System

Abstract

To provide a system for reflecting road conditions of the whole country in real time and directly and effectively issuing a warning to a driver according to the road conditions.SOLUTION: The system includes means for collecting driving data of a driver, means for acquiring nationwide road condition data, means for analyzing these data and evaluating a risk level of a traffic accident, and means for issuing a warning to the driver when the risk level is increased, and further includes means for analyzing a driving behavior pattern of the driver based on the driving data and means for performing individual risk evaluation according to the driving behavior pattern.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] Traffic accidents cause significant damage throughout the year, making their prevention a socially important issue. While existing traffic safety systems have achieved a certain degree of effectiveness, they lack sufficient risk assessment and preventative measures based on individual driver characteristics, making it difficult to completely prevent accidents. Furthermore, there is a lack of systems that reflect nationwide road conditions in real time and issue direct and effective warnings to drivers accordingly. This presents a problem in that accidents caused by specific road conditions or driving behaviors cannot be prevented before they occur. [Means for solving the problem]

[0005] The present invention provides a system including a first means for collecting driver driving data, a second means for acquiring nationwide road condition data, a third means for analyzing the data and assessing the risk level of traffic accidents, and a fourth means for issuing a warning to the driver if the risk level increases. Furthermore, the system includes a fifth means for analyzing the driver's driving behavior patterns based on the driving data, a sixth means for conducting individual risk assessments based on the driving behavior patterns, a seventh means for collecting data in real time and notifying the driver of the analysis results, and an eighth means for receiving feedback from the driver and updating the data analysis model based on that data. This makes it possible to accurately assess the risk of each driver and issue a warning at an appropriate time. As a result, traffic accidents can be prevented and safe driving can be promoted.

[0006] "Driving data" refers to data that indicates the driving conditions of the vehicle, and specifically includes information such as speed, number of sudden brakings, acceleration, distance traveled, and driving time.

[0007] "Road condition data" refers to data relating to real-time traffic conditions on specific roads or in specific areas, and specifically includes information such as traffic volume, congestion, road conditions, weather, and accident history.

[0008] "Risk assessment" is the process of quantitatively calculating the possibility of a traffic accident occurring by analyzing driving data and road condition data.

[0009] A "warning" is an informational message to alert the driver based on the results of a risk assessment, and may be provided in the form of an audio alert, visual alert, haptic feedback, or other similar message.

[0010] A "generative AI model" is an artificial intelligence model that learns from large amounts of driving data and road condition data and has an algorithm for predicting the risk of traffic accidents.

[0011] "Feedback" refers to responsive information provided by the driver regarding driving behavior and the effectiveness of warnings, which is used to improve the system's accuracy and retrain the model.

[0012] "Real-time" refers to a process in which data is collected, analyzed, and communicated immediately, without delay. [Brief explanation of the drawings]

[0013] [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

[0014] 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.

[0015] First, the terms used in the following description will be explained.

[0016] 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).

[0017] 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.

[0018] 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.

[0019] 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.

[0020] 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."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 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.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] 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.

[0028] 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.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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."

[0034] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. Below, we will explain in detail the program processing of this system in natural language.

[0035] Data collection and analysis program

[0036] The server first periodically acquires nationwide road condition data. This data includes congestion levels, weather, accident history, and other information provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, and other organizations. Meanwhile, the device uses on-board sensors and a smartphone app to collect driver data. This data includes speed, frequency of sudden braking, driving time, and other information. The collected data is then sent to the server in real time.

[0037] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0038] Warning program

[0039] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are several ways to issue the warning, mainly audio alerts, visual alerts, haptic feedback, or a combination of these. For example, a voice prompt saying, "There is a high-accident zone ahead. Please be careful," will be played, and a warning message will be displayed on the dashboard.

[0040] Feedback and continuous improvement of the system

[0041] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0042] Specific examples

[0043] 1. Data Collection Examples

[0044] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0045] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0046] 2. Data analysis examples

[0047] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0048] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0049] 3. Examples of warnings

[0050] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0051] At the same time, a warning light on the dashboard will flash to alert you.

[0052] In this way, this invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about the risk of traffic accidents.

[0053] The processing flow will be explained below.

[0054] Step 1: Data collection

[0055] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0056] The device collects the driver's driving data (speed, frequency of sudden braking, driving time, etc.) through on-board sensors and a smartphone app and transmits it to a server in real time.

[0057] Step 2: Data Preprocessing

[0058] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, complements missing values, detects and removes outliers, and organizes the data appropriately as time series data.

[0059] Step 3: Data analysis

[0060] The server inputs the preprocessed data into a generative AI model. The generative AI model learns from a large amount of past data and integrates and analyzes road condition data and driving data. Specifically, it evaluates the current driving situation based on factors such as traffic volume, weather conditions, and past accident history, and calculates the risk level of a traffic accident.

[0061] Step 4: Risk assessment

[0062] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model, which are then classified into categories such as "low risk," "medium risk," and "high risk."

[0063] Step 5: Send an alert

[0064] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is given by audio alert, visual alert, haptic feedback, or a combination of these. For example, a voice prompt saying "There is a high-accident zone ahead. Please be careful" is played, and a warning message is displayed on the dashboard at the same time.

[0065] Step 6: Adjust your driving behavior

[0066] The user receives a warning and is encouraged to review their driving behavior and take steps to drive safely, such as slowing down and paying attention to their surroundings.

[0067] Step 7: Gather feedback

[0068] If the user receives multiple warnings, they provide feedback on the effectiveness and areas for improvement. This feedback includes information on changes in driving behavior and the effectiveness of the system. The feedback data is sent to the server via the device.

[0069] Step 8: Update the Model

[0070] The server analyzes the received feedback data and uses it to retrain the generative AI model, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0071] Example 1

[0072] 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."

[0073] In order to reduce the risk of traffic accidents, it is essential to provide drivers with accurate warnings in real time. However, current systems have difficulty effectively integrating data on road conditions and driver behavior nationwide to make highly accurate risk predictions. To address this issue, there is a need for a system that can accurately predict the risk of traffic accidents and issue warnings to drivers in real time.

[0074] 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.

[0075] In this invention, the server includes a means for acquiring nationwide road condition data, a means for collecting driver driving data, a means for preprocessing the acquired and collected data, a means including a generative AI model for analyzing the preprocessed data to evaluate the risk level of a traffic accident, and a means for issuing a warning to the driver in real time when the risk level increases. This enables highly accurate risk prediction based on individual driving characteristics and nationwide road conditions, and allows drivers to receive real-time warnings to prevent the risk of a traffic accident.

[0076] "National road condition data" refers to data including road congestion, weather conditions, and past accident history obtained from data providers such as the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[0077] "Driver driving data" refers to data related to driving behavior, such as vehicle speed, number of sudden braking attempts, driving time, and location information, collected using in-vehicle sensors and smartphone apps.

[0078] "Preprocessing" refers to the process of filling in missing values ​​in collected data, removing outliers, and converting the data into a format suitable for analysis.

[0079] A "generative AI model" is an artificial intelligence model built on a large dataset, and is a machine learning or deep learning model used to assess the risk of traffic accidents.

[0080] A "warning" is a real-time notification to the driver when the risk of a traffic accident increases, and may be provided via audio alerts, visual alerts, and / or haptic feedback.

[0081] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. This system is mainly composed of a server, a terminal, and a user, and specific embodiments are described below.

[0082] Data collection methods

[0083] The server periodically obtains nationwide road condition data provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. To obtain this data, it receives data in JSON format from multiple API endpoints.

[0084] Meanwhile, the device collects driving data from the driver using on-board sensors and a smartphone app. Collected data includes the vehicle's speed, number of sudden braking attempts, driving time, and location information. This data is sent to a server in real time. The smartphone app on the device periodically collects data from the sensors in the background and transfers it to the server every few seconds.

[0085] Data analysis format

[0086] The server preprocesses the received data. This preprocessing involves filling in missing values, removing outliers, and converting the data into a format suitable for analysis. Specifically, data cleaning is performed using a Python program and the Pandas library. Missing values ​​are filled in using the average of the preceding and following data, and abnormal values ​​(e.g., speed data that suddenly reaches several hundred km / h) are removed as outliers.

[0087] The preprocessed data is then fed into a generative AI model to assess the risk level of a traffic accident. This generative AI model is a deep learning model built using TensorFlow or PyTorch and trained on a large dataset. The model's output includes a risk score and specific warnings.

[0088] Warning issuance format

[0089] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. This warning can be provided by audio alerts, visual alerts, and / or haptic feedback. For example, a voice prompt such as "There is a high-accident zone ahead. Please be careful" may be played from the smartphone, while a warning message is simultaneously displayed on the in-car display.

[0090] The device uses push notifications or Bluetooth connectivity to display information on the in-car display or smartphone, allowing the driver to quickly recognize important information on the spot.

[0091] Forms of feedback and system improvement

[0092] When users receive a warning from the system, they can review their driving behavior and strive for safer driving. If they receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback is given in the form of a survey via a smartphone app. For example, questions such as "Was the warning helpful?" and "Has it changed your driving behavior?" are displayed.

[0093] The feedback data is sent to a server and used to retrain the generative AI model. The server retrains the model based on new data, improving the accuracy of the system. This process is repeated periodically to ensure continuous improvement of the system.

[0094] Examples of specific examples and prompts

[0095] 1. Data Collection Example

[0096] The server obtains real-time road congestion data for Tokyo at 5 p.m. from the traffic management center.

[0097] The device collects driving data from the driver's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to a server.

[0098] 2. Data Analysis Example

[0099] The server checks that a particular intersection in Tokyo has had many accidents over the past three years.

[0100] The analysis results show that the risk level is high because the particular driver frequently brakes suddenly during evening traffic jams.

[0101] 3. Examples of warnings

[0102] Just before the user (driver) approaches a specific intersection, the device issues a voice alert saying, "There is a high-accident area ahead. Please slow down and be careful."

[0103] At the same time, a warning light on the dashboard will flash to alert you.

[0104] Example prompt sentence:

[0105] "Please explain how you collect and analyze driver behavior data and nationwide road condition data to predict the risk of traffic accidents, and how you issue warnings in real time."

[0106] The above is a specific embodiment for carrying out the present invention. This system enables drivers to accurately predict risks based on their individual driving characteristics and nationwide road conditions, and provides real-time warnings to prevent the risk of traffic accidents.

[0107] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0108] Step 1:

[0109] The server periodically obtains road condition data from across the country. As input, it receives data in JSON format from API endpoints provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. As output, the received data is stored in the server's database.

[0110] Specifically, the server periodically sends an API request like this:

[0111] GET / road_conditions?region=tokyo&datetime=2023-10-12T17:00:00

[0112] GET / weather?region=tokyo&datetime=2023-10-12T17:00:00

[0113] GET / accident_history?region=tokyo&datetime=2023-10-12

[0114] This allows the latest road condition data to be acquired.

[0115] Step 2:

[0116] The device collects the driver's driving data using on-board sensors and a smartphone app and transmits it to a server in real time. The inputs include speed, number of sudden braking attempts, driving time, location information, etc. As an output, the collected data is transferred to the server and added to a database.

