harvesting operations. The system combines a generative AI model and an emotion engine to analyze weather, soil, and crop growth data to suggest optimal harvest times and methods to farmers.

[0973] Data collection

[0974] The server obtains weather data daily through a weather API. This data includes temperature, precipitation, humidity, wind speed, and wind direction. For example, the temperature might be 20°C, the probability of precipitation 40%, and humidity 70%. The server also receives real-time soil data from soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. For example, the server verifies that the soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. The server also collects crop growth status data from drone photography. Specifically, this data includes image data such as crop height, leaf color, leaf area, and fruit size. For example, the server verifies that the crop height is 30 cm, leaf color is a healthy green, leaf area is 15 square centimeters, and fruit size is 5 cm.

[0975] Data Preprocessing

[0976] The server converts the collected raw data into an analyzable form. First, it standardizes the data format and fills in missing data with the average or median. For example, missing temperature data for one day is filled in with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, it fills in the 70% soil humidity data detected as an outlier with the average value (60%) of the other normal values. It also removes noise to create a preprocessed dataset.

[0977] Training generative AI models

[0978] The server splits the preprocessed dataset into training data and test data and trains a generative AI model. This model is used to predict the optimal harvest time and method based on historical weather, soil, and crop growth data. Machine learning techniques are used for training, and iterative training is performed to achieve high accuracy. For example, the model is trained using weather data, soil data, and harvest results from the past five years.

[0979] Predicting optimal harvest times and methods

[0980] The server inputs the latest weather, soil, and crop growth data into the generative AI model, which predicts the optimal harvest time and method. The generative AI model then provides specific predictions, such as "The optimal harvest time is next Tuesday at 8:00 AM."

[0981] Collecting and analyzing user emotion data

[0982] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine the user's emotion (satisfaction, dissatisfaction, stress, etc.) in response to the harvesting suggestion.

[0983] Proposal Notification

[0984] The server then sends farmers optimal harvesting suggestions based on the predictions of the optimal harvesting time and method, as well as the user's emotional data. The suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best possible harvest is expected." The tone and detailed information of the suggestions may also be adjusted based on the user's emotional data.

[0985] Gathering feedback and retraining

[0986] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and retrains the generative AI model to further improve prediction accuracy next time.

[0987] This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on user emotion data also increase the likelihood of suggestions being accepted.

[0988] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0989] Step 1:

[0990] The server collects weather data through the weather API at 5:00 AM every day. Specifically, it obtains data such as temperature, precipitation, humidity, wind speed, and wind direction. It receives this weather data as input and stores it in a database. For example, it obtains data such as a temperature of 20 degrees, a 40% chance of precipitation, and a 70% humidity. As output, the weather data is recorded in the database.

[0991] Step 2:

[0992] The server collects soil data in real time from soil sensors installed in farmland. Specifically, data such as soil humidity, pH, moisture content, and nutrient content are acquired. The server receives the soil data as input and stores it in a database. For example, data is collected showing that soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. As output, the soil data is recorded in the database.

[0993] Step 3:

[0994] The server uses drone image data to collect information on the growth status of crops. Specifically, it analyzes image data such as crop height, leaf color, leaf area, and fruit size. It receives this crop growth data as input and stores it in a database. For example, it obtains data such as a crop height of 30 cm, a healthy green leaf color, a leaf area of ​​15 square centimeters, and a fruit size of 5 cm. As output, the crop growth data is recorded in the database.

[0995] Step 4:

[0996] The server preprocesses the collected weather, soil, and crop growth data. First, it standardizes the data format and imputes missing data with the mean or median. For example, missing temperature data for one day is imputed with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, soil moisture data detected as an outlier at 70% is imputed with 60%, the average of the other normal values. This preprocessed data is received as input and noise is removed. A preprocessed dataset is created as output.

[0997] Step 5:

[0998] The server splits the preprocessed dataset into training data and test data and trains the generative AI model. For example, it uses weather data, soil data, and harvest results from the past five years. It receives the training data as input and trains the AI ​​model. It then evaluates the model's performance using the test data. The output is a highly accurate generative AI model.

[0999] Step 6:

[1000] The server inputs the latest weather, soil, and crop growth data into the AI ​​model to predict the optimal harvest time and method. For example, it generates a specific prediction result, such as "8:00 AM next Tuesday is the optimal harvest time." It receives this prediction result as input and generates information to notify farmers. The output is the predicted harvest time and method.

[1001] Step 7:

[1002] The server uses an emotion engine to collect and analyze the user's emotional data. Specifically, it uses voice recognition and face recognition technology to determine the user's emotions from their tone of voice and facial expressions. It receives the emotional data as input and analyzes the user's emotional state. The output is the user's emotional data.

[1003] Step 8:

[1004] The server notifies the farmer of optimal harvesting suggestions based on the prediction results of the optimal harvesting time and method and the user's emotional data. Specifically, the suggestion is sent via push notification or email to the farmer's device (smartphone, PC, etc.). For example, the suggestion might be, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The suggestion notification is sent to the farmer as output.

[1005] Step 9:

[1006] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The feedback data is received as input and used to improve the AI ​​model. The output is a retrained generative AI model.

[1007] The above is a specific example of how to implement the invention. Using this system, farmers can predict the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotional data also increase the likelihood of their suggestions being accepted.

[1008] (Application example 2)

[1009] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1010] In conventional agricultural harvesting work, determining the optimal harvesting time and method is difficult, and in many cases workers have had to rely on experience and intuition. As a result, harvest volume and quality are unstable, making it difficult for agricultural workers to harvest at the appropriate time. Furthermore, when harvesting suggestions are simply notified mechanically, they do not take into account the emotions and reactions of agricultural workers, making them less likely to accept the suggestions. This has led to a demand for a system that can provide optimal harvesting suggestions and reduce the burden on agricultural workers.

[1011] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth status, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining the generative AI model to improve it, means for recognizing the emotions of users who receive the prediction notification and collecting emotion data, means for adjusting the notification content based on the emotion data, and means for using the collected emotion data to improve the generative AI model. This makes it possible to simultaneously optimize harvesting operations and increase the likelihood that agricultural workers will accept suggestions based on their emotions.

[1012] "Weather data" refers to information related to weather, including data such as temperature, precipitation, humidity, wind speed, and wind direction.

[1013] "Soil data" refers to information about the condition of the soil, including data on soil humidity, pH, water content, nutrient content, etc.

[1014] "Crop growth status" is information that indicates how the crop is growing, and includes data such as the height of the crop, leaf color, leaf area, and fruit size.

[1015] A "generative AI model" is an artificial intelligence model that is trained on past data to predict specific outcomes.

[1016] "Emotional data" refers to information about a user's emotional state, including data derived from vocal tone and facial expressions.

