Information processing method, program, and information processing apparatus

What is AI technical title?
AI technical title is built by Patsnap AI team. It summarizes the technical point description of the patent document.
The information processing method addresses the limitation of pre-stored crop countermeasures by using a language model to integrate diverse data sources, enabling adaptive agricultural strategies for improved crop quality and yield.

JP2026011384APending Publication Date: 2026-01-23KUBOTA CORP

View PDF 1 Cites 0 Cited by

Patent Information

Application Number: JP2024111941
Authority / Receiving Office: JP · JP
Patent Type: Applications
Current Assignee / Owner
Filing Date: 2024-07-11
Publication Date: 2026-01-23

AI Technical Summary

Technical Problem

Existing programs are limited in providing flexible countermeasures for crop information on various crops, as they primarily select from pre-stored methods without adaptability.

Method used

An information processing method that acquires crop information using a language model to output flexible countermeasures, incorporating inputs from farmer terminals, agricultural machines, field sensors, and weather servers, and utilizes a language model to generate tailored agricultural recommendations.

Benefits of technology

Enables the output of flexible countermeasures for various crops, enhancing crop quality and yield by integrating real-time data from multiple sources and generating adaptive agricultural strategies.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure 2026011384000001_ABST

Patent Text Reader

Abstract

Outputting flexible countermeasures for crop information of various crops SOLUTION: An information processing method according to an embodiment of the present disclosure includes acquiring crop information regarding a crop, and outputting a countermeasure for the acquired crop information by a language model using the crop information. Further, an information processing method according to an embodiment of the present disclosure includes acquiring a query designating content of output, and outputting a countermeasure by the language model using the acquired query and the crop information. In addition, the information processing method according to an embodiment of the present disclosure acquires the crop information including the harvest information regarding the harvested crop, the tillage information regarding the agricultural machine, or the sensor information acquired from the sensor provided in the field, and outputs a countermeasure by the language model using the crop information including the harvest information, the tillage information, or the sensor information SELECTED DRAWING: Figure 1

Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present technology relates to an information processing method, a program, and an information processing device. [Background technology]

[0002] Conventionally, programs that output apple cultivation methods have been proposed. For example, the apple cultivation method suggestion program described in Patent Document 1 searches for and outputs an apple cultivation method that corresponds to reference information from information related to apple cultivation methods that has been compiled into data in advance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-111050 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the program described in Patent Document 1 merely selects an appropriate cultivation method for apples from pre-stored (digitized) methods, and is unable to output flexible countermeasures for crop information on various crops.

[0005] The present disclosure has been made in consideration of the above circumstances, and aims to output flexible countermeasures for crop information on various crops. [Means for solving the problem]

[0006] An information processing method according to an embodiment of the present disclosure acquires crop information related to crops, and outputs countermeasures for the acquired crop information using a language model that uses the crop information. [Effects of the Invention]

[0007] In the information processing method according to an embodiment of the present disclosure, flexible countermeasures for crop information on various crops can be output. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a schematic diagram of an information processing system. [Figure 2] FIG. 1 is a block diagram illustrating an example of the configuration of an information processing device. [Figure 3] FIG. 2 is a block diagram showing a configuration example of a farmer terminal. [Figure 4] FIG. 10 is an explanatory diagram illustrating an example of a farm product information table. [Figure 5] FIG. 10 is an explanatory diagram of an example of a harvest information table. [Figure 6] FIG. 10 is an explanatory diagram illustrating an example of a tillage information table. [Figure 7] FIG. 10 is an explanatory diagram illustrating an example of a sensor information table. [Figure 8] FIG. 4 is an explanatory diagram illustrating an example of a weather information table. [Figure 9] FIG. 10 is an explanatory diagram illustrating an example of a photographed image table. [Figure 10] FIG. 10 is an explanatory diagram showing an example of a harvest yield prediction model. [Figure 11] FIG. 2 is an explanatory diagram illustrating an example of a language model. [Figure 12] FIG. 10 is an explanatory diagram showing an example of a harvest information and photographed image input screen. [Figure 13] FIG. 10 is an explanatory diagram illustrating an example of a countermeasure display screen. [Figure 14] 10 is a flowchart illustrating an example of a process for registering farm product information. [Figure 15] 10 is a flowchart illustrating an example of a process for outputting a countermeasure. [Figure 16] FIG. 10 is an explanatory diagram illustrating an example of a countermeasure database. [Figure 17] FIG. 10 is an explanatory diagram illustrating an example of a language model according to the second embodiment. [Figure 18] 10 is a flowchart showing an example of a process for outputting a countermeasure according to the second embodiment. [Figure 19] 10 is a flowchart showing an example of a process for outputting a countermeasure according to a modification of the second embodiment. [Figure 20] FIG. 11 is an explanatory diagram illustrating an example of a language model according to the third embodiment. [Figure 21] FIG. 11 is an explanatory diagram showing an example of a countermeasure display screen according to the third embodiment. [Figure 22] 11 is a flowchart showing an example of a process for outputting a countermeasure according to the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] (Embodiment 1) FIG. 1 is a schematic diagram of an information processing system S. The information processing system S includes an information processing device 1, a language model server 2, a farmer terminal 3, an agricultural machine 4, a field sensor 5, and a weather server 6. In the information processing system S, the information processing device 1, the language model server 2, the farmer terminal 3, the agricultural machine 4, the field sensor 5, and the weather server 6 can communicate with each other via an external network N such as the Internet. The information processing device 1 is, for example, a server computer, and is owned by a business operator (administrator) that manages information about farmers. The farmer terminal 3 is a terminal owned by a farmer who produces agricultural products. The farmer terminal 3 is, for example, a smartphone, a tablet terminal, or a personal computer. The farmer terminal 3 accepts input of agricultural product harvest information or takes photographs, and transmits the input harvest information or photographed images to the information processing device 1 via the network N. The agricultural machine 4 is owned by the farmer and has a processing unit that acquires information about the drive of the agricultural machine 4 (tilling information) from the drive unit of the agricultural machine 4 and transmits the information to the information processing device 1 via the network N. The field sensor 5 is a sensor installed in a field where agricultural crops are grown or used in the field to detect the condition of the field. The condition of the field includes, for example, the presence or absence or severity of pest or disease infestation, as well as the water level or moisture content of the field. The field sensor 5 transmits the detected field condition as sensor information to the information processing device 1 via the network N. The weather server 6 is a server operated by the Japan Meteorological Agency or a private weather forecasting company. The information processing device 1 obtains meteorological information about the field from the weather server 6 via the network N.