[0117] Specifically, the device app periodically sends sensor data to the server in the background:

[0118] POST / upload_driving_data

[0119] Content-Type: application / json

[0120] {

[0121] "speed": 60,

[0122] "brake_count": 3,

[0123] "driving_time": "30 minutes",

[0124] "location": "35.6895, 139.6917"

[0125] }

[0126] Step 3:

[0127] The server preprocesses the received data. The input is the road condition data and driving data collected in Step 1 and Step 2. This data may contain missing values ​​and outliers, so these are imputed and removed. The output is the preprocessed, clean data.

[0128] Specifically, the server runs a Python script that uses the Pandas library to clean the data:

[0129] python

[0130] import pandas as pd

[0131] Data loading

[0132] data = pd.read_json("driving_data.json")

[0133] Missing value imputation

[0134] data['speed'].fillna(data['speed'].mean(), inplace=True)

[0135] Outlier removal

[0136] data = data[data['speed'] < 150]

[0137] Step 4:

[0138] The server inputs the preprocessed data into the generative AI model to perform risk assessment. The input includes preprocessed road condition data and driving data. The generative AI model integrates and analyzes this data to assess the risk level of traffic accidents. The output is a risk score and specific warning content.

[0139] Specifically, the server runs models built using TensorFlow or PyTorch:

[0140] python

[0141] import tensorflow as tf

[0142] Model Loading

[0143] model = tf.keras.models.load_model('risk_model.h5')

[0144] prediction

[0145] predictions = model.predict(preprocessed_data)

[0146] risk_score = predictions[0]

[0147] Step 5:

[0148] As a result of the risk assessment, the device issues a warning to the driver. The inputs include a risk score and warning content obtained from the generative AI model. The output is real-time notification to the driver in the form of audio alerts, visual alerts, and haptic feedback.

[0149] Specifically, the device will display a warning message on the car's display via Bluetooth:

[0150] python

[0151] Audio alerts

[0152] play_sound("warning_audio.mp3")

[0153] Visual Alerts

[0154] show_message("Accident-prone area ahead. Please slow down and take care.")

[0155] Haptic Feedback

[0156] vibrate_device()

[0157] Step 6:

[0158] Users receive warnings from the system and are encouraged to review their driving behavior and drive safely. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. Feedback data in the form of a questionnaire is included as input. The feedback data is sent to the server as output and used to improve the system.

[0159] Specifically, users answer a survey via a smartphone app:

[0160] POST / feedback

[0161] Content-Type: application / json

[0162] {

[0163] "useful": true,

[0164] "behavior_change": "Yes"

[0165] }

[0166] Through these processing steps, the system can use driver behavior data and nationwide road condition data to predict the risk of traffic accidents and issue warnings in real time.

[0167] (Application example 1)

[0168] 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."

[0169] There is a challenge to promote safe driving and prevent traffic accidents by providing a system that predicts the risk of traffic accidents in real time and issues effective warnings to drivers and autonomous vehicles. In addition, with conventional systems, it is difficult for autonomous vehicles to properly assess risks and automatically take appropriate control, so we would like to improve this aspect as well.

[0170] 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.

[0171] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for analyzing this data and assessing the risk level of traffic accidents, means for issuing a warning to the driver when the risk level increases, and means for intervening in the automatic control system of the autonomous vehicle, thereby enabling the driver and the autonomous driving system to receive risk information in real time and take appropriate measures.

[0172] "Driving data" refers to information relating to the driver's driving behavior, including data such as speed, frequency of sudden braking, and driving time.

[0173] "Road condition data" is information about roads across the country, including congestion levels, weather, accident history, and so on.

[0174] "Analysis" refers to the process of processing and analyzing collected data to assess the risk of traffic accidents.

[0175] The "risk level" is an evaluation index that indicates the possibility of a traffic accident occurring.

[0176] "Warnings" are messages or signals that alert the driver when the risk level increases, and can take the form of audio alerts, visual alerts, or haptic feedback.

[0177] An "automatic control system" is a system that automatically controls the driving operations of an autonomous vehicle. It receives information on traffic risks and adjusts vehicle speed and changes course.

[0178] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and conduct risk assessments.

[0179] A "prompt sentence" is an instruction sentence used as input to a generative AI model.

[0180] This invention is a system that collects and analyzes driver's driving behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers and self-driving vehicles in real time. This system includes the following means.

[0181] Data collection and analysis

[0182] The server periodically collects nationwide road condition data. This data includes congestion levels, weather, and accident history provided by the Traffic Management Center and the Japan Meteorological Agency. The device also collects driver driving data using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, and more. The collected data is sent to the server in real time.

[0183] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0184] Warning

[0185] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver or the autonomous vehicle's control system. The warning can be issued in multiple ways, primarily via audio alerts, visual alerts, or haptic feedback, or a combination of these. For example, a voice prompt saying, "Be careful. Accident hotspot ahead," may be played, along with a warning message displayed on the dashboard. The device may also instruct the autonomous vehicle to adjust speed or change course based on pre-set prompts.

[0186] Feedback and continuous improvement of the system

[0187] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0188] Specific examples

[0189] Data Collection Examples

[0190] The server obtains real-time data on road congestion in urban areas at 5 p.m. on a given day from the traffic management center.

[0191] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0192] Data analysis examples

[0193] The server checks that a particular intersection has had a high number of accidents over the past three years.

[0194] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0195] Examples of warnings issued

[0196] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0197] At the same time, a warning light on the dashboard will flash to alert you.

[0198] In the case of self-driving vehicles, the terminal sends instructions to the automatic control system to adjust speed and change course.

[0199] Example prompts to input to the generative AI model

[0200] Enter your driving and road condition data to assess your risk of traffic accidents.

[0201] Driving data: speed, frequency of hard braking, driving time

[0202] Road condition data: congestion, weather, accident history

[0203] This invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about traffic accident risks. This enables not only drivers but also automated vehicles to take appropriate action in situations with a high risk of accidents.

[0204] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0205] Step 1:

[0206] The server acquires road condition data from across the country. Specifically, it periodically collects data such as congestion levels, weather, and accident history from public institutions such as traffic management centers and the Japan Meteorological Agency via API. The acquired data is in formats such as JSON or XML, which it then parses appropriately and stores in an internal data store. The input for this step is road condition data from the external API, and the output is internal data ready for analysis.

[0207] Step 2:

[0208] The terminal collects the driver's driving data. This is done using on-board sensors and a smartphone app. Collected data items include speed, frequency of sudden braking, and driving time. The collected data is sent to the server in real time. The input for this step is driving data from the on-board sensors and smartphone, and the output is the data before analysis that is sent to the server.

[0209] Step 3:

[0210] The server integrates the received driving data and road condition data and performs preprocessing. Specifically, it complements missing values ​​in the data and removes abnormal values ​​(outliers). The preprocessed data is then standardized and normalized, and converted into a format suitable for the generative AI model. The input for this step is the driving data and road condition data, and the output is the preprocessed data.

[0211] Step 4:

[0212] The server inputs preprocessed data into a generative AI model to evaluate the risk level of a traffic accident. This model uses a machine learning algorithm to perform risk assessment based on the preprocessed data. As a specific example, the risk assessment algorithm calculates a risk score using a neural network. If the risk score exceeds a certain threshold, the risk is determined to be high. The input to this step is preprocessed data, and the output is the risk level assessment result (risk score).

[0213] Step 5:

[0214] The device issues a warning to the driver based on the risk level evaluation results. Warning methods include audio alerts, visual alerts, and haptic feedback, and may be a combination of these. For example, the driver may hear a voice prompt saying, "There is an accident-prone area ahead. Please be careful," and a warning message may be displayed on the dashboard. The input of this step is the risk level evaluation result, and the output is a warning to the driver.

[0215] Step 6:

[0216] The terminal intervenes in the automated control system of the autonomous vehicle and instructs it to take appropriate action. Specifically, if it determines that the risk is high, it sends instructions to adjust the vehicle's speed or change course. The input to this step is the risk level assessment result, and the output is an operational instruction to the automated control system.

[0217] Step 7:

[0218] After receiving the warning, the user reviews their driving behavior and strives for safe driving. They also provide information about the effectiveness of the warning and areas for improvement to the system through the feedback function. This feedback data is sent to the server and used to retrain the generative AI model. The input of this step is user feedback, and the output is data used for retraining.

[0219] Step 8:

[0220] The server receives feedback data from users and retrains the generative AI model. This allows the system to continuously improve its accuracy and enable more effective risk assessment and warning. The input of this step is the feedback data from users, and the output is the retrained generative AI model.

[0221] 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.

[0222] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. Below, we will explain in detail the program processing of this system in natural language.

[0223] Data collection and analysis program

[0224] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, and more. Meanwhile, the terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. The collected data is sent to the server in real time.

[0225] The server preprocesses the received data, converting it into an appropriate format, imputing missing values, and detecting and removing outliers, making the data suitable for analysis.

[0226] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[0227] Real-time alert program

[0228] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, and conversely, if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[0229] Feedback and continuous improvement of the system

[0230] The user is encouraged to review their driving behavior and strive for safe driving after receiving the warning. If the user receives multiple warnings, the system provides feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device.

[0231] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0232] Specific examples

[0233] 1. Data Collection Examples

[0234] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0235] The device collects driving data from the past week from Driver A's smartphone and emotional data (e.g., stress levels and changes in facial expressions) obtained from a facial recognition camera, and sends the data to a server.

[0236] 2. Data analysis examples

[0237] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0238] The emotion engine assesses Driver A's stress level as "high," and the generative AI model determines the risk level as "high."

[0239] 3. Examples of warnings

[0240] Just before the user, Driver A, approaches that particular intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0241] At the same time, a warning light on the dashboard will flash to alert you.

[0242] In this way, the present invention predicts and warns about the risk of traffic accidents by analyzing the individual driving characteristics of drivers and nationwide road conditions in real time, and also taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents from occurring.

[0243] The processing flow will be explained below.

[0244] Step 1: Data collection

[0245] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0246] The device uses in-vehicle sensors, a smartphone app, and emotion recognition devices (e.g., cameras and microphones) to collect the driver's driving and emotional data, including speed, frequency of sudden braking, driving time, facial expressions, tone of voice, etc. The collected data is sent to a server in real time.

[0247] Step 2: Data Preprocessing

[0248] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, fills in missing values, and detects and removes outliers. It also organizes the data appropriately as time-series data. Meanwhile, it also performs format conversion and preprocessing on emotion data.

[0249] Step 3: Data analysis

[0250] The server inputs the preprocessed data into the generative AI model and emotion engine. The generative AI model learns from a large amount of past data and integrates road condition data and driving data for analysis. The emotion engine analyzes the driver's emotional state from their facial expressions and voice and feeds the results back to the generative AI model. For example, if the driver is under stress, the engine will reflect this and adjust the risk assessment.

[0251] Step 4: Risk assessment

[0252] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model and emotion engine. The evaluation results are classified into categories such as "low risk," "medium risk," and "high risk." If the emotion data indicates high stress, the risk level is adjusted higher.

[0253] Step 5: Send an alert

[0254] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used saying, "There is an accident hotspot ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0255] Step 6: Adjust your driving behavior

[0256] Users are encouraged to review their driving behavior and focus on safe driving by receiving the warning. This includes slowing down and paying attention to their surroundings. The system also includes voice guidance and changes to the alert content to help drivers relax.

[0257] Step 7: Gather feedback

[0258] If the user receives multiple warnings, they are asked to provide feedback on the effectiveness of the warnings and areas for improvement. This feedback includes information on the impact of the warnings, changes in driving behavior, and changes in emotions. The feedback data is sent to the server via the device.