[1017] The "means for adjusting the notification content" is a means for changing the notification content regarding the optimal harvesting time and harvesting method based on the user's emotional data.

[1018] "Feedback Data" refers to information regarding actual harvest results and the effectiveness of recommendations, including data collected from users.

[1019] The "training means" is the process of training a generative AI model using preprocessed data.

[1020] "Preprocessing" refers to the process of converting the raw data acquired into an analyzable form.

[1021] "Retraining" is a training process that uses collected feedback data to improve the accuracy of an existing generative AI model.

[1022] The present invention is a system for highly optimizing agricultural harvesting operations, and is realized by including the following means.

[1023] First, the server collects weather data (temperature, precipitation, humidity, wind speed, wind direction, etc.) using a weather API at a specific time each day. Next, soil data is collected in real time using soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. Additionally, cameras and drones installed in the farmland are used to collect information on the growth status of the crops (plant height, leaf color, leaf area, fruit size, etc.).

[1024] The server preprocesses the collected data. Specifically, it uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, unify multiple data formats, and remove noise.

[1025] The preprocessed data is then used to train a generative AI model using machine learning frameworks such as TensorFlow and PyTorch, based on historical weather, soil, and crop growth data, which is then used to predict the optimal harvest time and method.

[1026] Using the trained generative AI model, the server inputs the latest data and predicts the best time and method for harvesting, generating a specific prediction result, for example, "The best time to harvest is next Tuesday at 8:00 AM."

[1027] In addition, the server uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to collect the user's emotional data. The server analyzes the user's tone of voice and facial expressions when receiving the notification and collects emotional data. Based on this emotional data, the server determines the user's emotional state, such as "comfortable," "anxious," or "confused," and adjusts the content of the notification accordingly.

[1028] Finally, the server notifies the farmer of the best time and method to harvest. The notification is sent via push notification or email. The content of the notification is adjusted based on the sentiment data, so it is presented in a way that is easy for the farmer to accept and understand.

[1029] Users harvest based on the suggestions and provide feedback on the results via their smartphone or computer. This feedback data is sent to the server and used to retrain the existing generative AI model, allowing the system to further improve its prediction accuracy in the future.

[1030] For example, a weather API retrieves the weather forecast for the coming week at 8:00 a.m. every morning, and a soil sensor collects soil moisture data every hour. The user's reaction to the harvest suggestion can be collected and the next notification can be tailored based on that reaction. A prompt such as "Please retrieve the weather forecast for next week and collect data to predict the optimal harvest time" can be used to create input data for a generative AI model.

[1031] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1032] Step 1:

[1033] The server collects weather data. Specifically, it uses a weather API to obtain data such as temperature, precipitation, humidity, wind speed, and wind direction at a specific time each day. This data is stored on the server in JSON format. The input is raw data obtained from the weather API, and the output is weather data in a unified JSON format.

[1034] Step 2:

[1035] The server collects soil data. Soil sensors acquire data such as soil humidity, pH, moisture content, and nutrient content in real time and send it to the server. This data is stored in the server in RAW format. The input is raw data from the soil sensors, and the output is formatted soil data.

[1036] Step 3:

[1037] The server collects information on the growth status of crops. It analyzes image data taken using cameras and drones installed on farmland to extract information such as crop height, leaf color, leaf area, and fruit size. It preprocesses the acquired image data to generate the necessary growth data. The input is image data from the cameras and drones, and the output is analyzed growth data.

[1038] Step 4:

[1039] The server integrates preprocessed weather data, soil data, and crop growth data. It uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, and unify multiple data formats. The input is individual data before preprocessing, and the output is the integrated, preprocessed data.

[1040] Step 5:

[1041] The server trains a generative AI model based on the preprocessed data. Using TensorFlow and PyTorch, the model is trained based on historical weather, soil, and crop growth data to generate a model that can predict the optimal harvest time and method. The input is the preprocessed data, and the output is the trained generative AI model.

[1042] Step 6:

[1043] The server uses a generative AI model to predict the optimal harvest time and method. The latest data is input into the model to generate a specific prediction. For example, the output may be "The optimal harvest time is 8:00 a.m. next Tuesday." The input is the latest data and a trained generative AI model, and the output is a prediction of the harvest time and method.

[1044] Step 7:

[1045] The server collects the user's emotional data. It uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to analyze the user's tone of voice and facial expression when they receive the suggestion notification. This emotional data is stored on the server. The input is the user's voice and facial image data, and the output is the analyzed emotional data.

[1046] Step 8:

[1047] The server adjusts the notification content based on the emotional data. Based on the collected emotional data, it modifies the suggestions to make them more acceptable to the user, providing the optimal tone and detailed information. The input is the emotional data and prediction results, and the output is the adjusted notification content.

[1048] Step 9:

[1049] The server notifies the farmer of the best time and method to harvest, sending specific suggestions via push notification or email. For example, it might say, "Start harvesting next Tuesday morning. Current soil moisture and forecasted weather suggest the best possible harvest." The input is the tailored notification content, and the output is the notification message.

[1050] Step 10:

[1051] The user harvests based on the suggestions and provides feedback on the results. Feedback data is sent to the server via smartphone or PC. The input is the user's harvest results and information on the validity of the suggestions, and the output is new feedback data.

[1052] Step 11:

[1053] The server uses the collected feedback data to retrain the generative AI model. By adding new data, the accuracy of the model improves. The input is the feedback data, and the output is an improved generative AI model.

[1054] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1055] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1056] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1057] [Fourth embodiment]

[1058] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1059] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1060] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1061] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1062] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1063] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1064] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1065] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1066] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1067] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1068] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1069] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1070] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1071] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system can be implemented as follows.

[1072] Data collection

[1073] The server collects weather, soil, and crop growth data via the internet and sensors. Weather data is automatically obtained daily through a weather API, and soil data is received in real time from soil sensors installed in the farmland. Crop growth status is collected from drone photography data and periodically captured image data.

[1074] Data Preprocessing

[1075] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces a single outlier, it is imputed with the average of the other normal values.

[1076] Training generative AI models

[1077] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[1078] Predicting optimal harvest times and methods

[1079] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[1080] Proposal Notification

[1081] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[1082] Gathering feedback and retraining

[1083] The user harvests based on the suggestions and provides feedback on the results. This sends data about the actual results of the harvest and the effectiveness of the suggestions to the server. The server then uses this feedback data to retrain the generative AI model, further improving the accuracy of the next prediction.

[1084] Specific examples

[1085] 1. Data collection example: The server retrieves one-week weather forecast data for Tokyo from the weather API at 5:00 AM every day. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from sensors installed in farmland.

[1086] 2. Preprocessing: The server interpolates the 70% of humidity data detected as abnormal values ​​with the average value (60%) of the other normal values.