[0010] The information processing device 1 acquires crop information related to crops from the farmer terminal 3, the agricultural machines 4, the field sensors 5, and the weather server 6. The crop information includes the above-mentioned harvest information, tillage information, sensor information, or weather information. The farmer terminal 3 may acquire crop information from the agricultural machines 4, the field sensors 5, and the weather server 6, and transmit the acquired crop information to the information processing device 1. The information processing device 1 may also acquire tillage information or sensor information from multiple agricultural machines 4 or field sensors 5 of different types.

[0011] The information processing device 1 also transmits the acquired crop information to the language model server 2, and inputs the crop information into the language model LM stored in the language model server 2. The language model LM outputs countermeasures for improving the quality or yield of the crops based on the input crop information. The information processing device 1 acquires the output countermeasures and transmits them to the farmer terminal 3.

[0012] FIG. 2 is a block diagram showing an example configuration of the information processing device 1. The information processing device 1 includes a processing unit 11, a storage unit 12, and a communication unit 13. The processing unit 11 is configured with a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphical Processing Unit), a quantum processor, or the like, and performs various control processes, arithmetic processes, and the like by reading and executing a program P (program product) and a database pre-stored in the storage unit 12. Note that a database server or the like may be provided outside the information processing device 1, and the processing unit 11 may read the database from the database server or the like. The information processing device 1 may also be one whose functions are realized by multiple server devices or computers. The information processing device 1 may also correspond to a node on a blockchain. The information processing device 1 may also execute some or all of the processes performed by the language model server 2 or the farmer terminal 3.

[0013] The storage unit 12 of the information processing device 1 is, for example, a volatile memory and a non-volatile memory. The storage unit 12 stores a program P, a crop information table 121, a harvest information table 122, a tillage information table 123, a sensor information table 124, a weather information table 125, a captured image table 126, and a harvest yield prediction model PM. The program P may be provided to the information processing device 1 using a computer-readable storage medium 12a. The storage medium 12a may be, for example, a portable memory. Examples of the portable memory include a CD-ROM, a USB (Universal Serial Bus) memory, an SD card, a micro SD card, and a CompactFlash (registered trademark) memory. When the storage medium 12a is a portable memory, the processing element of the processing unit 11 may read the program P from the storage medium 12a using a reading device (not shown). The read program P is written to the storage unit 12. The program P may also be provided to the information processing device 1 by the communication unit 13 communicating with an external device. The crop information table 121, harvest information table 122, tillage information table 123, sensor information table 124, weather information table 125, captured image table 126, and harvest yield prediction model PM will be described in detail later.

[0014] The communication unit 13 of the information processing device 1 is a communication module or communication interface for communicating with the language model server 2, farmer terminal 3, agricultural machine 4, field sensor 5, or weather server 6 via wired or wireless communication, and is, for example, a wide-area wireless communication module such as LTE (registered trademark), 4G, or 5G. The processing unit 11 communicates with the language model server 2, farmer terminal 3, agricultural machine 4, field sensor 5, or weather server 6 via the communication unit 13 and an external network N such as the Internet.

[0015] FIG. 3 is a block diagram showing an example of the configuration of the farmer terminal 3. In this embodiment, an example will be described in which the farmer terminal 3 is a smartphone. The farmer terminal 3 includes a terminal processing unit 31, a memory unit 32, a communication unit 33, a display unit 34, an input unit 35, and a photographing unit 36. The terminal processing unit 31 is configured with a CPU or MPU, etc., and performs various control processes, arithmetic processes, etc. The functions of the farmer terminal 3 may be realized by multiple devices. Furthermore, the farmer terminal 3 may execute some or all of the processes performed by the information processing device 1 or the language model server 2.