[0259] Step 8: Update the Model

[0260] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0261] Example 2

[0262] 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."

[0263] Conventional driver assistance systems not only collect and analyze driver data and nationwide road condition data in real time, but also have the problem of being unable to perform risk assessments that take into account the driver's emotional state. Furthermore, data pre-processing during risk assessments is often insufficient, potentially reducing the accuracy of the analysis. Furthermore, warning methods for increased risk are limited, making them unable to adapt to a variety of situations.

[0264] 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.

[0265] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for preprocessing this data and converting it into an appropriate format, means for completing missing values ​​and detecting and removing outliers, means for inputting the preprocessed data into a generative AI model to evaluate the risk level of a traffic accident, means for collecting driver emotion data and reflecting it in risk assessment, and means for issuing a warning to the driver when the risk level increases. This enables highly accurate risk assessment and diverse warning methods that take into account the driver's emotional state while improving data accuracy.

[0266] "Driving data" refers to data related to the driver's driving behavior, specifically data including speed, frequency of sudden braking, driving time, and the like.

[0267] "Road condition data" refers to data on traffic conditions across the country, including congestion levels, weather, accident history, and the like.

[0268] "Preprocessing" is the process of converting collected data into a format suitable for analysis, and is a procedure for filling in missing values ​​and detecting and removing outliers.

[0269] A "generative AI model" is an algorithm or software that uses machine learning technology to learn from large amounts of past data and perform risk assessments based on new data.

[0270] "Risk level" is an index that evaluates the probability and degree of danger of a traffic accident, and is a number calculated using a generative AI model.

[0271] "Emotion data" refers to data relating to emotions acquired from the driver's facial expressions and voice, and includes stress levels, emotional states, and the like.

[0272] A "warning" is a notification to a driver that the risk of a traffic accident has increased, and may take the form of an audio alert, visual alert, or haptic feedback.

[0273] "Feedback" is information provided by the driver about changes in driving behavior and the impact of warnings, and is data that is sent to the server and used to improve the system.

[0274] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. The implementation method of this system will be described in detail below.

[0275] Data collection methods

[0276] The device collects the driver's driving data using on-board sensors and smartphones. This driving data includes speed, frequency of sudden braking, and driving time. It also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. This data is sent to a server in real time. As a specific example of collection, speed data is obtained using the smartphone's GPS function, and the frequency of sudden braking is detected using on-board sensors. Facial expression data is collected using a camera, and the driver's voice is recorded using a microphone, and voice tone analysis is performed.

[0277] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, etc. As a specific example, it obtains real-time road congestion data from the Traffic Management Center, receives current and forecast weather information from the Japan Meteorological Agency, and downloads past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0278] Data preprocessing and analysis

[0279] The server preprocesses the received data and converts it into an appropriate format. First, it performs missing value imputation and outlier detection and removal to prepare the data for analysis. For example, preprocessing involves generating a histogram of the velocity data and identifying and removing anomalous data points.

[0280] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model uses an algorithm trained on a large amount of past data. The model integrates driving data, road condition data, and driver emotion data to calculate a risk score. For example, if the driver is experiencing high stress, the model will assess the risk level as high.

[0281] Real-time alerts and feedback

[0282] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are multiple ways to warn the driver, including providing audio alerts, visual alerts, and haptic feedback depending on the driver's emotional state. For example, if the driver is nervous, a gentle audio alert will be played saying, "There is a high-accident zone ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0283] The user receives the warning and is encouraged to review their driving behavior and drive safely. After receiving multiple warnings, the user is also given feedback on the impact of the warning and areas for improvement. The feedback data includes information on the impact of the warning and changes in driving behavior, and is sent to the server via the device.

[0284] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0285] Examples and prompts

[0286] As an example of data collection, a server obtains road congestion data for Tokyo in real time from a traffic management center at 5:00 PM on a certain day. A device also collects driving data from Driver A's smartphone for the past week and emotion data obtained from a facial recognition camera, and sends these data to the server. Specific examples of prompts include a voice alert such as "There have been many accidents at the next intersection over the past three years. Please be careful," or a visual message such as "Be careful when making the next left turn."

[0287] In this way, the present invention provides a system that predicts and warns about traffic accident risks by analyzing individual driver characteristics and nationwide road conditions in real time, taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents.

[0288] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0289] Step 1: Data collection

[0290] The device collects driving data from the driver using on-board sensors and smartphones. Specifically, it obtains speed data from GPS and the frequency of sudden braking from the brake pedal sensor. It also detects facial expressions with a camera and records voice data with a microphone for emotion recognition. This data is sent to a server in real time.

[0291] Input: Driving data (speed, frequency of sudden braking, driving time), emotional data (facial expressions, voice)

[0292] Output: Driving and emotion data sent to a server in real time

[0293] Step 2: External Data Acquisition

[0294] The server receives nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Specifically, it receives road congestion data from the Traffic Management Center, weather information from the Japan Meteorological Agency, and past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0295] Input: Road condition data (congestion, weather, accident history)

[0296] Output: Road condition data stored on the server

[0297] Step 3: Data Preprocessing

[0298] The server preprocesses the received driving data, emotion data, and road condition data, including data format conversion, missing value imputation, and outlier removal. Specifically, it removes outliers in speed data and imputes missing emotion data using statistical methods.

[0299] Input: Received driving data, emotion data, road condition data

[0300] Output: Preprocessed data

[0301] Step 4: Risk assessment

[0302] The server inputs the preprocessed data into a generative AI model to assess traffic accident risk. The generative AI model calculates a risk score by integrating driving data, road condition data, and emotional data. For example, if the driver is experiencing high stress, the model will assess the risk level higher.

[0303] Input: Preprocessed data

[0304] Output: Risk score

[0305] Step 5: Send an alert

[0306] If the risk score exceeds a certain threshold as a result of the risk assessment, the device will issue a real-time warning to the driver. The warning method is a voice alert, playing a gentle message saying, "There is a high-accident zone ahead. Please be careful." The visual alert is a warning message displayed on the dashboard.

[0307] Input: Risk Score

[0308] Output: Real-time warning (audio alert, visual alert)

[0309] Step 6: Gather feedback

[0310] After receiving the warning, the user reviews their driving behavior and provides feedback, including the impact of the warning and changes in their driving behavior, which is sent to the server via the device.

[0311] Input: Feedback data on driving behavior revision

[0312] Output: Feedback data sent to the server

[0313] Step 7: Update the model

[0314] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, thereby improving the accuracy of risk assessment. Once retrained, the generative AI model is applied to the system and used for subsequent risk assessments.

[0315] Input: Feedback data

[0316] Output: An updated generative AI model

[0317] Through the above steps, the system of the present invention takes into account the driver's driving characteristics, national road conditions, and the driver's emotional state, predicts traffic accident risks, and issues warnings, thereby promoting safe driving.

[0318] (Application example 2)

[0319] 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."

[0320] Conventional traffic accident prevention systems assessed risk using only driver driving data and road condition data, making it impossible to accurately assess risk by taking the driver's emotional state into account. Furthermore, the warning format was uniform, making it difficult to issue a warning appropriate to the driver's emotional state. Therefore, improving the effectiveness of traffic accident risk prediction and warnings remains a challenge.

[0321] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0322] In this invention, the server includes a means for collecting driving data and emotional data of the driver, a means for acquiring nationwide road condition data and weather data, a means for analyzing these data and using a generative AI model to evaluate the risk level of a traffic accident, and a means for issuing an audio or visual alert according to the emotional state of the driver when the risk level increases. This makes it possible to integrate the driving data and emotional data of the driver to evaluate the risk of a traffic accident with high accuracy and to issue a warning appropriate to the emotional state of the driver.

[0323] "Driving data" refers to information such as the driver's speed, frequency of sudden braking, and driving time.

[0324] "Emotion data" refers to information about the emotional state of the driver as recognized from their facial expressions and voice.

[0325] "Road condition data" refers to information such as road congestion, weather, and accident history obtained from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[0326] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of past data to assess the risk of traffic accidents.

[0327] "Audio alert" refers to a means of issuing an audio warning to the driver.

[0328] "Visual alert" refers to a means of visually displaying a warning to the driver.

[0329] "Haptic feedback" refers to a means of providing tactile feedback such as vibration to the driver.

[0330] "Risk level" refers to an index that indicates the risk of a traffic accident occurring.

[0331] "Driving behavior patterns" refers to information that indicates a driver's driving habits and characteristic driving behavior.

[0332] The present invention is a system that collects driving data and emotional data of drivers and evaluates and warns about traffic accident risks in real time. An embodiment of this system will be described below with specific examples.

[0333] Data collection and analysis program

[0334] The server obtains road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Road condition data includes congestion levels, weather, and accident history. The terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice.

[0335] The collected data is sent to a server in real time. The server preprocesses the received data, converting it into an appropriate format, and also complements missing values ​​and removes outliers. The preprocessed data is then input into a generative AI model to assess traffic accident risk. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[0336] Real-time alert program

[0337] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, while if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[0338] Feedback and continuous improvement of the system

[0339] Upon receiving the warning, the user will review their driving behavior and strive for safer driving. If the user receives multiple warnings, the system will provide feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device. The server analyzes the received feedback data and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and enables the system to provide more effective warnings.

[0340] Specific examples

[0341] 1. Data Collection Examples

[0342] The server obtains real-time data on road congestion in Tokyo at 5:00 PM on a given day from the traffic management center. The device collects driving data from the past week from the driver's smartphone and emotional data obtained from a facial recognition camera (e.g., stress level and changes in facial expression), and sends the data to the server.

[0343] 2. Data analysis examples

[0344] The server checks that a particular intersection in Tokyo has had a high number of accidents over the past three years. The emotion engine evaluates the driver's stress level as "high," and the generative AI model determines the risk level as "high."

[0345] 3. Examples of warnings

[0346] Just before the user (driver) approaches a specific intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful." At the same time, a warning light on the dashboard flashes to draw attention.

[0347] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[0348] As described above, this invention analyzes the individual driving characteristics of drivers and nationwide road conditions in real time, and also takes into account the emotional state of the driver, thereby predicting and warning about the risk of traffic accidents, thereby promoting safe driving and preventing traffic accidents from occurring.

[0349] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0350] Step 1:

[0351] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, and accident history. As input, it receives real-time road condition information from each data source, and as output, it integrates and stores this data. Specifically, it sends API requests, analyzes the response data, and extracts the necessary information.

[0352] Step 2:

[0353] The device collects driving data from the driver using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, etc. It also uses a camera and microphone to collect emotional data from the driver's facial expressions and voice. As input, it receives real-time driving data and emotional data from the sensors and devices, and as output, it sends this data to a server. Specifically, it acquires the data stream from the device, unifies the format, and transmits it.

[0354] Step 3:

[0355] The server preprocesses the received driving data and emotion data. This preprocessing includes converting the data format, filling in missing values, and removing outliers. It receives the driving data and emotion data sent from the device as input, and generates preprocessed data as output. Specifically, it stores the data in a database while processing missing values ​​and outliers within the program.

[0356] Step 4:

[0357] The server inputs the preprocessed data into a generative AI model to evaluate the risk of traffic accidents. The generative AI model learns from a huge amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It receives preprocessed driving data and emotion data as input, and generates risk assessment results as output. Specifically, it inputs data into the generative AI model, analyzes the prediction results, and calculates the risk level.