[1087] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[1088] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[1089] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[1090] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[1091] The above is a specific embodiment for carrying out the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing the yield and quality of their crops.

[1092] The processing flow will be explained below.

[1093] Step 1: Data collection

[1094] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[1095] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[1096] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[1097] Step 2: Data Preprocessing

[1098] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format.

[1099] The server imputes missing data with the mean or median, detects outliers and replaces them with reasonable values.

[1100] The server performs noise removal and creates a preprocessed dataset, for example, by interpolating outliers from a sensor with the average of normal values ​​from other sensors.

[1101] Step 3: Model generation

[1102] The server splits the preprocessed dataset into a training dataset and a test dataset.

[1103] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[1104] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[1105] Step 4: Optimization

[1106] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[1107] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[1108] Step 5: Notification

[1109] The server then notifies the farmer of the results of the predictions via push notification, email, or an application.

[1110] The device will then display the received notification to the user, for example, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected."

[1111] Step 6: Feedback

[1112] The user carries out harvesting work based on the suggestions and provides feedback on the results through the application.

[1113] The server collects feedback data on harvest results and the effectiveness of the proposals, specifically on harvest volume, quality, and work efficiency.

[1114] The server retrains the generative AI model based on the collected feedback data to improve the accuracy of the next prediction, including harvest result feedback data.

[1115] The above is the processing procedure of the program of the present invention.

[1116] Example 1

[1117] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1118] Agricultural harvesting depends on complex factors such as weather conditions, soil conditions, and crop growth, making it difficult to properly evaluate these and determine the optimal harvest time and method. Furthermore, methods that rely on experience and intuition are prone to variations in harvest volume and quality, making efficient agricultural management difficult. Furthermore, conventional systems lack the ability to utilize feedback for improvements, making it difficult to improve prediction accuracy.

[1119] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1120] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying farmers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data from farmers and retraining the generative AI model to improve it, means for detecting and imputing outliers and imputing missing data with mean or median values, and means for iteratively training based on past data to evaluate and improve model performance. This makes it possible to improve the accuracy of harvest time and method predictions, maximize harvest volume and quality, and streamline agricultural management.

[1121] "Weather data" refers to data collected based on meteorological observations that indicates atmospheric conditions such as temperature, humidity, precipitation, wind speed, and sunshine hours.

[1122] "Soil data" refers to data measuring the condition of soil in agricultural or cultivated areas, such as humidity, pH, temperature, and nutrient content.

[1123] "Crop growth status" refers to data indicating the growth and health status of the crop under cultivation, such as its height, leaf area, color, and the presence or absence of pests and diseases.

[1124] "Preprocessing" refers to a series of processes that supplement missing values ​​and correct outliers in collected data to prepare it in a format suitable for analysis and model training.

[1125] A "generative AI model" is an artificial intelligence model trained from pre-processed data using machine learning algorithms to automatically perform specific tasks or predictions.

[1126] The "optimal harvest time and harvest method" refers to the most suitable time and method for harvesting a crop, determined by a comprehensive assessment of weather conditions, soil conditions, and the growth state of the crop.

[1127] "Notification" refers to the act or means of transmitting information to inform users of predictions or suggestions.

[1128] "Feedback data" refers to data on the results and evaluation of harvesting work carried out by farmers based on the suggestions.

[1129] "Retraining" is the training process of updating and improving existing generative AI models using new collected data.

[1130] "Outlier detection and imputation" is the process of detecting values ​​in a dataset that fall outside the normal range and, if necessary, replacing them with other reasonable values.

[1131] "Iterative training" is a technique in which a generative AI model is trained multiple times using the same or extended dataset to improve its performance.

[1132] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data, and suggests optimal harvest times and methods to farmers. This system is implemented using the following hardware and software.

[1133] Hardware and software used

[1134] Weather data collection method: Daily weather data is collected using a weather API (e.g., OpenWeatherMap API).

[1135] Soil data collection methods: Soil sensors (e.g., soil moisture sensors, pH sensors) are used to collect real-time soil data.

[1136] Methods for collecting crop growth status: Drones and ground cameras are used to photograph crop growth status and collect image data.

[1137] Generative AI models: Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models to predict the optimal time and method of harvesting.

[1138] Data preprocessing: Programs for imputing missing data and detecting and correcting outliers (e.g., Python's Pandas library).

[1139] Explanation of the program processing flow

[1140] Data collection

[1141] The server collects data in the following ways:

[1142] Weather data is automatically collected via a weather API at 5:00 AM every day.

[1143] Soil data is received in real time from soil sensors installed on farmland.

[1144] The growth status of the crops is collected from image data taken by drones on a regular basis and from image data taken by ground cameras.

[1145] Data Preprocessing

[1146] The server pre-processes the collected raw data:

[1147] If there are missing data, they are imputed with the mean or median.

[1148] Anomalies are detected and interpolated with the average of other normal values. For example, if 70% of the data obtained from a soil moisture sensor is abnormal, it is interpolated with the average of other normal values.

[1149] Training generative AI models

[1150] The server trains a generative AI model using the preprocessed dataset:

[1151] Use machine learning frameworks (e.g., TensorFlow, PyTorch) to train models based on historical weather, soil, and crop growth data.

[1152] Cross-validation is performed to evaluate the model's performance and hyperparameters are adjusted as needed.

[1153] Predicting optimal harvest times and methods

[1154] The server inputs the latest data into the generative AI model and makes a prediction:

[1155] For example, new weather forecast data and soil moisture data are input to generate a prediction that "the best time to harvest is next Tuesday at 8:00 a.m."

[1156] Proposal Notification

[1157] The server then uses the prediction results to provide farmers with optimal harvesting suggestions:

[1158] Notifications are sent to farmers' devices (smartphones or computers) via push notifications or email.

[1159] For example, make a specific suggestion like, "Please start harvesting next Tuesday at 8:00 a.m. The current soil moisture and weather forecast will result in the best possible harvest."

[1160] Gathering feedback and retraining

[1161] The user harvests based on the suggestions and provides feedback on the results:

[1162] Users input data on harvest yield and quality through the terminal.

[1163] The server uses the collected feedback data to retrain the generative AI model and improve the accuracy of its predictions.

[1164] Specific examples

[1165] Data collection example: Every day at 5:00 AM, the server retrieves one-week weather forecast data from the weather API. At the same time, it receives soil moisture data (e.g., 60%, 61%, 62%, 59%) from the soil sensor.

[1166] Preprocessing: The 70% of humidity data detected as abnormal values ​​are interpolated with the average normal value (60%).

[1167] Model generation: Train a generative AI model using weather data, soil data, and crop results from the past five years.

[1168] Optimization: Input the latest data and predict that "8:00 a.m. next Tuesday is the best time to harvest."