[0016] The storage unit 32 of the farmer terminal 3 stores an application program (application program) Pa that accepts input of harvest information related to agricultural products or takes images of agricultural products or fields and transmits the harvest information or the taken images to the information processing device 1. The application program Pa is provided to the farmer terminal 3 using, for example, a storage medium 32a. The terminal processing unit 31 of the farmer terminal 3 may obtain the application program Pa using the Internet and store it in the storage unit 32.

[0017] The communication unit 33 of the farmer terminal 3 is a communication module or communication interface for wirelessly communicating with the information processing device 1 or the language model server 2. The terminal processing unit 31 communicates with the information processing device 1 through an external network N via the communication unit 33. The farmer terminal 3 may also communicate with the agricultural machine 4 or the field sensor 5 and acquire crop information from the agricultural machine 4 or the field sensor 5.

[0018] The display unit 34 of the farmer's terminal 3 displays, for example, a harvest information input screen, countermeasures acquired from the information processing device 1, and the like.

[0019] The input unit 35 of the farmer terminal 3 accepts input of harvest information related to agricultural crops. In this embodiment, the farmer terminal 3 is a smartphone, and the display unit 34 and the input unit 35 are integrally configured as a touch panel.

[0020] The photographing unit 36 of the farmer terminal 3 takes images (photographed images) of the crops or the field. When the farmer terminal 3 is a smartphone, the photographing unit 36 is configured by a camera built into the smartphone. The terminal processing unit 31 of the farmer terminal 3 may transmit the photographed images stored in the memory unit 32 of the farmer terminal 3 to the information processing device 1. The terminal processing unit 31 may also acquire the photographed images from a camera, which is an external device, and transmit the acquired photographed images to the information processing device 1. The photographed images may be those taken by an external camera device or a camera device mounted on a drone. The photographed images may also be satellite images of the field taken by an artificial satellite.

[0021] 4 is an explanatory diagram showing an example of the crop information table 121. The crop information table 121 stores information about farmers and information for identifying tables in which each piece of information included in the crop information (harvest information, tillage information, sensor information, and weather information) is stored. The management items (fields) of the crop information table include, for example, a farmer ID field, a farmer name field, a crop field, a field location field, a harvest information table ID field, a tillage information table ID field, a sensor information table ID field, a weather information table ID field, and a captured image table ID field.

[0022] The farmer ID field of the crop information table 121 stores an ID assigned to a farmer. The farmer name field stores the name of the farmer. The crop field stores the name of the crop produced by the farmer. The field location field stores the location of the field where the farmer produces the crop.

[0023] The harvest information table ID field of the crop information table 121 stores an ID that identifies harvest information table 122 (see FIG. 5) in which the harvest information included in the crop information is stored. The tillage information table ID field stores an ID that identifies tillage information table 123 (see FIG. 6) in which the tillage information included in the crop information is stored. The sensor information table ID field stores an ID that identifies sensor information table 124 (see FIG. 7) in which the sensor information included in the crop information is stored. The weather information table ID field stores an ID that identifies weather information table 125 (see FIG. 8) in which the weather information included in the crop information is stored. The photographed image table ID field stores an ID that identifies a photographed image table in which photographed images included in the crop information are stored.

[0024] 5 is an explanatory diagram showing an example of the harvest information table 122. The harvest information table 122 stores harvest information inputted at the farmer terminal 3, including the yield of the crop and the quality of the harvested crop. The harvest information table 122 also includes the farmer ID, the crop, and the harvest information table ID as attribute information. That is, the storage unit 12 of the information processing device 1 stores multiple harvest information tables 122 for each farmer and each crop. The management items of the harvest information table 122 include, for example, a harvest period field, a yield field, a quality field, and a remarks field.

[0025] The harvest period field of the harvest information table 122 stores the harvest period of the crop. The yield field stores the yield of the crop harvested during the harvest period. The quality field stores an evaluation of the quality of the harvested crop, for example, expressed on a three-level scale of A, B, or C. The quality of the crop is evaluated by the farmer or a third-party organization. The quality field may be subdivided into multiple evaluation items depending on the crop to be harvested. For example, the quality field may include a sugar content evaluation field, an amino acid evaluation field, a color evaluation field, or a trait evaluation field. The remarks field stores remarks about the crop entered by the farmer on the farmer terminal 3 (for example, a comparison with crops harvested last year, or the farmer's impressions, etc.).

[0026] FIG. 6 is an explanatory diagram showing an example of the tillage information table 123. The tillage information table 123 stores information about the operation of the agricultural machine 4, which is acquired by the information processing device 1 from the agricultural machine 4. In this embodiment, the agricultural machine 4 is a tractor capable of spraying fertilizer or pesticides while tilling a field. The agricultural machine 4 may also be a rice planter, combine harvester, sowing machine, transplanter, harvester, or drone. The tillage information table 123 includes, as attribute information, a farmer ID, a crop, and a tillage information table ID. The memory unit 12 of the information processing device 1 stores multiple tillage information tables 123 for each farmer and crop. The management items (fields) of the tillage information table 123 include a tillage implementation date field, a tillage depth field, a spread fertilizer field, a fertilizer application amount field, a spread pesticide field, and a pesticide application amount field.