[0358] Step 5:

[0359] The server also reflects the driver's emotional data in the generative AI model and determines the risk level based on the risk assessment results. It receives the emotional assessment data from the emotion engine and the risk assessment results from the generative AI model as input, and generates the final risk level as output. Specifically, it integrates the emotional assessment data with the output of the generative AI model and modifies the risk level.

[0360] Step 6:

[0361] If the risk assessment determines that there is a high risk of a traffic accident, the device issues a real-time warning to the driver. The format of the warning is adjusted according to the driver's emotional state. As input, the device receives the risk assessment results and emotional evaluation data from the server, and as output, issues an appropriate warning to the driver. Specifically, the device selects and executes the method of issuing an audio or visual alert based on the driver's emotional state.

[0362] Step 7:

[0363] When the user receives the warning, they will review their driving behavior and strive for safe driving. If they receive multiple warnings, they will be given feedback on the effectiveness of the warnings and areas for improvement. The system receives driver feedback data as input and sends this data to the server as output. Specific operations include providing an interface for collecting feedback and sending the input information to the server.

[0364] Step 8:

[0365] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and provides more effective warnings. It receives user feedback and emotion data as input and creates improved versions of the generative AI model and emotion engine as output. Specifically, it uses the collected data to retrain the algorithm and update and apply the new model.

[0366] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[0367] 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.

[0368] 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.

[0369] 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.

[0370] [Second embodiment]

[0371] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0372] 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.

[0373] 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).

[0374] 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.

[0375] 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.

[0376] 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).

[0377] 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.

[0378] 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.

[0379] 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.

[0380] 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.

[0381] 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.

[0382] 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."

[0383] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. Below, we will explain in detail the program processing of this system in natural language.

[0384] Data collection and analysis program

[0385] The server first periodically acquires nationwide road condition data. This data includes congestion levels, weather, accident history, and other information provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, and other organizations. Meanwhile, the device uses on-board sensors and a smartphone app to collect driver data. This data includes speed, frequency of sudden braking, driving time, and other information. The collected data is then sent to the server in real time.

[0386] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0387] Warning program

[0388] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are several ways to issue the warning, mainly audio alerts, visual alerts, haptic feedback, or a combination of these. For example, a voice prompt saying, "There is a high-accident zone ahead. Please be careful," will be played, and a warning message will be displayed on the dashboard.

[0389] Feedback and continuous improvement of the system

[0390] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0391] Specific examples

[0392] 1. Data Collection Examples

[0393] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0394] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0395] 2. Data analysis examples

[0396] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0397] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0398] 3. Examples of warnings

[0399] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0400] At the same time, a warning light on the dashboard will flash to alert you.

[0401] In this way, this invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about the risk of traffic accidents.

[0402] The processing flow will be explained below.

[0403] Step 1: Data collection

[0404] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0405] The device collects the driver's driving data (speed, frequency of sudden braking, driving time, etc.) through on-board sensors and a smartphone app and transmits it to a server in real time.

[0406] Step 2: Data Preprocessing

[0407] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, complements missing values, detects and removes outliers, and organizes the data appropriately as time series data.

[0408] Step 3: Data analysis

[0409] The server inputs the preprocessed data into a generative AI model. The generative AI model learns from a large amount of past data and integrates and analyzes road condition data and driving data. Specifically, it evaluates the current driving situation based on factors such as traffic volume, weather conditions, and past accident history, and calculates the risk level of a traffic accident.

[0410] Step 4: Risk assessment

[0411] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model, which are then classified into categories such as "low risk," "medium risk," and "high risk."

[0412] Step 5: Send an alert

[0413] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is given by audio alert, visual alert, haptic feedback, or a combination of these. For example, a voice prompt saying "There is a high-accident zone ahead. Please be careful" is played, and a warning message is displayed on the dashboard at the same time.

[0414] Step 6: Adjust your driving behavior

[0415] The user receives a warning and is encouraged to review their driving behavior and take steps to drive safely, such as slowing down and paying attention to their surroundings.

[0416] Step 7: Gather feedback

[0417] If the user receives multiple warnings, they provide feedback on the effectiveness and areas for improvement. This feedback includes information on changes in driving behavior and the effectiveness of the system. The feedback data is sent to the server via the device.

[0418] Step 8: Update the Model

[0419] The server analyzes the received feedback data and uses it to retrain the generative AI model, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0420] Example 1

[0421] 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."

[0422] In order to reduce the risk of traffic accidents, it is essential to provide drivers with accurate warnings in real time. However, current systems have difficulty effectively integrating data on road conditions and driver behavior nationwide to make highly accurate risk predictions. To address this issue, there is a need for a system that can accurately predict the risk of traffic accidents and issue warnings to drivers in real time.

[0423] 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.

[0424] In this invention, the server includes a means for acquiring nationwide road condition data, a means for collecting driver driving data, a means for preprocessing the acquired and collected data, a means including a generative AI model for analyzing the preprocessed data to evaluate the risk level of a traffic accident, and a means for issuing a warning to the driver in real time when the risk level increases. This enables highly accurate risk prediction based on individual driving characteristics and nationwide road conditions, and allows drivers to receive real-time warnings to prevent the risk of a traffic accident.

[0425] "National road condition data" refers to data including road congestion, weather conditions, and past accident history obtained from data providers such as the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[0426] "Driver driving data" refers to data related to driving behavior, such as vehicle speed, number of sudden braking attempts, driving time, and location information, collected using in-vehicle sensors and smartphone apps.

[0427] "Preprocessing" refers to the process of filling in missing values ​​in collected data, removing outliers, and converting the data into a format suitable for analysis.

[0428] A "generative AI model" is an artificial intelligence model built on a large dataset, and is a machine learning or deep learning model used to assess the risk of traffic accidents.

[0429] A "warning" is a real-time notification to the driver when the risk of a traffic accident increases, and may be provided via audio alerts, visual alerts, and / or haptic feedback.

[0430] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. This system is mainly composed of a server, a terminal, and a user, and specific embodiments are described below.

[0431] Data collection methods

[0432] The server periodically obtains nationwide road condition data provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. To obtain this data, it receives data in JSON format from multiple API endpoints.

[0433] Meanwhile, the device collects driving data from the driver using on-board sensors and a smartphone app. Collected data includes the vehicle's speed, number of sudden braking attempts, driving time, and location information. This data is sent to a server in real time. The smartphone app on the device periodically collects data from the sensors in the background and transfers it to the server every few seconds.

[0434] Data analysis format

[0435] The server preprocesses the received data. This preprocessing involves filling in missing values, removing outliers, and converting the data into a format suitable for analysis. Specifically, data cleaning is performed using a Python program and the Pandas library. Missing values ​​are filled in using the average of the preceding and following data, and abnormal values ​​(e.g., speed data that suddenly reaches several hundred km / h) are removed as outliers.

[0436] The preprocessed data is then fed into a generative AI model to assess the risk level of a traffic accident. This generative AI model is a deep learning model built using TensorFlow or PyTorch and trained on a large dataset. The model's output includes a risk score and specific warnings.

[0437] Warning issuance format

[0438] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. This warning can be provided by audio alerts, visual alerts, and / or haptic feedback. For example, a voice prompt such as "There is a high-accident zone ahead. Please be careful" may be played from the smartphone, while a warning message is simultaneously displayed on the in-car display.

[0439] The device uses push notifications or Bluetooth connectivity to display information on the in-car display or smartphone, allowing the driver to quickly recognize important information on the spot.

[0440] Forms of feedback and system improvement

[0441] When users receive a warning from the system, they can review their driving behavior and strive for safer driving. If they receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback is given in the form of a survey via a smartphone app. For example, questions such as "Was the warning helpful?" and "Has it changed your driving behavior?" are displayed.

[0442] The feedback data is sent to a server and used to retrain the generative AI model. The server retrains the model based on new data, improving the accuracy of the system. This process is repeated periodically to ensure continuous improvement of the system.

[0443] Examples of specific examples and prompts

[0444] 1. Data Collection Example

[0445] The server obtains real-time road congestion data for Tokyo at 5 p.m. from the traffic management center.

[0446] The device collects driving data from the driver's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to a server.

[0447] 2. Data Analysis Example

[0448] The server checks that a particular intersection in Tokyo has had many accidents over the past three years.

[0449] The analysis results show that the risk level is high because the particular driver frequently brakes suddenly during evening traffic jams.

[0450] 3. Examples of warnings

[0451] Just before the user (driver) approaches a specific intersection, the device issues a voice alert saying, "There is a high-accident area ahead. Please slow down and be careful."

[0452] At the same time, a warning light on the dashboard will flash to alert you.

[0453] Example prompt sentence:

[0454] "Please explain how you collect and analyze driver behavior data and nationwide road condition data to predict the risk of traffic accidents, and how you issue warnings in real time."

[0455] The above is a specific embodiment for carrying out the present invention. This system enables drivers to accurately predict risks based on their individual driving characteristics and nationwide road conditions, and provides real-time warnings to prevent the risk of traffic accidents.

[0456] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0457] Step 1:

[0458] The server periodically obtains road condition data from across the country. As input, it receives data in JSON format from API endpoints provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. As output, the received data is stored in the server's database.

[0459] Specifically, the server periodically sends an API request like this:

[0460] GET / road_conditions?region=tokyo&datetime=2023-10-12T17:00:00

[0461] GET / weather?region=tokyo&datetime=2023-10-12T17:00:00

[0462] GET / accident_history?region=tokyo&datetime=2023-10-12

[0463] This allows the latest road condition data to be acquired.

[0464] Step 2:

[0465] The device collects the driver's driving data using on-board sensors and a smartphone app and transmits it to a server in real time. The inputs include speed, number of sudden braking attempts, driving time, location information, etc. As an output, the collected data is transferred to the server and added to a database.

[0466] Specifically, the device app periodically sends sensor data to the server in the background:

[0467] POST / upload_driving_data

[0468] Content-Type: application / json

[0469] {

[0470] "speed": 60,

[0471] "brake_count": 3,

[0472] "driving_time": "30 minutes",

[0473] "location": "35.6895, 139.6917"

[0474] }

[0475] Step 3:

[0476] The server preprocesses the received data. The input is the road condition data and driving data collected in Step 1 and Step 2. This data may contain missing values ​​and outliers, so these are imputed and removed. The output is the preprocessed, clean data.

[0477] Specifically, the server runs a Python script that uses the Pandas library to clean the data:

[0478] python

[0479] import pandas as pd

[0480] Data loading

[0481] data = pd.read_json("driving_data.json")

[0482] Missing value imputation

[0483] data['speed'].fillna(data['speed'].mean(), inplace=True)

[0484] Outlier removal

[0485] data = data[data['speed'] < 150]

[0486] Step 4:

[0487] The server inputs the preprocessed data into the generative AI model to perform risk assessment. The input includes preprocessed road condition data and driving data. The generative AI model integrates and analyzes this data to assess the risk level of traffic accidents. The output is a risk score and specific warning content.

[0488] Specifically, the server runs models built using TensorFlow or PyTorch:

[0489] python

[0490] import tensorflow as tf

[0491] Model Loading

[0492] model = tf.keras.models.load_model('risk_model.h5')

[0493] prediction

[0494] predictions = model.predict(preprocessed_data)

[0495] risk_score = predictions[0]

[0496] Step 5:

[0497] As a result of the risk assessment, the device issues a warning to the driver. The inputs include a risk score and warning content obtained from the generative AI model. The output is real-time notification to the driver in the form of audio alerts, visual alerts, and haptic feedback.