[1169] Notification: The prediction results will be sent to the user's smartphone via push notification.

[1170] Feedback: Users harvest and input their results into the app, and the server uses that data to retrain the generative AI model.

[1171] As described above, farmers can use the system to know the optimal harvesting time and method in advance, maximizing harvest volume and quality.

[1172] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1173] Step 1: Data collection

[1174] The server collects weather data, soil data, and crop growth status data. Inputs are data from weather APIs, soil sensors, drones, and ground cameras. Weather data is automatically acquired at 5:00 AM every day, and soil data is received in real time. Crop growth status data is captured periodically. The output of the data collection step is a raw dataset of weather data, soil data, and crop growth status data.

[1175] Step 2: Data Preprocessing

[1176] The server preprocesses the collected raw data. The input is the raw data collected in step 1. First, it imputes missing data with the mean or median, and detects and corrects outliers. For example, if the data from a soil moisture sensor contains an outlier of 70%, it imputes it with the mean of the other normal values. The output is a preprocessed dataset.

[1177] Step 3: Training the generative AI model

[1178] The server trains a generative AI model using the preprocessed dataset. The input is the dataset preprocessed in step 2. Using a machine learning framework (e.g., TensorFlow, PyTorch), the model is trained based on historical weather data, soil data, and crop growth data. Cross-validation is performed during the training process to evaluate and improve the model's performance. The output is a trained generative AI model.

[1179] Step 4: Predict the optimal harvest time and method

[1180] The server inputs the latest data into the generative AI model to predict the optimal harvest time and method. The inputs are the latest weather data, soil data, and growth data. For example, by combining weather data and soil moisture data, a prediction is generated that "the optimal harvest time is 8:00 a.m. next Tuesday." The output is a prediction of the specific harvest time and method.

[1181] Step 5: Proposal Notification

[1182] The server notifies the farmer of the optimal harvesting suggestion based on the prediction results. The input is the prediction result from step 4. The notification is sent to the farmer's device (smartphone or PC) via push notification or email. For example, a specific suggestion may be made such as, "Please start harvesting at 8:00 a.m. next Tuesday. Based on the current soil moisture and weather forecast, the best harvest is expected." The output is the notified suggestion.

[1183] Step 6: Gather feedback and retrain

[1184] The user harvests based on the suggestions and provides feedback on the results. The input is the harvest result data. The user inputs data on the harvest quantity and quality through their terminal, and the server uses the feedback data to retrain the generative AI model. The output is an improved generative AI model.

[1185] (Application example 1)

[1186] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1187] Optimizing harvest times in agriculture is a crucial issue for improving yield and quality. Traditionally, farmers decide when to harvest based on their own experience and intuition, which can lead to errors in judgment and oversights. Furthermore, there are limited concrete methods for improving the efficiency and automating harvesting work. Therefore, there is a need for a system that automatically optimizes harvest times and methods and enables autonomous vehicles to carry out harvesting work based on the results.

[1188] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1189] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining to improve the generative AI model, means for generating a driving plan for an autonomous vehicle based on the prediction and for the autonomous vehicle to automatically start harvesting work based on the driving plan, and means for collecting driving results of the autonomous vehicle as feedback and retraining to improve the generative AI model. This automatically optimizes the harvest time and harvesting method, realizes automation and efficiency of harvesting work, and enables improved harvest volume and quality.

[1190] "Weather data" is information indicating weather conditions such as weather, temperature, precipitation, humidity, and wind speed.

[1191] "Soil data" refers to information about agricultural soil, including soil moisture, pH value, nutrient content, etc.

[1192] "Crop growth status" is information indicating the growth stage and health status of the crop, and includes, for example, germination, growth, flowering, and harvestable status.

[1193] "Preprocessing" refers to the process of preparing raw data in an analyzable format, and includes filling in missing data and correcting outliers.

[1194] A "generative AI model" is an artificial intelligence model trained to use collected data to predict optimal harvest times and methods.

[1195] "Training" refers to the process by which an AI model learns from collected data so that it can make accurate predictions.

[1196] "Prediction" means that a generative AI model infers future situations based on input data.

[1197] "Notification" refers to the method of informing farmers of the prediction results, and includes push notifications and emails.

[1198] An "autonomous vehicle" is a vehicle that operates automatically based on programmed instructions to perform agricultural tasks.

[1199] An "operation plan" is a plan that shows the route and work procedures required for an autonomous vehicle to carry out harvesting work.

[1200] "Feedback" refers to collecting information based on the results of harvesting operations and actual conditions, and using it to improve the system and retrain the AI ​​model.

[1201] "Retraining" is the process of re-learning a generative AI model to improve its performance using newly collected data.

[1202] This invention is a system for optimizing agricultural harvesting operations, which uses a generative AI model to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. It also has the function of generating operation plans for autonomous vehicles based on the prediction results and automatically carrying out harvesting operations.

[1203] Data collection

[1204] The server uses a weather API to collect weather data. This automatically obtains the latest weather, temperature, precipitation, humidity, wind speed, and other weather conditions every day. It also collects soil data in real time using soil sensors installed in the farmland. This soil data includes soil moisture, pH, and nutrient content. Furthermore, drones are used to capture images of the growing conditions of the crops and collect the data. This allows the growth stage and health of the crops to be understood.

[1205] Data Preprocessing

[1206] The server preprocesses the collected raw data: missing data is imputed with the mean or median, and outliers are detected and corrected. For example, if a soil moisture sensor produces an abnormal value, it is imputed with the average of the other normal values.

[1207] Training generative AI models

[1208] The server uses the preprocessed dataset to train a generative AI model, which uses historical weather, soil, and crop growth data to predict optimal harvest times and methods. Machine learning techniques are used for training, and the model undergoes iterative training to evaluate and improve its performance.

[1209] Predicting optimal harvest times and methods

[1210] The server inputs the latest data into a generative AI model to predict the optimal harvest time and method. For example, by combining weather forecasts and soil moisture data, a specific prediction can be made, such as "The best time to harvest is next Tuesday morning."

[1211] Proposal Notification

[1212] The server then sends farmers optimal harvest suggestions based on the forecast results. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, specific suggestions may be made such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected."

[1213] Autonomous vehicles perform harvesting operations

[1214] The server generates a driving plan for the autonomous vehicle based on the predictions. Based on this driving plan, the autonomous vehicle automatically starts harvesting work. For example, instructions such as "Harvesting work will automatically start at 8:00 a.m. next Tuesday" are sent to the autonomous vehicle.

[1215] Gathering feedback and retraining

[1216] Users harvest based on the suggestions and provide feedback on the results. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and uses this data to retrain the generative AI model, further improving the accuracy of the next prediction.