[0027] The tillage date field of the tillage information table 123 stores the date when the agricultural machine 4 was driven and tilled. The tillage depth field stores the depth to which the agricultural machine 4 tilled the soil. The applied fertilizer field stores the type of fertilizer applied to the field by the agricultural machine 4. The applied fertilizer amount field stores, for example, the mass of fertilizer applied per square meter. The applied pesticide field stores the type of pesticide applied to the field by the agricultural machine 4. The applied pesticide amount field stores, for example, the mass of pesticide applied per square meter. In this embodiment, the processing unit 11 of the information processing device 1 records the tillage information acquired from the agricultural machine 4 in the tillage information table 123, but this is not limited to this. The processing unit 11 may also acquire tillage information from the farmer terminal 3. At this time, the farmer terminal 3 may transmit the tillage information input by the farmer to the information processing device 1, or may transmit the tillage information acquired from the agricultural machine 4 to the information processing device 1. Also, a null value is stored in fields relating to work that the agricultural machine 4 is not performing during tillage. The tillage information may include information about the agricultural machine, including, for example, the model name of the agricultural machine owned by the farmer. At this time, the information about the agricultural machine is stored in the tillage information table 123.

[0028] FIG. 7 is an explanatory diagram showing an example of the sensor information table 124. The sensor information table 124 stores sensor information acquired by the information processing device 1 from the field sensor 5. In this embodiment, the field sensor 5 is composed of multiple sensor devices that measure or detect the pH of the field soil, the water level of the field, the moisture content of the field soil, pests, or epidemics. The sensor information table includes, as attribute information, a farmer ID, a crop, and a sensor information table ID. In other words, the memory unit 12 of the information processing device 1 stores multiple sensor information tables 124 for each farmer and crop. The management items (fields) of the sensor information table 124 include a measurement date field, a soil pH field, a water level field, a moisture content field, a pest outbreak field, a pest outbreak field, an epidemic outbreak field, and an epidemic outbreak field.

[0029] The measurement date field of the sensor information table 124 stores the date on which the sensor information was measured. The soil pH field stores the pH of the soil in the field. The water level field stores the water level if the field is a rice paddy. The moisture content field stores the volume water content (VWC) of the soil in the field. The pest occurrence field stores the type of pest that has occurred in the field. The pest infestation field stores the degree of infestation of pests that has occurred in the field, expressed in three levels, for example, A, B, and C. The disease field stores the type of disease that has occurred in crops in the field. The disease infestation field stores the degree of infestation of disease that has occurred in the field, expressed in three levels, for example, A, B, and C. In this embodiment, the processing unit 11 of the information processing device 1 records the tillage information acquired from the field sensor 5 in the sensor information table 124, but this is not limited to this. The processing unit 11 may acquire sensor information from the farmer's terminal 3. At this time, the farmer's terminal 3 may transmit the sensor information input by the farmer to the information processing device 1, or may transmit tillage information acquired from the field sensor 5 to the information processing device 1. Furthermore, a null value is stored in a field relating to an item that the field sensor 5 has not measured or detected.

[0030] 8 is an explanatory diagram showing an example of the weather information table 125. The weather information table 125 stores weather information for the field where the crops are produced (weather information for the area where the field is located), which the processing unit 11 of the information processing device 1 acquires from the weather server 6. The weather information table 125 also includes, as attribute information, a farmer ID, a crop, and a weather information table ID. That is, the storage unit 12 of the information processing device 1 stores multiple weather information tables 125 for each farmer and crop. The management items of the weather information table 125 include a date field, a maximum temperature field, a minimum temperature field, a humidity field, a precipitation field, and a sunshine duration field.

[0031] The date field of the weather information table 125 stores the date on which the weather information was observed. The maximum temperature field stores the maximum temperature in the field on the observation date. The maximum temperature field stores the minimum temperature in the field on the observation date. The humidity field stores the average humidity on the observation date. The precipitation field stores the amount of precipitation on the observation date. The sunshine hours field stores the sunshine hours on the observation date. In this embodiment, when the processing unit 11 of the information processing device 1 acquires harvest information from the farmer's terminal 3, it acquires weather information for each day of the harvest period included in the harvest information from the weather information server and stores it in the weather information table 125, but this is not limited to this. The processing unit 11 may, for example, acquire the weather information for the previous day from the weather server 6 at a predetermined time every day and store it in the weather information table 125.

[0032] FIG. 9 is an explanatory diagram showing an example of the photographed image table 126. The photographed image table stores photographed images relating to agricultural crops or farm fields. The photographed image table 126 includes, as attribute information, a farmer ID, a crop, and a photographed image table ID. That is, the storage unit 12 of the information processing device 1 stores a plurality of photographed image tables 126 for each farmer and each crop. The management items (fields) of the photographed image table 126 include a photography date field, a photographed object field, and a photographed image field.