[0498] Specifically, the device will display a warning message on the car's display via Bluetooth:

[0499] python

[0500] Audio alerts

[0501] play_sound("warning_audio.mp3")

[0502] Visual Alerts

[0503] show_message("Accident-prone area ahead. Please slow down and take care.")

[0504] Haptic Feedback

[0505] vibrate_device()

[0506] Step 6:

[0507] Users receive warnings from the system and are encouraged to review their driving behavior and drive safely. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. Feedback data in the form of a questionnaire is included as input. The feedback data is sent to the server as output and used to improve the system.

[0508] Specifically, users answer a survey via a smartphone app:

[0509] POST / feedback

[0510] Content-Type: application / json

[0511] {

[0512] "useful": true,

[0513] "behavior_change": "Yes"

[0514] }

[0515] Through these processing steps, the system can use driver behavior data and nationwide road condition data to predict the risk of traffic accidents and issue warnings in real time.

[0516] (Application example 1)

[0517] 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."

[0518] There is a challenge to promote safe driving and prevent traffic accidents by providing a system that predicts the risk of traffic accidents in real time and issues effective warnings to drivers and autonomous vehicles. In addition, with conventional systems, it is difficult for autonomous vehicles to properly assess risks and automatically take appropriate control, so we would like to improve this aspect as well.

[0519] 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.

[0520] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for analyzing this data and assessing the risk level of traffic accidents, means for issuing a warning to the driver when the risk level increases, and means for intervening in the automatic control system of the autonomous vehicle, thereby enabling the driver and the autonomous driving system to receive risk information in real time and take appropriate measures.

[0521] "Driving data" refers to information relating to the driver's driving behavior, including data such as speed, frequency of sudden braking, and driving time.

[0522] "Road condition data" is information about roads across the country, including congestion levels, weather, accident history, and so on.

[0523] "Analysis" refers to the process of processing and analyzing collected data to assess the risk of traffic accidents.

[0524] The "risk level" is an evaluation index that indicates the possibility of a traffic accident occurring.

[0525] "Warnings" are messages or signals that alert the driver when the risk level increases, and can take the form of audio alerts, visual alerts, or haptic feedback.

[0526] An "automatic control system" is a system that automatically controls the driving operations of an autonomous vehicle. It receives information on traffic risks and adjusts vehicle speed and changes course.

[0527] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and conduct risk assessments.

[0528] A "prompt sentence" is an instruction sentence used as input to a generative AI model.

[0529] This invention is a system that collects and analyzes driver's driving behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers and self-driving vehicles in real time. This system includes the following means.

[0530] Data collection and analysis

[0531] The server periodically collects nationwide road condition data. This data includes congestion levels, weather, and accident history provided by the Traffic Management Center and the Japan Meteorological Agency. The device also collects driver driving data using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, and more. The collected data is sent to the server in real time.

[0532] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0533] Warning

[0534] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver or the autonomous vehicle's control system. The warning can be issued in multiple ways, primarily via audio alerts, visual alerts, or haptic feedback, or a combination of these. For example, a voice prompt saying, "Be careful. Accident hotspot ahead," may be played, along with a warning message displayed on the dashboard. The device may also instruct the autonomous vehicle to adjust speed or change course based on pre-set prompts.

[0535] Feedback and continuous improvement of the system

[0536] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0537] Specific examples

[0538] Data Collection Examples

[0539] The server obtains real-time data on road congestion in urban areas at 5 p.m. on a given day from the traffic management center.

[0540] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0541] Data analysis examples

[0542] The server checks that a particular intersection has had a high number of accidents over the past three years.

[0543] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0544] Examples of warnings issued

[0545] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0546] At the same time, a warning light on the dashboard will flash to alert you.

[0547] In the case of self-driving vehicles, the terminal sends instructions to the automatic control system to adjust speed and change course.

[0548] Example prompts to input to the generative AI model

[0549] Enter your driving and road condition data to assess your risk of traffic accidents.

[0550] Driving data: speed, frequency of hard braking, driving time

[0551] Road condition data: congestion, weather, accident history

[0552] This invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about traffic accident risks. This enables not only drivers but also automated vehicles to take appropriate action in situations with a high risk of accidents.

[0553] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0554] Step 1:

[0555] The server acquires road condition data from across the country. Specifically, it periodically collects data such as congestion levels, weather, and accident history from public institutions such as traffic management centers and the Japan Meteorological Agency via API. The acquired data is in formats such as JSON or XML, which it then parses appropriately and stores in an internal data store. The input for this step is road condition data from the external API, and the output is internal data ready for analysis.

[0556] Step 2:

[0557] The terminal collects the driver's driving data. This is done using on-board sensors and a smartphone app. Collected data items include speed, frequency of sudden braking, and driving time. The collected data is sent to the server in real time. The input for this step is driving data from the on-board sensors and smartphone, and the output is the data before analysis that is sent to the server.

[0558] Step 3:

[0559] The server integrates the received driving data and road condition data and performs preprocessing. Specifically, it complements missing values ​​in the data and removes abnormal values ​​(outliers). The preprocessed data is then standardized and normalized, and converted into a format suitable for the generative AI model. The input for this step is the driving data and road condition data, and the output is the preprocessed data.

[0560] Step 4:

[0561] The server inputs preprocessed data into a generative AI model to evaluate the risk level of a traffic accident. This model uses a machine learning algorithm to perform risk assessment based on the preprocessed data. As a specific example, the risk assessment algorithm calculates a risk score using a neural network. If the risk score exceeds a certain threshold, the risk is determined to be high. The input to this step is preprocessed data, and the output is the risk level assessment result (risk score).

[0562] Step 5:

[0563] The device issues a warning to the driver based on the risk level evaluation results. Warning methods include audio alerts, visual alerts, and haptic feedback, and may be a combination of these. For example, the driver may hear a voice prompt saying, "There is an accident-prone area ahead. Please be careful," and a warning message may be displayed on the dashboard. The input of this step is the risk level evaluation result, and the output is a warning to the driver.

[0564] Step 6:

[0565] The terminal intervenes in the automated control system of the autonomous vehicle and instructs it to take appropriate action. Specifically, if it determines that the risk is high, it sends instructions to adjust the vehicle's speed or change course. The input to this step is the risk level assessment result, and the output is an operational instruction to the automated control system.

[0566] Step 7:

[0567] After receiving the warning, the user reviews their driving behavior and strives for safe driving. They also provide information about the effectiveness of the warning and areas for improvement to the system through the feedback function. This feedback data is sent to the server and used to retrain the generative AI model. The input of this step is user feedback, and the output is data used for retraining.

[0568] Step 8:

[0569] The server receives feedback data from users and retrains the generative AI model. This allows the system to continuously improve its accuracy and enable more effective risk assessment and warning. The input of this step is the feedback data from users, and the output is the retrained generative AI model.

[0570] 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.

[0571] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. Below, we will explain in detail the program processing of this system in natural language.

[0572] Data collection and analysis program

[0573] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, and more. Meanwhile, the terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. The collected data is sent to the server in real time.

[0574] The server preprocesses the received data, converting it into an appropriate format, imputing missing values, and detecting and removing outliers, making the data suitable for analysis.

[0575] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[0576] Real-time alert program

[0577] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, and conversely, if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[0578] Feedback and continuous improvement of the system

[0579] The user is encouraged to review their driving behavior and strive for safe driving after receiving the warning. If the user receives multiple warnings, the system provides feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device.

[0580] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0581] Specific examples

[0582] 1. Data Collection Examples

[0583] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0584] The device collects driving data from the past week from Driver A's smartphone and emotional data (e.g., stress levels and changes in facial expressions) obtained from a facial recognition camera, and sends the data to a server.

[0585] 2. Data analysis examples

[0586] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0587] The emotion engine assesses Driver A's stress level as "high," and the generative AI model determines the risk level as "high."

[0588] 3. Examples of warnings

[0589] Just before the user, Driver A, approaches that particular intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0590] At the same time, a warning light on the dashboard will flash to alert you.

[0591] In this way, the present invention predicts and warns about the risk of traffic accidents by analyzing the individual driving characteristics of drivers and nationwide road conditions in real time, and also taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents from occurring.

[0592] The processing flow will be explained below.

[0593] Step 1: Data collection

[0594] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0595] The device uses in-vehicle sensors, a smartphone app, and emotion recognition devices (e.g., cameras and microphones) to collect the driver's driving and emotional data, including speed, frequency of sudden braking, driving time, facial expressions, tone of voice, etc. The collected data is sent to a server in real time.

[0596] Step 2: Data Preprocessing

[0597] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, fills in missing values, and detects and removes outliers. It also organizes the data appropriately as time-series data. Meanwhile, it also performs format conversion and preprocessing on emotion data.

[0598] Step 3: Data analysis

[0599] The server inputs the preprocessed data into the generative AI model and emotion engine. The generative AI model learns from a large amount of past data and integrates road condition data and driving data for analysis. The emotion engine analyzes the driver's emotional state from their facial expressions and voice and feeds the results back to the generative AI model. For example, if the driver is under stress, the engine will reflect this and adjust the risk assessment.

[0600] Step 4: Risk assessment

[0601] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model and emotion engine. The evaluation results are classified into categories such as "low risk," "medium risk," and "high risk." If the emotion data indicates high stress, the risk level is adjusted higher.

[0602] Step 5: Send an alert

[0603] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used saying, "There is an accident hotspot ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0604] Step 6: Adjust your driving behavior

[0605] Users are encouraged to review their driving behavior and focus on safe driving by receiving the warning. This includes slowing down and paying attention to their surroundings. The system also includes voice guidance and changes to the alert content to help drivers relax.

[0606] Step 7: Gather feedback

[0607] If the user receives multiple warnings, they are asked to provide feedback on the effectiveness of the warnings and areas for improvement. This feedback includes information on the impact of the warnings, changes in driving behavior, and changes in emotions. The feedback data is sent to the server via the device.

[0608] Step 8: Update the Model

[0609] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0610] Example 2

[0611] 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."

[0612] Conventional driver assistance systems not only collect and analyze driver data and nationwide road condition data in real time, but also have the problem of being unable to perform risk assessments that take into account the driver's emotional state. Furthermore, data pre-processing during risk assessments is often insufficient, potentially reducing the accuracy of the analysis. Furthermore, warning methods for increased risk are limited, making them unable to adapt to a variety of situations.

[0613] 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.

[0614] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for preprocessing this data and converting it into an appropriate format, means for completing missing values ​​and detecting and removing outliers, means for inputting the preprocessed data into a generative AI model to evaluate the risk level of a traffic accident, means for collecting driver emotion data and reflecting it in risk assessment, and means for issuing a warning to the driver when the risk level increases. This enables highly accurate risk assessment and diverse warning methods that take into account the driver's emotional state while improving data accuracy.

[0615] "Driving data" refers to data related to the driver's driving behavior, specifically data including speed, frequency of sudden braking, driving time, and the like.

[0616] "Road condition data" refers to data on traffic conditions across the country, including congestion levels, weather, accident history, and the like.

[0617] "Preprocessing" is the process of converting collected data into a format suitable for analysis, and is a procedure for filling in missing values ​​and detecting and removing outliers.