[1217] Specific examples

[1218] 1. Data Collection Example: The server retrieves a week's worth of weather data for a city from a weather API at 5:00 AM every day. At the same time, it receives soil moisture data from sensors installed in farmland.

[1219] 2. Data preprocessing: The server interpolates the humidity data detected as abnormal values ​​with the average value of other normal values.

[1220] 3. Model generation: The server trains a generative AI model that uses historical weather data, soil data, and harvest results to predict optimal harvest times.

[1221] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[1222] 5. Notification: The server sends the prediction results to the user's smartphone via push notification.

[1223] 6. Feedback: Users harvest crops and input their results into the app. The server receives this feedback data and retrains the generative AI model.

[1224] Prompt Sentence Examples

[1225] Enter the weather, soil and crop growth data as follows:

[1226] Weather data: Temperature=30, Humidity=70, Precipitation=0

[1227] Soil data: soil moisture=60, pH=6.5

[1228] Crop growth data: Growth stage=vegetative, Image data=base64_encoded_image_string

[1229] By inputting this data into a generative AI model, the optimal harvest time and method can be predicted.

[1230] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1231] Step 1: Data collection

[1232] The server obtains weather data from the weather API and collects soil and crop growth data from soil sensors and drones installed in the farmland. The server sends requests to the weather API to obtain information such as temperature, humidity, and precipitation, and obtains soil humidity and pH data from the soil sensors. It also operates a drone to take images of the crops and collect data on their growth status. The input data consists of weather data, soil sensor data, and crop image data, and these are used to proceed to the next step.

[1233] Step 2: Data Preprocessing

[1234] The server preprocesses the collected weather data, soil data, and crop growth data. If there are missing data or outliers, it fills them in with the mean or median and corrects the outliers. For example, if data from a soil moisture sensor contains an outlier, the server fills it in using the mean value of other normal data. The input data is the collected raw data, which is then processed to output preprocessed data.

[1235] Step 3: Training the generative AI model

[1236] The server uses the preprocessed dataset to train a generative AI model. It uses historical weather, soil, and crop growth data to build a multi-layer neural network and train a model to predict optimal harvest times and methods. The server uses the preprocessed dataset as input data and outputs a trained generative AI model.

[1237] Step 4: Predict harvest time and method

[1238] The server inputs the latest data into a trained generative AI model to predict the optimal harvest time and method. For example, it uses next week's weather data and current soil moisture data to obtain a specific prediction result, such as "The optimal time to harvest is next Tuesday morning." The input data is the latest weather data, soil data, and crop growth data, and the output data is the prediction result of the optimal harvest time and method.

[1239] Step 5: Proposal Notification

[1240] Based on the prediction results, the server notifies the farmer of the optimal harvest time and harvesting method. Notifications are sent to the farmer's device (smartphone, PC, etc.) via push notification or email. For example, a message could be sent saying, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The input data is the prediction result, and the output data is the notification message.

[1241] Step 6: Generate a driving plan for the autonomous vehicle

[1242] The server generates a driving plan for the autonomous vehicle based on the predictions. The autonomous vehicle then automatically starts harvesting according to the generated driving plan. For example, an instruction such as "Harvesting will begin automatically at 8:00 AM next Tuesday" is sent to the autonomous vehicle. The input data is the predicted results of the optimal harvesting time and method, and the output data is the driving plan and specific instructions for the autonomous vehicle.

[1243] Step 7: Gather feedback

[1244] The user harvests based on the suggestions and sends the results as feedback to the server. The server collects data on the actual results of the harvest and the effectiveness of the suggestions, and retrains the generative AI model to improve future prediction accuracy. The input data are the harvest results and feedback data, and the output data is an improved generative AI model.

[1245] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1246] This invention is a system for optimizing agricultural harvesting operations, combining a generative AI model and an emotion engine to analyze weather data, soil data, and crop growth data to suggest optimal harvest times and methods to farmers. This system recognizes the user's emotions and uses them as feedback, enabling even more accurate suggestions.

[1247] Data collection

[1248] The server obtains weather data daily through a weather API. The obtained data includes temperature, precipitation, humidity, wind speed, and wind direction. The server also receives soil data in real time from soil sensors. The soil data includes soil humidity, pH, moisture content, and nutrient content. The server also collects information on the growth status of crops from drone photography data and periodically captured image data. Specifically, the image data includes crop height, leaf color, leaf area, and fruit size.

[1249] Data Preprocessing

[1250] The server converts the collected raw data into an analyzable form. For example, it standardizes the acquired data format and fills in missing data with the mean or median. Outliers are detected and replaced with reasonable values. It also performs noise removal and creates a preprocessed data set. For example, it fills in outliers obtained from a soil moisture sensor with the mean of other normal values.

[1251] Training generative AI models

[1252] The server splits the preprocessed dataset into a training dataset and a test dataset and trains a generative AI model that uses historical weather, soil, and crop growth data to predict the optimal time and method of harvesting. Machine learning techniques are used for training, and iterative training is performed to evaluate and improve the model's performance.

[1253] Predicting optimal harvest times and methods

[1254] The server inputs the latest weather, soil, and crop growth data into the generative AI model to predict the optimal harvest time and method, generating specific predictions such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[1255] Collecting and analyzing user emotion data

[1256] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine how the user feels about the harvesting suggestions (satisfaction, dissatisfaction, stress, etc.).

[1257] Proposal Notification

[1258] The server then uses the prediction results and emotion data to provide farmers with optimal harvest suggestions. These suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best possible harvest is expected." The tone and details of the suggestion may also be adjusted based on the user's emotion data.

[1259] Gathering feedback and retraining

[1260] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and uses it to retrain the generative AI model, further improving the accuracy of the next prediction.

[1261] Specific examples

[1262] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[1263] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[1264] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[1265] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[1266] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[1267] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[1268] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[1269] The above is a specific example of how to implement the invention. This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotions also increase the likelihood of suggestions being accepted.

[1270] The processing flow will be explained below.

[1271] Program processing

[1272] Step 1: Data collection

[1273] The server retrieves weather data daily through a weather API, including temperature, precipitation, humidity, wind speed, and wind direction.

[1274] The server collects real-time soil data from soil sensors, specifically soil humidity, pH, water content, and nutrient content.

[1275] The server collects data on the growth status of the crops from drones and periodic photographs, including image data on the height, leaf color, leaf area, and fruit size of the crops.

[1276] Step 2: Data Preprocessing

[1277] The server converts the collected raw data into an analyzable form, for example by standardizing the acquired data format and filling in missing data with the mean or median.

[1278] The server detects outliers and replaces them with reasonable values, for example, interpolating outliers from a soil moisture sensor with the average of other normal values.

[1279] The server performs noise removal and creates a preprocessed dataset.

[1280] Step 3: Model generation

[1281] The server splits the preprocessed dataset into a training dataset and a test dataset.