[0033] The date on which the image was taken is stored in the image date field of the image table 126. The photographed object field stores the part of the crop that was the subject of the photograph (harvested product, the entire crop, leaves, flowers, fruit, etc.) or the part of the field (the entire field, soil, etc.). The photographed image field stores the photographed image acquired by the information processing device 1 from the farmer's terminal 3, for example, in file format.

[0034] FIG. 10 is an explanatory diagram showing an example of a crop yield prediction model PM. The crop yield prediction model PM is generated by machine learning using a neural network, for example, a recurrent neural network (RNN) that outputs values based on time-series data. Note that the machine learning may be performed using a neural network such as a convolutional neural network (CNN) or a deep neural network (DNN). Furthermore, the machine learning may be performed using a method other than a neural network. For example, various machine learning methods such as a long short-term memory (LSTM), a transformer, a support vector machine (SVM), or a k-nearest neighbor method may be employed.

[0035] When the yield prediction model PM is configured using a neural network such as an RNN, the input layer of the yield prediction model PM has multiple neurons that accept input of time-series data on harvest information, tillage information, sensor information, and weather information included in the crop information, and passes the input crop information to the middle layer. The middle layer has multiple neurons that extract features of the crop information and passes the extracted features to the output layer. The output layer has neurons that output a predicted yield, and outputs the predicted yield based on the features output from the middle layer. Note that the yield prediction model PM may not only use time-series data on crop information, but also use information related to the latest records of harvest information, tillage information, sensor information, or weather information. The crop information input to the yield prediction model PM may also include photographed images.

[0036] For example, if the harvest period of the harvest information input to the harvest forecasting model PM is recorded by month, the predicted harvest output by the harvest forecasting model PM is the predicted harvest for the month following the harvest period of the input harvest information (scheduled harvest time). Note that the harvest forecasting model PM may output predicted harvests for each month up to one year in the future. Furthermore, if the harvest period of the harvest information input to the harvest forecasting model PM is recorded by year, the harvest forecasting model PM may output the predicted harvest for the following year.

[0037] FIG. 11 is an explanatory diagram showing an example of a language model LM. The language model LM is configured, for example, by a Generative Pretrained Transformer (GPT). When a prompt including crop information and a query is input, the language model LM outputs a countermeasure for the crop information based on the input prompt. The query includes an instruction to output a countermeasure for the crop information. When the predicted yield output (calculated) by the yield prediction model PM is lower than a predetermined value, the processing unit 11 of the information processing device 1 transmits the crop information and the query to the language model server and inputs them into the language model LM.

[0038] The prompt crop information input to the language model LM includes crop information including harvest information, tillage information, sensor information, weather information, and captured images that the processing unit 11 of the information processing device 1 has acquired within a predetermined period (e.g., the past year) and stored in the memory unit 12. The crop information input to the language model LM may include at least one of harvest information, tillage information, sensor information, weather information, and captured images. The language model LM may also be input with features extracted from captured images using a learning model such as CNN or Vision Transformer, a predicted crop yield calculated based on the features, or text data obtained by inputting captured images into an Image to Text model.

[0039] The prompt query input to the language model LM includes, for example, the type of agricultural product produced by the farmer and an instruction to propose countermeasures to ensure that the yield of the agricultural product is equal to or greater than a predetermined value. The storage unit 12 of the information processing device 1 stores a query template that serves as the basis for the query input to the language model LM. The query template is expressed, for example, as a sentence such as, "You are a farmer who produces X. Please tell me what countermeasures you can take to ensure that the yield of X is equal to or greater than Y tons." The processing unit 11 of the information processing device 1 generates a query by substituting the type of agricultural product produced by the farmer for X in the query template and substituting a predetermined value, which is a threshold value for the expected yield, for Y as a target value for the yield, which the processing unit 11 uses to input the agricultural product information and the query into the language model LM. In the example shown in FIG. 11 , X (agricultural product) is broccoli, and Y (target yield) is 1 (ton). Note that the value substituted for Y may be a value greater than the predetermined threshold value for the expected yield, or may be a target value for the yield per unit field area (e.g., 1 hectare). The query input to the language model LM may be input each time at the farmer's terminal 3. Furthermore, the query may instruct the output of measures to improve the quality of the agricultural products, including the size, color, shape, taste, moisture content, or amount of pesticide used.

[0040] When agricultural product information and a query (prompt) are input, the language model LM outputs countermeasures to increase the yield of the agricultural product specified in the query above a target value. For example, the language model LM outputs countermeasures such as adjusting the water level if the field is a rice paddy, the amount of water to be sprayed if the field is a field, the type of fertilizer to be added to the soil, the amount of fertilizer to be added to the soil, the type of pesticide to be sprayed, the amount of pesticide to be sprayed, or precautions to take when producing agricultural products. In the example shown in FIG. 11, the language model LM outputs the countermeasures such as "adjust the water level at 3 L / m" 2 Spread "Fertilizer C at 0.7 kg / m 2 Please spray." is displayed as a countermeasure.

[0041] 12 is an explanatory diagram showing an example of the harvest information input and photographed image screen. The harvest information input and photographed image screen includes a farmer ID input field, a crop input field, a harvest period input field, a harvest volume input field, a quality input field, a remarks input field, an object to be photographed input field, and a photographed image input field.