[0618] A "generative AI model" is an algorithm or software that uses machine learning technology to learn from large amounts of past data and perform risk assessments based on new data.

[0619] "Risk level" is an index that evaluates the probability and degree of danger of a traffic accident, and is a number calculated using a generative AI model.

[0620] "Emotion data" refers to data relating to emotions acquired from the driver's facial expressions and voice, and includes stress levels, emotional states, and the like.

[0621] A "warning" is a notification to a driver that the risk of a traffic accident has increased, and may take the form of an audio alert, visual alert, or haptic feedback.

[0622] "Feedback" is information provided by the driver about changes in driving behavior and the impact of warnings, and is data that is sent to the server and used to improve the system.

[0623] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. The implementation method of this system will be described in detail below.

[0624] Data collection methods

[0625] The device collects the driver's driving data using on-board sensors and smartphones. This driving data includes speed, frequency of sudden braking, and driving time. It also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. This data is sent to a server in real time. As a specific example of collection, speed data is obtained using the smartphone's GPS function, and the frequency of sudden braking is detected using on-board sensors. Facial expression data is collected using a camera, and the driver's voice is recorded using a microphone, and voice tone analysis is performed.

[0626] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, etc. As a specific example, it obtains real-time road congestion data from the Traffic Management Center, receives current and forecast weather information from the Japan Meteorological Agency, and downloads past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0627] Data preprocessing and analysis

[0628] The server preprocesses the received data and converts it into an appropriate format. First, it performs missing value imputation and outlier detection and removal to prepare the data for analysis. For example, preprocessing involves generating a histogram of the velocity data and identifying and removing anomalous data points.

[0629] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model uses an algorithm trained on a large amount of past data. The model integrates driving data, road condition data, and driver emotion data to calculate a risk score. For example, if the driver is experiencing high stress, the model will assess the risk level as high.

[0630] Real-time alerts and feedback

[0631] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are multiple ways to warn the driver, including providing audio alerts, visual alerts, and haptic feedback depending on the driver's emotional state. For example, if the driver is nervous, a gentle audio alert will be played saying, "There is a high-accident zone ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0632] The user receives the warning and is encouraged to review their driving behavior and drive safely. After receiving multiple warnings, the user is also given feedback on the impact of the warning and areas for improvement. The feedback data includes information on the impact of the warning and changes in driving behavior, and is sent to the server via the device.

[0633] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0634] Examples and prompts

[0635] As an example of data collection, a server obtains road congestion data for Tokyo in real time from a traffic management center at 5:00 PM on a certain day. A device also collects driving data from Driver A's smartphone for the past week and emotion data obtained from a facial recognition camera, and sends these data to the server. Specific examples of prompts include a voice alert such as "There have been many accidents at the next intersection over the past three years. Please be careful," or a visual message such as "Be careful when making the next left turn."

[0636] In this way, the present invention provides a system that predicts and warns about traffic accident risks by analyzing individual driver characteristics and nationwide road conditions in real time, taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents.

[0637] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0638] Step 1: Data collection

[0639] The device collects driving data from the driver using on-board sensors and smartphones. Specifically, it obtains speed data from GPS and the frequency of sudden braking from the brake pedal sensor. It also detects facial expressions with a camera and records voice data with a microphone for emotion recognition. This data is sent to a server in real time.

[0640] Input: Driving data (speed, frequency of sudden braking, driving time), emotional data (facial expressions, voice)

[0641] Output: Driving and emotion data sent to a server in real time

[0642] Step 2: External Data Acquisition

[0643] The server receives nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Specifically, it receives road congestion data from the Traffic Management Center, weather information from the Japan Meteorological Agency, and past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0644] Input: Road condition data (congestion, weather, accident history)

[0645] Output: Road condition data stored on the server

[0646] Step 3: Data Preprocessing

[0647] The server preprocesses the received driving data, emotion data, and road condition data, including data format conversion, missing value imputation, and outlier removal. Specifically, it removes outliers in speed data and imputes missing emotion data using statistical methods.

[0648] Input: Received driving data, emotion data, road condition data

[0649] Output: Preprocessed data

[0650] Step 4: Risk assessment

[0651] The server inputs the preprocessed data into a generative AI model to assess traffic accident risk. The generative AI model calculates a risk score by integrating driving data, road condition data, and emotional data. For example, if the driver is experiencing high stress, the model will assess the risk level higher.

[0652] Input: Preprocessed data

[0653] Output: Risk score

[0654] Step 5: Send an alert

[0655] If the risk score exceeds a certain threshold as a result of the risk assessment, the device will issue a real-time warning to the driver. The warning method is a voice alert, playing a gentle message saying, "There is a high-accident zone ahead. Please be careful." The visual alert is a warning message displayed on the dashboard.

[0656] Input: Risk Score

[0657] Output: Real-time warning (audio alert, visual alert)

[0658] Step 6: Gather feedback

[0659] After receiving the warning, the user reviews their driving behavior and provides feedback, including the impact of the warning and changes in their driving behavior, which is sent to the server via the device.

[0660] Input: Feedback data on driving behavior revision

[0661] Output: Feedback data sent to the server

[0662] Step 7: Update the model

[0663] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, thereby improving the accuracy of risk assessment. Once retrained, the generative AI model is applied to the system and used for subsequent risk assessments.

[0664] Input: Feedback data

[0665] Output: An updated generative AI model

[0666] Through the above steps, the system of the present invention takes into account the driver's driving characteristics, national road conditions, and the driver's emotional state, predicts traffic accident risks, and issues warnings, thereby promoting safe driving.

[0667] (Application example 2)

[0668] 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."

[0669] Conventional traffic accident prevention systems assessed risk using only driver driving data and road condition data, making it impossible to accurately assess risk by taking the driver's emotional state into account. Furthermore, the warning format was uniform, making it difficult to issue a warning appropriate to the driver's emotional state. Therefore, improving the effectiveness of traffic accident risk prediction and warnings remains a challenge.

[0670] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0671] In this invention, the server includes a means for collecting driving data and emotional data of the driver, a means for acquiring nationwide road condition data and weather data, a means for analyzing these data and using a generative AI model to evaluate the risk level of a traffic accident, and a means for issuing an audio or visual alert according to the emotional state of the driver when the risk level increases. This makes it possible to integrate the driving data and emotional data of the driver to evaluate the risk of a traffic accident with high accuracy and to issue a warning appropriate to the emotional state of the driver.

[0672] "Driving data" refers to information such as the driver's speed, frequency of sudden braking, and driving time.

[0673] "Emotion data" refers to information about the emotional state of the driver as recognized from their facial expressions and voice.

[0674] "Road condition data" refers to information such as road congestion, weather, and accident history obtained from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[0675] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of past data to assess the risk of traffic accidents.

[0676] "Audio alert" refers to a means of issuing an audio warning to the driver.

[0677] "Visual alert" refers to a means of visually displaying a warning to the driver.

[0678] "Haptic feedback" refers to a means of providing tactile feedback such as vibration to the driver.

[0679] "Risk level" refers to an index that indicates the risk of a traffic accident occurring.

[0680] "Driving behavior patterns" refers to information that indicates a driver's driving habits and characteristic driving behavior.

[0681] The present invention is a system that collects driving data and emotional data of drivers and evaluates and warns about traffic accident risks in real time. An embodiment of this system will be described below with specific examples.

[0682] Data collection and analysis program

[0683] The server obtains road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Road condition data includes congestion levels, weather, and accident history. The terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice.

[0684] The collected data is sent to a server in real time. The server preprocesses the received data, converting it into an appropriate format, and also complements missing values ​​and removes outliers. The preprocessed data is then input into a generative AI model to assess traffic accident risk. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[0685] Real-time alert program

[0686] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, while if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[0687] Feedback and continuous improvement of the system

[0688] Upon receiving the warning, the user will review their driving behavior and strive for safer driving. If the user receives multiple warnings, the system will provide feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device. The server analyzes the received feedback data and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and enables the system to provide more effective warnings.

[0689] Specific examples

[0690] 1. Data Collection Examples

[0691] The server obtains real-time data on road congestion in Tokyo at 5:00 PM on a given day from the traffic management center. The device collects driving data from the past week from the driver's smartphone and emotional data obtained from a facial recognition camera (e.g., stress level and changes in facial expression), and sends the data to the server.

[0692] 2. Data analysis examples

[0693] The server checks that a particular intersection in Tokyo has had a high number of accidents over the past three years. The emotion engine evaluates the driver's stress level as "high," and the generative AI model determines the risk level as "high."

[0694] 3. Examples of warnings

[0695] Just before the user (driver) approaches a specific intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful." At the same time, a warning light on the dashboard flashes to draw attention.

[0696] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[0697] As described above, this invention analyzes the individual driving characteristics of drivers and nationwide road conditions in real time, and also takes into account the emotional state of the driver, thereby predicting and warning about the risk of traffic accidents, thereby promoting safe driving and preventing traffic accidents from occurring.

[0698] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0699] Step 1:

[0700] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, and accident history. As input, it receives real-time road condition information from each data source, and as output, it integrates and stores this data. Specifically, it sends API requests, analyzes the response data, and extracts the necessary information.

[0701] Step 2:

[0702] The device collects driving data from the driver using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, etc. It also uses a camera and microphone to collect emotional data from the driver's facial expressions and voice. As input, it receives real-time driving data and emotional data from the sensors and devices, and as output, it sends this data to a server. Specifically, it acquires the data stream from the device, unifies the format, and transmits it.

[0703] Step 3:

[0704] The server preprocesses the received driving data and emotion data. This preprocessing includes converting the data format, filling in missing values, and removing outliers. It receives the driving data and emotion data sent from the device as input, and generates preprocessed data as output. Specifically, it stores the data in a database while processing missing values ​​and outliers within the program.

[0705] Step 4:

[0706] The server inputs the preprocessed data into a generative AI model to evaluate the risk of traffic accidents. The generative AI model learns from a huge amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It receives preprocessed driving data and emotion data as input, and generates risk assessment results as output. Specifically, it inputs data into the generative AI model, analyzes the prediction results, and calculates the risk level.

[0707] Step 5:

[0708] The server also reflects the driver's emotional data in the generative AI model and determines the risk level based on the risk assessment results. It receives the emotional assessment data from the emotion engine and the risk assessment results from the generative AI model as input, and generates the final risk level as output. Specifically, it integrates the emotional assessment data with the output of the generative AI model and modifies the risk level.

[0709] Step 6:

[0710] If the risk assessment determines that there is a high risk of a traffic accident, the device issues a real-time warning to the driver. The format of the warning is adjusted according to the driver's emotional state. As input, the device receives the risk assessment results and emotional evaluation data from the server, and as output, issues an appropriate warning to the driver. Specifically, the device selects and executes the method of issuing an audio or visual alert based on the driver's emotional state.

[0711] Step 7:

[0712] When the user receives the warning, they will review their driving behavior and strive for safe driving. If they receive multiple warnings, they will be given feedback on the effectiveness of the warnings and areas for improvement. The system receives driver feedback data as input and sends this data to the server as output. Specific operations include providing an interface for collecting feedback and sending the input information to the server.

[0713] Step 8:

[0714] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and provides more effective warnings. It receives user feedback and emotion data as input and creates improved versions of the generative AI model and emotion engine as output. Specifically, it uses the collected data to retrain the algorithm and update and apply the new model.