[1282] The server trains a generative AI model that uses historical weather, soil, and crop growth data to predict the best time and method for harvesting.

[1283] The server evaluates the model's performance using a test dataset and adjusts the hyperparameters as necessary, using evaluation metrics such as precision, recall, and F-measure.

[1284] Step 4: Optimization

[1285] The server inputs the latest weather, soil, and crop growth data into the generative AI model.

[1286] The server calculates the optimal harvest time and method predicted by the model, generating a specific prediction result such as "8:00 a.m. next Tuesday is the optimal time to harvest."

[1287] Step 5: Collecting sentiment data

[1288] The server collects user emotion data using an emotion engine, which uses voice and face recognition technology to determine emotions from the user's tone of voice and facial expressions.

[1289] The device analyzes the acquired emotional data and evaluates the user's stress level and satisfaction.

[1290] Step 6: Generate and send notifications

[1291] The server then generates optimal harvesting suggestions for the user based on the prediction results and emotion data, such as "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best harvest is expected."

[1292] The server sends the generated suggestions to the user's device via push notification, email, or application.

[1293] Step 7: Gather feedback and retrain

[1294] The user harvests based on the suggestions and provides feedback on the results through the application, with data on the harvest results and the effectiveness of the suggestions being sent to the server.

[1295] The server retrains the generative AI model based on the collected feedback and emotion data to improve the accuracy of the next prediction.

[1296] Specific examples

[1297] 1. Data collection example: The server obtains one-week weather forecast data for Tokyo from the weather API at 5:00 a.m. every day. At the same time, it receives real-time soil moisture data (e.g., 60%, 61%, 62%, 59%) from soil sensors installed in farmland.

[1298] 2. Preprocessing: The server interpolates the 70% of soil moisture data detected as abnormal values ​​with the average value (60%) of the other normal values.

[1299] 3. Model generation: The server uses weather data, soil data, and harvest results from the past five years to train a generative AI model that predicts the optimal harvest time.

[1300] 4. Optimization: The server inputs the latest data into the model and obtains a prediction result, such as "8:00 a.m. next Tuesday is the best time to harvest."

[1301] 5. Emotion data collection: The emotion engine recognizes the user's voice and facial expressions when they receive the suggestion notification and collects that data.

[1302] 6. Notification: Based on the prediction results and emotion data, the server sends specific harvesting suggestions via push notification to the user's smartphone.

[1303] 7. Feedback: Users actually harvest crops and input their results and emotions into the app. The server receives this feedback and emotion data and retrains the generative AI model.

[1304] The above is a specific embodiment for carrying out the present invention.

[1305] Example 2

[1306] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1307] In agriculture, determining the optimal harvest time and method is extremely important, but it is difficult to make accurate predictions due to the influence of numerous factors, such as weather conditions, soil conditions, and crop growth status. Furthermore, adjusting suggestions based on the farmer's emotions and stress levels is also essential to improving work efficiency. There is a need for a system that can comprehensively analyze these complex factors and make optimal harvest suggestions for farmers.

[1308] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1309] In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth conditions, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvest method using the trained generative AI model, means for collecting and analyzing user emotion data, means for notifying farmers of the optimal harvest time and harvest method based on the prediction and the user emotion data, and means for collecting feedback data and emotion data and retraining to improve the generative AI model. This enables accurate harvest predictions based on the collected data and individual responses according to farmers' emotions, thereby maximizing harvest volume and quality and improving work efficiency.

[1310] "Weather data" refers to various information about weather conditions, such as temperature, precipitation, humidity, wind speed, and wind direction.

[1311] "Soil data" refers to information that indicates the condition of the soil, such as soil humidity, pH, water content, and nutrient content.

[1312] "Crop growth status" refers to information about the physical growth of the crop, such as crop height, leaf color, leaf area, and fruit size.

[1313] "Preprocessing" refers to converting collected data into an analyzable form, standardizing data formats, filling in missing data, detecting and replacing outliers, and removing noise.

[1314] A "generative AI model" is a model trained using machine learning techniques and is an algorithm that predicts the optimal harvest time and method based on collected data.

[1315] "User emotion data" refers to data related to emotions obtained from the user's tone of voice and facial expressions determined using voice recognition and facial recognition technology.

[1316] "Feedback data" refers to data on the results of actual harvesting work and the effectiveness of suggestions.

[1317] This invention is a system for optimizing agricultural harvesting operations. The system combines a generative AI model and an emotion engine to analyze weather, soil, and crop growth data to suggest optimal harvest times and methods to farmers.

[1318] Data collection

[1319] The server obtains weather data daily through a weather API. This data includes temperature, precipitation, humidity, wind speed, and wind direction. For example, the temperature might be 20°C, the probability of precipitation 40%, and humidity 70%. The server also receives real-time soil data from soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. For example, the server verifies that the soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. The server also collects crop growth status data from drone photography. Specifically, this data includes image data such as crop height, leaf color, leaf area, and fruit size. For example, the server verifies that the crop height is 30 cm, leaf color is a healthy green, leaf area is 15 square centimeters, and fruit size is 5 cm.

[1320] Data Preprocessing

[1321] The server converts the collected raw data into an analyzable form. First, it standardizes the data format and fills in missing data with the average or median. For example, missing temperature data for one day is filled in with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, it fills in the 70% soil humidity data detected as an outlier with the average value (60%) of the other normal values. It also removes noise to create a preprocessed dataset.

[1322] Training generative AI models

[1323] The server splits the preprocessed dataset into training data and test data and trains a generative AI model. This model is used to predict the optimal harvest time and method based on historical weather, soil, and crop growth data. Machine learning techniques are used for training, and iterative training is performed to achieve high accuracy. For example, the model is trained using weather data, soil data, and harvest results from the past five years.

[1324] Predicting optimal harvest times and methods

[1325] The server inputs the latest weather, soil, and crop growth data into the generative AI model, which predicts the optimal harvest time and method. The generative AI model then provides specific predictions, such as "The optimal harvest time is next Tuesday at 8:00 AM."

[1326] Collecting and analyzing user emotion data

[1327] The server collects user emotion data using an emotion engine. This emotion engine uses voice and face recognition technology to determine the user's emotion from their tone of voice and facial expression. For example, it can determine the user's emotion (satisfaction, dissatisfaction, stress, etc.) in response to the harvesting suggestion.

[1328] Proposal Notification

[1329] The server then sends farmers optimal harvesting suggestions based on the predictions of the optimal harvesting time and method, as well as the user's emotional data. The suggestions are sent to the farmer's device (such as a smartphone or PC) via push notification or email. For example, the suggestion might be specific, such as, "Please start harvesting next Tuesday morning. Based on the current soil humidity and weather forecast, the best possible harvest is expected." The tone and detailed information of the suggestions may also be adjusted based on the user's emotional data.