[0042] The farmer's farm ID is entered in the farmer ID input field on the harvest information and captured image input screen. The type of crop produced by the farmer is stored in the crop input field. The harvest period input field is entered with the period during which the farmer harvested the crop. The harvest yield input field is entered with the yield of the crop harvested by the farmer during the harvest period. The quality input field is entered with the quality of the crop as evaluated by the farmer or a third-party organization. The remarks input field is used to enter remarks regarding the crop (for example, a comparison with crops harvested last year, or the farmer's impressions, etc.). The harvest period, harvest yield, quality, and remarks entered on the harvest information input and captured image screen are sent to the information processing device 1 and stored in a harvest information table 122 that includes the entered farmer ID and crop as attribute information.

[0043] The part of the crop or the part of the field that was the subject of the photograph is input in the photographed object input field on the harvest information input and photographed image input screen. The photographed image input screen is input with the photographed image taken by the photographing unit 36 of the farmer terminal 3 or the photographed image stored in the memory unit 32. The photographed object and photographed image input on the harvest information input and photographed image screen are transmitted to the information processing device 1 and stored in the photographed image table 126, which includes the input farmer ID and crop as attribute information.

[0044] 13 is an explanatory diagram showing an example of a countermeasure output screen. The processing unit 11 of the information processing device 1 outputs the countermeasure output by the language model LM from the language model server 2 to the farmer terminal 3. The terminal processing unit 31 of the farmer terminal 3 displays the countermeasure acquired from the information processing device 1 on the display unit 34. The processing unit 11 of the information processing device 1 also outputs the predicted yield output by the yield prediction model PM and the target value of the yield included in the query input to the language model LM to the farmer terminal 3. The terminal processing unit 31 of the farmer terminal 3 displays the predicted yield and target value of the yield acquired from the information processing device 1 on the display unit 34. The processing unit 11 of the information processing device 1 may transmit the crop information input to the language model LM to the farmer terminal 3, and the terminal processing unit 31 of the farmer terminal 3 may display the crop information acquired from the information processing device 1 on the display unit 34. Furthermore, the processing unit 11 of the information processing device 1 may output and display the countermeasures output by the language model LM on a display device such as a personal computer, a VT terminal, a dashboard monitor or pillar of the agricultural machine 4, or a tablet for automatic operation of the agricultural machine 4. Furthermore, the processing unit 11 of the information processing device 1 may output only the predicted yield to the farmer terminal 3 when the predicted yield output by the yield prediction model PM is equal to or greater than the target yield value.

[0045] 14 is a flowchart showing an example of a process for registering agricultural crop information. The processing unit 11 of the information processing device 1 acquires harvest information and captured images from the farmer's terminal 3 (S1). The processing unit 11 stores the acquired harvest information in the harvest information table 122 (S2) and stores the captured images in the captured image table 126 (S3). The processing unit 11 acquires tillage information from the agricultural machine 4 (S4). The processing unit 11 stores the acquired tillage information in the tillage information table 123 (S5). The processing unit 11 acquires sensor information from the field sensor 5 (S6). The processing unit 11 stores the acquired sensor information in the sensor information table 124 (S7). The processing unit 11 acquires weather information from the weather server 6 (S8). The processing unit 11 stores the acquired weather information in the weather information table 125 (S9), and ends the process.

[0046] 15 is a flowchart showing an example of a process for outputting countermeasures. The processing unit 11 of the information processing device 1, for example, executes a process for registering crop information and then executes a process for outputting countermeasures. The processing unit 11 reads time-series crop information from the harvest information table 122, the tillage information table 123, the sensor information table 124, and the weather information table 125 (S11). The processing unit 11 inputs the crop information into the yield prediction model PM (S12) and outputs the predicted yield (S13). The processing unit 11 determines whether the predicted yield is equal to or greater than a predetermined value (S14). If the predicted yield is equal to or greater than the predetermined value (S14: YES), the processing unit 11 transmits (outputs) the predicted yield to the farmer's terminal 3 (S15), and ends the process.

[0047] If the expected harvest is lower than the predetermined value (S14: NO), the processing unit 11 of the information processing device 1 reads out crop information from the harvest information table 122, the tillage information table 123, the sensor information table 124, the weather information table 125, and the captured image table 126 (S16), and inputs the read out crop information and query as a prompt into the language model LM (sends it to the language model server 2) (S17). The processing unit 11 acquires the countermeasures output by the language model LM from the language model server 2 (S18). The processing unit 11 sends the expected harvest, the target harvest value, and the countermeasures to the farmer terminal 3 (S19), and ends the processing.

[0048] According to the above configuration and processing, the processing unit 11 of the information processing device 1 can input agricultural product information into the language model LM, thereby outputting flexible countermeasures according to the type of agricultural product and the agricultural product information. The agricultural product information may include a work schedule for the field or the agricultural product and the progress of the actual work. Furthermore, the storage unit 12 of the information processing device 1 may store a planned harvest volume for each predetermined period inputted, for example, to the farmer's terminal 3, and use the planned harvest volume as a predetermined value for outputting countermeasures.