[0715] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[0716] 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.

[0717] 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.

[0718] 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.

[0719] [Third embodiment]

[0720] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0721] 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.

[0722] 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).

[0723] 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.

[0724] 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.

[0725] 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).

[0726] 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.

[0727] 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.

[0728] 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.

[0729] 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.

[0730] 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.

[0731] 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."

[0732] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. Below, we will explain in detail the program processing of this system in natural language.

[0733] Data collection and analysis program

[0734] The server first periodically acquires nationwide road condition data. This data includes congestion levels, weather, accident history, and other information provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, and other organizations. Meanwhile, the device uses on-board sensors and a smartphone app to collect driver data. This data includes speed, frequency of sudden braking, driving time, and other information. The collected data is then sent to the server in real time.

[0735] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0736] Warning program

[0737] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are several ways to issue the warning, mainly audio alerts, visual alerts, haptic feedback, or a combination of these. For example, a voice prompt saying, "There is a high-accident zone ahead. Please be careful," will be played, and a warning message will be displayed on the dashboard.

[0738] Feedback and continuous improvement of the system

[0739] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0740] Specific examples

[0741] 1. Data Collection Examples

[0742] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0743] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0744] 2. Data analysis examples

[0745] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0746] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0747] 3. Examples of warnings

[0748] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0749] At the same time, a warning light on the dashboard will flash to alert you.

[0750] In this way, this invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about the risk of traffic accidents.

[0751] The processing flow will be explained below.

[0752] Step 1: Data collection

[0753] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0754] The device collects the driver's driving data (speed, frequency of sudden braking, driving time, etc.) through on-board sensors and a smartphone app and transmits it to a server in real time.

[0755] Step 2: Data Preprocessing

[0756] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, complements missing values, detects and removes outliers, and organizes the data appropriately as time series data.

[0757] Step 3: Data analysis

[0758] The server inputs the preprocessed data into a generative AI model. The generative AI model learns from a large amount of past data and integrates and analyzes road condition data and driving data. Specifically, it evaluates the current driving situation based on factors such as traffic volume, weather conditions, and past accident history, and calculates the risk level of a traffic accident.

[0759] Step 4: Risk assessment

[0760] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model, which are then classified into categories such as "low risk," "medium risk," and "high risk."

[0761] Step 5: Send an alert

[0762] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is given by audio alert, visual alert, haptic feedback, or a combination of these. For example, a voice prompt saying "There is a high-accident zone ahead. Please be careful" is played, and a warning message is displayed on the dashboard at the same time.

[0763] Step 6: Adjust your driving behavior

[0764] The user receives a warning and is encouraged to review their driving behavior and take steps to drive safely, such as slowing down and paying attention to their surroundings.

[0765] Step 7: Gather feedback

[0766] If the user receives multiple warnings, they provide feedback on the effectiveness and areas for improvement. This feedback includes information on changes in driving behavior and the effectiveness of the system. The feedback data is sent to the server via the device.

[0767] Step 8: Update the Model

[0768] The server analyzes the received feedback data and uses it to retrain the generative AI model, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0769] Example 1

[0770] 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."

[0771] In order to reduce the risk of traffic accidents, it is essential to provide drivers with accurate warnings in real time. However, current systems have difficulty effectively integrating data on road conditions and driver behavior nationwide to make highly accurate risk predictions. To address this issue, there is a need for a system that can accurately predict the risk of traffic accidents and issue warnings to drivers in real time.

[0772] 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.

[0773] In this invention, the server includes a means for acquiring nationwide road condition data, a means for collecting driver driving data, a means for preprocessing the acquired and collected data, a means including a generative AI model for analyzing the preprocessed data to evaluate the risk level of a traffic accident, and a means for issuing a warning to the driver in real time when the risk level increases. This enables highly accurate risk prediction based on individual driving characteristics and nationwide road conditions, and allows drivers to receive real-time warnings to prevent the risk of a traffic accident.

[0774] "National road condition data" refers to data including road congestion, weather conditions, and past accident history obtained from data providers such as the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[0775] "Driver driving data" refers to data related to driving behavior, such as vehicle speed, number of sudden braking attempts, driving time, and location information, collected using in-vehicle sensors and smartphone apps.

[0776] "Preprocessing" refers to the process of filling in missing values ​​in collected data, removing outliers, and converting the data into a format suitable for analysis.

[0777] A "generative AI model" is an artificial intelligence model built on a large dataset, and is a machine learning or deep learning model used to assess the risk of traffic accidents.

[0778] A "warning" is a real-time notification to the driver when the risk of a traffic accident increases, and may be provided via audio alerts, visual alerts, and / or haptic feedback.

[0779] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. This system is mainly composed of a server, a terminal, and a user, and specific embodiments are described below.

[0780] Data collection methods

[0781] The server periodically obtains nationwide road condition data provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. To obtain this data, it receives data in JSON format from multiple API endpoints.

[0782] Meanwhile, the device collects driving data from the driver using on-board sensors and a smartphone app. Collected data includes the vehicle's speed, number of sudden braking attempts, driving time, and location information. This data is sent to a server in real time. The smartphone app on the device periodically collects data from the sensors in the background and transfers it to the server every few seconds.

[0783] Data analysis format

[0784] The server preprocesses the received data. This preprocessing involves filling in missing values, removing outliers, and converting the data into a format suitable for analysis. Specifically, data cleaning is performed using a Python program and the Pandas library. Missing values ​​are filled in using the average of the preceding and following data, and abnormal values ​​(e.g., speed data that suddenly reaches several hundred km / h) are removed as outliers.

[0785] The preprocessed data is then fed into a generative AI model to assess the risk level of a traffic accident. This generative AI model is a deep learning model built using TensorFlow or PyTorch and trained on a large dataset. The model's output includes a risk score and specific warnings.

[0786] Warning issuance format

[0787] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. This warning can be provided by audio alerts, visual alerts, and / or haptic feedback. For example, a voice prompt such as "There is a high-accident zone ahead. Please be careful" may be played from the smartphone, while a warning message is simultaneously displayed on the in-car display.

[0788] The device uses push notifications or Bluetooth connectivity to display information on the in-car display or smartphone, allowing the driver to quickly recognize important information on the spot.

[0789] Forms of feedback and system improvement

[0790] When users receive a warning from the system, they can review their driving behavior and strive for safer driving. If they receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback is given in the form of a survey via a smartphone app. For example, questions such as "Was the warning helpful?" and "Has it changed your driving behavior?" are displayed.

[0791] The feedback data is sent to a server and used to retrain the generative AI model. The server retrains the model based on new data, improving the accuracy of the system. This process is repeated periodically to ensure continuous improvement of the system.

[0792] Examples of specific examples and prompts

[0793] 1. Data Collection Example

[0794] The server obtains real-time road congestion data for Tokyo at 5 p.m. from the traffic management center.

[0795] The device collects driving data from the driver's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to a server.

[0796] 2. Data Analysis Example

[0797] The server checks that a particular intersection in Tokyo has had many accidents over the past three years.

[0798] The analysis results show that the risk level is high because the particular driver frequently brakes suddenly during evening traffic jams.

[0799] 3. Examples of warnings

[0800] Just before the user (driver) approaches a specific intersection, the device issues a voice alert saying, "There is a high-accident area ahead. Please slow down and be careful."

[0801] At the same time, a warning light on the dashboard will flash to alert you.

[0802] Example prompt sentence:

[0803] "Please explain how you collect and analyze driver behavior data and nationwide road condition data to predict the risk of traffic accidents, and how you issue warnings in real time."

[0804] The above is a specific embodiment for carrying out the present invention. This system enables drivers to accurately predict risks based on their individual driving characteristics and nationwide road conditions, and provides real-time warnings to prevent the risk of traffic accidents.

[0805] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0806] Step 1:

[0807] The server periodically obtains road condition data from across the country. As input, it receives data in JSON format from API endpoints provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. As output, the received data is stored in the server's database.

[0808] Specifically, the server periodically sends an API request like this:

[0809] GET / road_conditions?region=tokyo&datetime=2023-10-12T17:00:00

[0810] GET / weather?region=tokyo&datetime=2023-10-12T17:00:00

[0811] GET / accident_history?region=tokyo&datetime=2023-10-12

[0812] This allows the latest road condition data to be acquired.

[0813] Step 2:

[0814] The device collects the driver's driving data using on-board sensors and a smartphone app and transmits it to a server in real time. The inputs include speed, number of sudden braking attempts, driving time, location information, etc. As an output, the collected data is transferred to the server and added to a database.

[0815] Specifically, the device app periodically sends sensor data to the server in the background:

[0816] POST / upload_driving_data

[0817] Content-Type: application / json

[0818] {

[0819] "speed": 60,

[0820] "brake_count": 3,

[0821] "driving_time": "30 minutes",

[0822] "location": "35.6895, 139.6917"

[0823] }

[0824] Step 3:

[0825] The server preprocesses the received data. The input is the road condition data and driving data collected in Step 1 and Step 2. This data may contain missing values ​​and outliers, so these are imputed and removed. The output is the preprocessed, clean data.

[0826] Specifically, the server runs a Python script that uses the Pandas library to clean the data:

[0827] python

[0828] import pandas as pd

[0829] Data loading

[0830] data = pd.read_json("driving_data.json")

[0831] Missing value imputation

[0832] data['speed'].fillna(data['speed'].mean(), inplace=True)

[0833] Outlier removal

[0834] data = data[data['speed'] < 150]

[0835] Step 4:

[0836] The server inputs the preprocessed data into the generative AI model to perform risk assessment. The input includes preprocessed road condition data and driving data. The generative AI model integrates and analyzes this data to assess the risk level of traffic accidents. The output is a risk score and specific warning content.

[0837] Specifically, the server runs models built using TensorFlow or PyTorch:

[0838] python

[0839] import tensorflow as tf

[0840] Model Loading

[0841] model = tf.keras.models.load_model('risk_model.h5')

[0842] prediction

[0843] predictions = model.predict(preprocessed_data)

[0844] risk_score = predictions[0]

[0845] Step 5:

[0846] As a result of the risk assessment, the device issues a warning to the driver. The inputs include a risk score and warning content obtained from the generative AI model. The output is real-time notification to the driver in the form of audio alerts, visual alerts, and haptic feedback.

[0847] Specifically, the device will display a warning message on the car's display via Bluetooth:

[0848] python

[0849] Audio alerts

[0850] play_sound("warning_audio.mp3")

[0851] Visual Alerts

[0852] show_message("Accident-prone area ahead. Please slow down and take care.")

[0853] Haptic Feedback

[0854] vibrate_device()

[0855] Step 6:

[0856] Users receive warnings from the system and are encouraged to review their driving behavior and drive safely. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. Feedback data in the form of a questionnaire is included as input. The feedback data is sent to the server as output and used to improve the system.

[0857] Specifically, users answer a survey via a smartphone app:

[0858] POST / feedback

[0859] Content-Type: application / json

[0860] {

[0861] "useful": true,

[0862] "behavior_change": "Yes"

[0863] }

[0864] Through these processing steps, the system can use driver behavior data and nationwide road condition data to predict the risk of traffic accidents and issue warnings in real time.

[0865] (Application example 1)

[0866] 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."

[0867] There is a challenge to promote safe driving and prevent traffic accidents by providing a system that predicts the risk of traffic accidents in real time and issues effective warnings to drivers and autonomous vehicles. In addition, with conventional systems, it is difficult for autonomous vehicles to properly assess risks and automatically take appropriate control, so we would like to improve this aspect as well.