[1330] Gathering feedback and retraining

[1331] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The server also collects user sentiment data and retrains the generative AI model to further improve prediction accuracy next time.

[1332] This system allows farmers to know the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on user emotion data also increase the likelihood of suggestions being accepted.

[1333] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1334] Step 1:

[1335] The server collects weather data through the weather API at 5:00 AM every day. Specifically, it obtains data such as temperature, precipitation, humidity, wind speed, and wind direction. It receives this weather data as input and stores it in a database. For example, it obtains data such as a temperature of 20 degrees, a 40% chance of precipitation, and a 70% humidity. As output, the weather data is recorded in the database.

[1336] Step 2:

[1337] The server collects soil data in real time from soil sensors installed in farmland. Specifically, data such as soil humidity, pH, moisture content, and nutrient content are acquired. The server receives the soil data as input and stores it in a database. For example, data is collected showing that soil humidity is 60%, pH is 6.5, moisture content is 25%, and nutrient content is within the appropriate range. As output, the soil data is recorded in the database.

[1338] Step 3:

[1339] The server uses drone image data to collect information on the growth status of crops. Specifically, it analyzes image data such as crop height, leaf color, leaf area, and fruit size. It receives this crop growth data as input and stores it in a database. For example, it obtains data such as a crop height of 30 cm, a healthy green leaf color, a leaf area of ​​15 square centimeters, and a fruit size of 5 cm. As output, the crop growth data is recorded in the database.

[1340] Step 4:

[1341] The server preprocesses the collected weather, soil, and crop growth data. First, it standardizes the data format and imputes missing data with the mean or median. For example, missing temperature data for one day is imputed with the average temperature data for other days. Next, it detects outliers and replaces them with reasonable values. For example, soil moisture data detected as an outlier at 70% is imputed with 60%, the average of the other normal values. This preprocessed data is received as input and noise is removed. A preprocessed dataset is created as output.

[1342] Step 5:

[1343] The server splits the preprocessed dataset into training data and test data and trains the generative AI model. For example, it uses weather data, soil data, and harvest results from the past five years. It receives the training data as input and trains the AI ​​model. It then evaluates the model's performance using the test data. The output is a highly accurate generative AI model.

[1344] Step 6:

[1345] The server inputs the latest weather, soil, and crop growth data into the AI ​​model to predict the optimal harvest time and method. For example, it generates a specific prediction result, such as "8:00 AM next Tuesday is the optimal harvest time." It receives this prediction result as input and generates information to notify farmers. The output is the predicted harvest time and method.

[1346] Step 7:

[1347] The server uses an emotion engine to collect and analyze the user's emotional data. Specifically, it uses voice recognition and face recognition technology to determine the user's emotions from their tone of voice and facial expressions. It receives the emotional data as input and analyzes the user's emotional state. The output is the user's emotional data.

[1348] Step 8:

[1349] The server notifies the farmer of optimal harvesting suggestions based on the prediction results of the optimal harvesting time and method and the user's emotional data. Specifically, the suggestion is sent via push notification or email to the farmer's device (smartphone, PC, etc.). For example, the suggestion might be, "Please start harvesting next Tuesday morning. Based on the current soil moisture and weather forecast, the best harvest is expected." The suggestion notification is sent to the farmer as output.

[1350] Step 9:

[1351] The user harvests based on the suggestions and provides feedback on the results through the application. Data on the harvest results and the effectiveness of the suggestions is sent to the server. The feedback data is received as input and used to improve the AI ​​model. The output is a retrained generative AI model.

[1352] The above is a specific example of how to implement the invention. Using this system, farmers can predict the optimal harvest time and method in advance, maximizing yield and quality. Furthermore, personalized responses based on the user's emotional data also increase the likelihood of their suggestions being accepted.

[1353] (Application example 2)

[1354] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1355] In conventional agricultural harvesting work, determining the optimal harvesting time and method is difficult, and in many cases workers have had to rely on experience and intuition. As a result, harvest volume and quality are unstable, making it difficult for agricultural workers to harvest at the appropriate time. Furthermore, when harvesting suggestions are simply notified mechanically, they do not take into account the emotions and reactions of agricultural workers, making them less likely to accept the suggestions. This has led to a demand for a system that can provide optimal harvesting suggestions and reduce the burden on agricultural workers.

[1356] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting weather data, means for collecting soil data, means for collecting crop growth status, means for preprocessing the weather data, the soil data, and the crop growth data, means for training a generative AI model based on the preprocessed data, means for predicting the optimal harvest time and harvesting method using the trained generative AI model, means for notifying agricultural workers of the optimal harvest time and harvesting method based on the prediction, means for collecting feedback data and retraining the generative AI model to improve it, means for recognizing the emotions of users who receive the prediction notification and collecting emotion data, means for adjusting the notification content based on the emotion data, and means for using the collected emotion data to improve the generative AI model. This makes it possible to simultaneously optimize harvesting operations and increase the likelihood that agricultural workers will accept suggestions based on their emotions.

[1357] "Weather data" refers to information related to weather, including data such as temperature, precipitation, humidity, wind speed, and wind direction.

[1358] "Soil data" refers to information about the condition of the soil, including data on soil humidity, pH, water content, nutrient content, etc.

[1359] "Crop growth status" is information that indicates how the crop is growing, and includes data such as the height of the crop, leaf color, leaf area, and fruit size.

[1360] A "generative AI model" is an artificial intelligence model that is trained on past data to predict specific outcomes.

[1361] "Emotional data" refers to information about a user's emotional state, including data derived from vocal tone and facial expressions.

[1362] The "means for adjusting the notification content" is a means for changing the notification content regarding the optimal harvesting time and harvesting method based on the user's emotional data.

[1363] "Feedback Data" refers to information regarding actual harvest results and the effectiveness of recommendations, including data collected from users.

[1364] The "training means" is the process of training a generative AI model using preprocessed data.

[1365] "Preprocessing" refers to the process of converting the raw data acquired into an analyzable form.

[1366] "Retraining" is a training process that uses collected feedback data to improve the accuracy of an existing generative AI model.

[1367] The present invention is a system for highly optimizing agricultural harvesting operations, and is realized by including the following means.

[1368] First, the server collects weather data (temperature, precipitation, humidity, wind speed, wind direction, etc.) using a weather API at a specific time each day. Next, soil data is collected in real time using soil sensors. This soil data includes soil humidity, pH, moisture content, and nutrient content. Additionally, cameras and drones installed in the farmland are used to collect information on the growth status of the crops (plant height, leaf color, leaf area, fruit size, etc.).

[1369] The server preprocesses the collected data. Specifically, it uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, unify multiple data formats, and remove noise.