[0049] (Embodiment 2) The storage unit 12 of the information processing device 1 according to the second embodiment stores a countermeasure database in which countermeasures for each piece of agricultural product information are stored in advance. The processing unit 11 of the information processing device 1 extracts agricultural product information similar to the acquired agricultural product information and countermeasures for the agricultural product information, and causes the language model LM to output the countermeasures using a prompt including the agricultural product information and the countermeasures extracted from the countermeasure database.

[0050] FIG. 16 is an explanatory diagram showing an example of the countermeasure database 127. The countermeasure database 127 is stored in the storage unit 12 of the information processing device 1 (see FIG. 2). The countermeasure database records countermeasures for agricultural product information when the predicted yield is lower than a predetermined value. The management items (fields) of the countermeasure database 127 include a condition field and a countermeasure field.

[0051] The condition field stores a condition for extracting the countermeasure of the corresponding record so that it is included in the prompt to be input to the language model LM. The countermeasure field stores a countermeasure to be included in the prompt to be input to the language model LM when the agricultural product information acquired by the processing unit 11 of the information processing device 1 meets the condition of the corresponding record.

[0052] FIG. 17 is an explanatory diagram showing an example of a language model LM according to the second embodiment. When the predicted yield output by the yield prediction model PM is lower than a predetermined value, the processing unit 11 of the information processing device 1 according to the second embodiment determines whether the crop information read from the harvest information table 122, the tillage information table 123, the sensor information table 124, and the weather information table 125 satisfies each condition stored in the condition field of the countermeasure database 127. When the crop information satisfies any of the conditions, the processing unit 11 reads a countermeasure related to the corresponding condition from the countermeasure database 127 and inputs the corresponding condition and countermeasure into the language model LM. The language model LM references the input condition and countermeasure and outputs a countermeasure for the input crop information. Note that the countermeasure output by the language model LM is not necessarily the same as the input countermeasure. Based on information on multiple items included in the input crop information, the processing unit 11 outputs a countermeasure that is more appropriate than the input countermeasure according to the condition.

[0053] Fig. 18 is a flowchart showing an example of a process for outputting countermeasures according to the second embodiment. The processes in S21 to S26 are the same as the processes in S11 to S16 shown in Fig. 15. The processing unit 11 of the information processing device 1 reads out countermeasures for the conditions that the read-out crop information falls under from the countermeasure database 127 (S27). The processing unit 11 inputs the read-out crop information, the query, the conditions that the crop information falls under, and the countermeasures for the conditions as prompts to the language model LM (sends them to the language model server 2) (S28). The processing unit 11 acquires the countermeasures output by the language model LM from the language model server 2 (S29). The processing unit 11 sends the expected yield, the target value of the yield, and the countermeasures to the farmer terminal 3 (S30), and ends the process.

[0054] (Variation) The processing unit 11 of the information processing device 1 may input training data including the conditions and countermeasures stored in the countermeasure database 127 into the language model LM in advance, and cause the language model LM to learn so as to output a countermeasure when agricultural product information is input. FIG. 19 is a flowchart showing an example of a countermeasure output process according to a modification of the second embodiment. The processing unit 11 of the information processing device 1 reads out conditions and countermeasures for the conditions from the countermeasure database 127 (S41), and inputs training data including the read conditions and countermeasures into the language model LM (S42). The processes of S43 to S51 are similar to the processes of S11 to S19 shown in FIG. 15.

[0055] According to the above configuration and processing, the processing unit 11 of the information processing device 1 can output more appropriate countermeasures using the language model LM based on the conditions of the agricultural product information and the countermeasures for those conditions. Note that the processing unit 11 may collect external knowledge data, such as agricultural machinery manuals or agricultural product guidelines, from an external data server using a Retrieval Augmented Generation (RAG) function and input the external knowledge data to the language model LM as prompts or training data. In this case, the processing unit 11 may input a link (URL) of a web page including the external knowledge data to the language model LM.

[0056] (Embodiment 3) The processing unit 11 of the information processing device 1 according to the third embodiment includes in the query input to the language model LM a request for work on the crops or the field, or a countermeasure related to the control amount of the agricultural machine 4. When the language model LM receives input of crop information and a query (prompt), it outputs work on the crops or the field, or a countermeasure related to the control amount of the agricultural machine 4. Furthermore, when the output of the language model LM includes a countermeasure related to the control amount of the agricultural machine 4, it outputs a drive command corresponding to the control amount to the agricultural machine 4.

[0057] FIG. 20 is an explanatory diagram showing an example of a language model LM according to the third embodiment. A query input to the language model LM according to the third embodiment includes a request for countermeasures regarding work on crops or a field, or the control amount of the agricultural machine 4 (the sentence shown in the query in FIG. 20 is, "Please tell me what work should be done on the crops or the field, or the control amount when driving the agricultural machine."). When this query is input, the language model LM outputs countermeasures regarding work on crops or a field, or the control amount of the agricultural machine 4. Countermeasures regarding work on crops include, for example, the thinning rate or artificial pollination. Countermeasures regarding work on a field include, for example, laying nets or sheets, adjusting the water level, or the amount of water to be sprayed on the field. Countermeasures regarding the control amount of the agricultural machine 4 include, for example, the tilling depth, the type and amount of fertilizer to be applied, or the type and amount of pesticide to be applied. If the agricultural machine 4 is a water gate in a rice field, the countermeasures related to the control amount of the agricultural machine 4 may include the opening degree and opening time of the water gate. If the agricultural machine 4 is a sprinkler, the countermeasures related to the control amount of the agricultural machine 4 may include the amount of water to be sprayed. In the example shown in FIG. 20, the countermeasures related to the control amount indicating the amount of water to be sprayed by the agricultural machine 4 (sprinkler) ("Water 3 L / m to the agricultural machine (sprinkler)") 2 ") and measures for controlling the amount of fertilizer to be spread by agricultural machine 4 (tractor) ("Fertilizer C is to be spread at 0.7 kg / m by agricultural machine (tractor)." 2 The language model LM may output a program code for driving the agricultural machine 4 in accordance with the control amount.

[0058] FIG. 21 is an explanatory diagram showing an example of a countermeasure display screen according to the third embodiment. When the language model LM outputs countermeasures related to the control variables of the agricultural machinery 4, the countermeasure display screen according to the third embodiment displays an approval button for receiving input of approval information approving the implementation of each output countermeasure. When the terminal processing unit 31 of the farmer terminal 3 receives a press of the approval button, it transmits to the information processing device 1 a message indicating that approval information for the output countermeasures has been input. The processing unit 11 of the information processing device 1 that has acquired the approval information outputs a drive command corresponding to the control variable of the agricultural machinery 4 output by the language model LM to the agricultural machinery 4, thereby driving the agricultural machinery 4. The drive command is a command to drive the agricultural machinery 4 according to the control variable of the agricultural machinery 4 output by the language model LM.

[0059] The terminal processing unit 31 of the farmer terminal 3 may accept input of correction of the control amount for each countermeasure on the countermeasure display screen. For example, in the countermeasure shown in FIG. 20, 2 " or "0.7kg / m 2 " may be input to change the numerical value of ". The processing unit 11 may also accept input of countermeasures other than the output countermeasures. In this case, the processing unit 11 of the information processing device 1 outputs to the agricultural machine 4 a drive command corresponding to the control amount of the agricultural machine 4 included in the countermeasures input at the farmer's terminal 3, and drives the agricultural machine 4. Furthermore, the terminal processing unit 31 may respond to inquiries about countermeasures (for example, "I want to spray pesticide G, but how many L / m 2 In this case, the processing unit 11 of the information processing device 1 may input the inquiry input at the farmer terminal 3 into the language model LM and transmit the output countermeasure to the farmer terminal 3.

[0060] FIG. 22 is a flowchart illustrating an example of a process for outputting a countermeasure according to the third embodiment. The processing of S61 to S69 is the same as the processing of S11 to S19 shown in Fig. 15. The processing unit 11 of the information processing device 1 acquires the approval information input at the farmer's terminal 3 (S70). The processing unit 11 outputs to the agricultural machine 4 a drive command corresponding to the control amount of the agricultural machine 4 output by the language model LM (S71), and ends the processing.

[0061] According to the above configuration and processing, the farmer can operate the agricultural machine 4 to increase the yield of agricultural crops by approving the countermeasure output by the language model LM.

[0062] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The technical features described in each embodiment may be combined with one another, and the scope of the present invention is intended to include all modifications within the scope of the claims and equivalents thereto. Furthermore, independent and dependent claims described in the claims may be combined with one another in any and all combinations, regardless of the reference format. Furthermore, while the claims use a format in which a claim references two or more other claims (multiple claim format), this is not limiting. Multiple claims (multiple multiple claims) that reference at least one other multiple claim may also be used. [Explanation of symbols]

[0063] 1: Information processing equipment 2: Language model server 3: Farmer terminal 4: Agricultural machinery 5: Field sensor 6: Weather server 11: Processing section 12: Storage section 12a:Storage medium 13: Communications Department 24: Display section 26: Photography Department 31: Terminal processing section 32: Storage section 32a:Storage medium 33: Communications Department 34: Display section 35: Input section 36: Photography Department 121: Crop information table 122: Harvest information table 123: Tillage information table 124: Sensor information table 125: Weather information table 126: Photographed image table 127: Countermeasures database LM: Language Model N: Network P: Program PM: Yield prediction model Pa: Application program Pa: Application program S: Information Processing System

Claims

1. Obtaining crop information about agricultural crops; A language model using the crop information is used to output a countermeasure for the acquired crop information. Information processing methods.

2. Get the query that specifies what to output, Using the acquired query and the crop information, a countermeasure is output using the language model. The information processing method according to claim 1 .

3. Acquire the crop information, which includes harvest information on the harvested crops, tillage information on the agricultural machinery, or sensor information acquired from a sensor installed in the field; Using the crop information including the harvest information, the tillage information, or the sensor information, outputting countermeasures using the language model.