[1370] The preprocessed data is then used to train a generative AI model using machine learning frameworks such as TensorFlow and PyTorch, based on historical weather, soil, and crop growth data, which is then used to predict the optimal harvest time and method.

[1371] Using the trained generative AI model, the server inputs the latest data and predicts the best time and method for harvesting, generating a specific prediction result, for example, "The best time to harvest is next Tuesday at 8:00 AM."

[1372] In addition, the server uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to collect the user's emotional data. The server analyzes the user's tone of voice and facial expressions when receiving the notification and collects emotional data. Based on this emotional data, the server determines the user's emotional state, such as "comfortable," "anxious," or "confused," and adjusts the content of the notification accordingly.

[1373] Finally, the server notifies the farmer of the best time and method to harvest. The notification is sent via push notification or email. The content of the notification is adjusted based on the sentiment data, so it is presented in a way that is easy for the farmer to accept and understand.

[1374] Users harvest based on the suggestions and provide feedback on the results via their smartphone or computer. This feedback data is sent to the server and used to retrain the existing generative AI model, allowing the system to further improve its prediction accuracy in the future.

[1375] For example, a weather API retrieves the weather forecast for the coming week at 8:00 a.m. every morning, and a soil sensor collects soil moisture data every hour. The user's reaction to the harvest suggestion can be collected and the next notification can be tailored based on that reaction. A prompt such as "Please retrieve the weather forecast for next week and collect data to predict the optimal harvest time" can be used to create input data for a generative AI model.

[1376] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1377] Step 1:

[1378] The server collects weather data. Specifically, it uses a weather API to obtain data such as temperature, precipitation, humidity, wind speed, and wind direction at a specific time each day. This data is stored on the server in JSON format. The input is raw data obtained from the weather API, and the output is weather data in a unified JSON format.

[1379] Step 2:

[1380] The server collects soil data. Soil sensors acquire data such as soil humidity, pH, moisture content, and nutrient content in real time and send it to the server. This data is stored in the server in RAW format. The input is raw data from the soil sensors, and the output is formatted soil data.

[1381] Step 3:

[1382] The server collects information on the growth status of crops. It analyzes image data taken using cameras and drones installed on farmland to extract information such as crop height, leaf color, leaf area, and fruit size. It preprocesses the acquired image data to generate the necessary growth data. The input is image data from the cameras and drones, and the output is analyzed growth data.

[1383] Step 4:

[1384] The server integrates preprocessed weather data, soil data, and crop growth data. It uses Python libraries (Pandas, NumPy, etc.) to fill in missing data, detect and correct outliers, and unify multiple data formats. The input is individual data before preprocessing, and the output is the integrated, preprocessed data.

[1385] Step 5:

[1386] The server trains a generative AI model based on the preprocessed data. Using TensorFlow and PyTorch, the model is trained based on historical weather, soil, and crop growth data to generate a model that can predict the optimal harvest time and method. The input is the preprocessed data, and the output is the trained generative AI model.

[1387] Step 6:

[1388] The server uses a generative AI model to predict the optimal harvest time and method. The latest data is input into the model to generate a specific prediction. For example, the output may be "The optimal harvest time is 8:00 a.m. next Tuesday." The input is the latest data and a trained generative AI model, and the output is a prediction of the harvest time and method.

[1389] Step 7:

[1390] The server collects the user's emotional data. It uses a speech recognition API (Google Cloud Speech-to-Text) and a facial recognition API (Microsoft Azure Face API) to analyze the user's tone of voice and facial expression when they receive the suggestion notification. This emotional data is stored on the server. The input is the user's voice and facial image data, and the output is the analyzed emotional data.

[1391] Step 8:

[1392] The server adjusts the notification content based on the emotional data. Based on the collected emotional data, it modifies the suggestions to make them more acceptable to the user, providing the optimal tone and detailed information. The input is the emotional data and prediction results, and the output is the adjusted notification content.

[1393] Step 9:

[1394] The server notifies the farmer of the best time and method to harvest, sending specific suggestions via push notification or email. For example, it might say, "Start harvesting next Tuesday morning. Current soil moisture and forecasted weather suggest the best possible harvest." The input is the tailored notification content, and the output is the notification message.

[1395] Step 10:

[1396] The user harvests based on the suggestions and provides feedback on the results. Feedback data is sent to the server via smartphone or PC. The input is the user's harvest results and information on the validity of the suggestions, and the output is new feedback data.

[1397] Step 11:

[1398] The server uses the collected feedback data to retrain the generative AI model. By adding new data, the accuracy of the model improves. The input is the feedback data, and the output is an improved generative AI model.

[1399] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1400] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1401] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1402] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1403] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1404] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1405] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1406] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1407] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1408] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1409] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1410] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1411] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1412] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1413] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1414] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1415] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1416] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1417] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1418] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1419] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1420] The following is further disclosed regarding the above embodiment.

[1421] (Claim 1)

[1422] a means for collecting meteorological data;

[1423] a means for collecting soil data;

[1424] means for collecting information on the growth status of the crop;

[1425] means for preprocessing the meteorological data, the soil data, and the crop growth data;

[1426] A means for training a generative AI model based on the preprocessed data; and

[1427] a means for predicting optimal harvest times and methods using a trained generative AI model;

[1428] a means for informing an agricultural worker of the optimum harvesting time and harvesting method based on said prediction;

[1429] A system including means for collecting feedback data and retraining said generative AI model to improve it.

[1430] (Claim 2)

[1431] The system of claim 1 , wherein the weather data collection means utilizes a weather API.

[1432] (Claim 3)

[1433] 10. The system of claim 1, wherein said soil data gathering means utilizes a soil sensor.

[1434] "Example 1"

[1435] (Claim 1)

[1436] a means for collecting meteorological data;

[1437] a means for collecting soil data;

[1438] means for collecting information on the growth status of the crop;

[1439] means for preprocessing the meteorological data, the soil data, and the crop growth data;

[1440] A means for training a generative AI model based on the preprocessed data; and

[1441] a means for predicting optimal harvest times and methods using a trained generative AI model;

[1442] a means for informing a farmer of the optimal harvesting time and harvesting method based...

Claims

1. a means for collecting meteorological data; a means for collecting soil data; means for collecting information on the growth status of the crop; means for preprocessing the meteorological data, the soil data, and the crop growth data; A means for training a generative AI model based on the preprocessed data; and a means for predicting optimal harvest times and methods using a trained generative AI model; a means for informing an agricultural worker of the optimum harvesting time and harvesting method based on said prediction; A system including means for collecting feedback data and retraining said generative AI model to improve it.

2. The system of claim 1 , wherein the weather data collection means utilizes a weather API.

3. 10. The system of claim 1, wherein said soil data gathering means utilizes a soil sensor.

Images