[0868] 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.

[0869] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for analyzing this data and assessing the risk level of traffic accidents, means for issuing a warning to the driver when the risk level increases, and means for intervening in the automatic control system of the autonomous vehicle, thereby enabling the driver and the autonomous driving system to receive risk information in real time and take appropriate measures.

[0870] "Driving data" refers to information relating to the driver's driving behavior, including data such as speed, frequency of sudden braking, and driving time.

[0871] "Road condition data" is information about roads across the country, including congestion levels, weather, accident history, and so on.

[0872] "Analysis" refers to the process of processing and analyzing collected data to assess the risk of traffic accidents.

[0873] The "risk level" is an evaluation index that indicates the possibility of a traffic accident occurring.

[0874] "Warnings" are messages or signals that alert the driver when the risk level increases, and can take the form of audio alerts, visual alerts, or haptic feedback.

[0875] An "automatic control system" is a system that automatically controls the driving operations of an autonomous vehicle. It receives information on traffic risks and adjusts vehicle speed and changes course.

[0876] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and conduct risk assessments.

[0877] A "prompt sentence" is an instruction sentence used as input to a generative AI model.

[0878] This invention is a system that collects and analyzes driver's driving behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers and self-driving vehicles in real time. This system includes the following means.

[0879] Data collection and analysis

[0880] The server periodically collects nationwide road condition data. This data includes congestion levels, weather, and accident history provided by the Traffic Management Center and the Japan Meteorological Agency. The device also collects driver driving data using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, and more. The collected data is sent to the server in real time.

[0881] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[0882] Warning

[0883] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver or the autonomous vehicle's control system. The warning can be issued in multiple ways, primarily via audio alerts, visual alerts, or haptic feedback, or a combination of these. For example, a voice prompt saying, "Be careful. Accident hotspot ahead," may be played, along with a warning message displayed on the dashboard. The device may also instruct the autonomous vehicle to adjust speed or change course based on pre-set prompts.

[0884] Feedback and continuous improvement of the system

[0885] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[0886] Specific examples

[0887] Data Collection Examples

[0888] The server obtains real-time data on road congestion in urban areas at 5 p.m. on a given day from the traffic management center.

[0889] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[0890] Data analysis examples

[0891] The server checks that a particular intersection has had a high number of accidents over the past three years.

[0892] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[0893] Examples of warnings issued

[0894] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0895] At the same time, a warning light on the dashboard will flash to alert you.

[0896] In the case of self-driving vehicles, the terminal sends instructions to the automatic control system to adjust speed and change course.

[0897] Example prompts to input to the generative AI model

[0898] Enter your driving and road condition data to assess your risk of traffic accidents.

[0899] Driving data: speed, frequency of hard braking, driving time

[0900] Road condition data: congestion, weather, accident history

[0901] This invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about traffic accident risks. This enables not only drivers but also automated vehicles to take appropriate action in situations with a high risk of accidents.

[0902] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0903] Step 1:

[0904] The server acquires road condition data from across the country. Specifically, it periodically collects data such as congestion levels, weather, and accident history from public institutions such as traffic management centers and the Japan Meteorological Agency via API. The acquired data is in formats such as JSON or XML, which it then parses appropriately and stores in an internal data store. The input for this step is road condition data from the external API, and the output is internal data ready for analysis.

[0905] Step 2:

[0906] The terminal collects the driver's driving data. This is done using on-board sensors and a smartphone app. Collected data items include speed, frequency of sudden braking, and driving time. The collected data is sent to the server in real time. The input for this step is driving data from the on-board sensors and smartphone, and the output is the data before analysis that is sent to the server.

[0907] Step 3:

[0908] The server integrates the received driving data and road condition data and performs preprocessing. Specifically, it complements missing values ​​in the data and removes abnormal values ​​(outliers). The preprocessed data is then standardized and normalized, and converted into a format suitable for the generative AI model. The input for this step is the driving data and road condition data, and the output is the preprocessed data.

[0909] Step 4:

[0910] The server inputs preprocessed data into a generative AI model to evaluate the risk level of a traffic accident. This model uses a machine learning algorithm to perform risk assessment based on the preprocessed data. As a specific example, the risk assessment algorithm calculates a risk score using a neural network. If the risk score exceeds a certain threshold, the risk is determined to be high. The input to this step is preprocessed data, and the output is the risk level assessment result (risk score).

[0911] Step 5:

[0912] The device issues a warning to the driver based on the risk level evaluation results. Warning methods include audio alerts, visual alerts, and haptic feedback, and may be a combination of these. For example, the driver may hear a voice prompt saying, "There is an accident-prone area ahead. Please be careful," and a warning message may be displayed on the dashboard. The input of this step is the risk level evaluation result, and the output is a warning to the driver.

[0913] Step 6:

[0914] The terminal intervenes in the automated control system of the autonomous vehicle and instructs it to take appropriate action. Specifically, if it determines that the risk is high, it sends instructions to adjust the vehicle's speed or change course. The input to this step is the risk level assessment result, and the output is an operational instruction to the automated control system.

[0915] Step 7:

[0916] After receiving the warning, the user reviews their driving behavior and strives for safe driving. They also provide information about the effectiveness of the warning and areas for improvement to the system through the feedback function. This feedback data is sent to the server and used to retrain the generative AI model. The input of this step is user feedback, and the output is data used for retraining.

[0917] Step 8:

[0918] The server receives feedback data from users and retrains the generative AI model. This allows the system to continuously improve its accuracy and enable more effective risk assessment and warning. The input of this step is the feedback data from users, and the output is the retrained generative AI model.

[0919] 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.

[0920] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. Below, we will explain in detail the program processing of this system in natural language.

[0921] Data collection and analysis program

[0922] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, and more. Meanwhile, the terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. The collected data is sent to the server in real time.

[0923] The server preprocesses the received data, converting it into an appropriate format, imputing missing values, and detecting and removing outliers, making the data suitable for analysis.

[0924] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[0925] Real-time alert program

[0926] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, and conversely, if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[0927] Feedback and continuous improvement of the system

[0928] The user is encouraged to review their driving behavior and strive for safe driving after receiving the warning. If the user receives multiple warnings, the system provides feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device.

[0929] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0930] Specific examples

[0931] 1. Data Collection Examples

[0932] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[0933] The device collects driving data from the past week from Driver A's smartphone and emotional data (e.g., stress levels and changes in facial expressions) obtained from a facial recognition camera, and sends the data to a server.

[0934] 2. Data analysis examples

[0935] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[0936] The emotion engine assesses Driver A's stress level as "high," and the generative AI model determines the risk level as "high."

[0937] 3. Examples of warnings

[0938] Just before the user, Driver A, approaches that particular intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[0939] At the same time, a warning light on the dashboard will flash to alert you.

[0940] In this way, the present invention predicts and warns about the risk of traffic accidents by analyzing the individual driving characteristics of drivers and nationwide road conditions in real time, and also taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents from occurring.

[0941] The processing flow will be explained below.

[0942] Step 1: Data collection

[0943] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[0944] The device uses in-vehicle sensors, a smartphone app, and emotion recognition devices (e.g., cameras and microphones) to collect the driver's driving and emotional data, including speed, frequency of sudden braking, driving time, facial expressions, tone of voice, etc. The collected data is sent to a server in real time.

[0945] Step 2: Data Preprocessing

[0946] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, fills in missing values, and detects and removes outliers. It also organizes the data appropriately as time-series data. Meanwhile, it also performs format conversion and preprocessing on emotion data.

[0947] Step 3: Data analysis

[0948] The server inputs the preprocessed data into the generative AI model and emotion engine. The generative AI model learns from a large amount of past data and integrates road condition data and driving data for analysis. The emotion engine analyzes the driver's emotional state from their facial expressions and voice and feeds the results back to the generative AI model. For example, if the driver is under stress, the engine will reflect this and adjust the risk assessment.

[0949] Step 4: Risk assessment

[0950] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model and emotion engine. The evaluation results are classified into categories such as "low risk," "medium risk," and "high risk." If the emotion data indicates high stress, the risk level is adjusted higher.

[0951] Step 5: Send an alert

[0952] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used saying, "There is an accident hotspot ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0953] Step 6: Adjust your driving behavior

[0954] Users are encouraged to review their driving behavior and focus on safe driving by receiving the warning. This includes slowing down and paying attention to their surroundings. The system also includes voice guidance and changes to the alert content to help drivers relax.

[0955] Step 7: Gather feedback

[0956] If the user receives multiple warnings, they are asked to provide feedback on the effectiveness of the warnings and areas for improvement. This feedback includes information on the impact of the warnings, changes in driving behavior, and changes in emotions. The feedback data is sent to the server via the device.

[0957] Step 8: Update the Model

[0958] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0959] Example 2

[0960] 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."

[0961] Conventional driver assistance systems not only collect and analyze driver data and nationwide road condition data in real time, but also have the problem of being unable to perform risk assessments that take into account the driver's emotional state. Furthermore, data pre-processing during risk assessments is often insufficient, potentially reducing the accuracy of the analysis. Furthermore, warning methods for increased risk are limited, making them unable to adapt to a variety of situations.

[0962] 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.

[0963] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for preprocessing this data and converting it into an appropriate format, means for completing missing values ​​and detecting and removing outliers, means for inputting the preprocessed data into a generative AI model to evaluate the risk level of a traffic accident, means for collecting driver emotion data and reflecting it in risk assessment, and means for issuing a warning to the driver when the risk level increases. This enables highly accurate risk assessment and diverse warning methods that take into account the driver's emotional state while improving data accuracy.

[0964] "Driving data" refers to data related to the driver's driving behavior, specifically data including speed, frequency of sudden braking, driving time, and the like.

[0965] "Road condition data" refers to data on traffic conditions across the country, including congestion levels, weather, accident history, and the like.

[0966] "Preprocessing" is the process of converting collected data into a format suitable for analysis, and is a procedure for filling in missing values ​​and detecting and removing outliers.

[0967] A "generative AI model" is an algorithm or software that uses machine learning technology to learn from large amounts of past data and perform risk assessments based on new data.

[0968] "Risk level" is an index that evaluates the probability and degree of danger of a traffic accident, and is a number calculated using a generative AI model.

[0969] "Emotion data" refers to data relating to emotions acquired from the driver's facial expressions and voice, and includes stress levels, emotional states, and the like.

[0970] A "warning" is a notification to a driver that the risk of a traffic accident has increased, and may take the form of an audio alert, visual alert, or haptic feedback.

[0971] "Feedback" is information provided by the driver about changes in driving behavior and the impact of warnings, and is data that is sent to the server and used to improve the system.

[0972] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. The implementation method of this system will be described in detail below.

[0973] Data collection methods

[0974] The device collects the driver's driving data using on-board sensors and smartphones. This driving data includes speed, frequency of sudden braking, and driving time. It also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. This data is sent to a server in real time. As a specific example of collection, speed data is obtained using the smartphone's GPS function, and the frequency of sudden braking is detected using on-board sensors. Facial expression data is collected using a camera, and the driver's voice is recorded using a microphone, and voice tone analysis is performed.

[0975] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, etc. As a specific example, it obtains real-time road congestion data from the Traffic Management Center, receives current and forecast weather information from the Japan Meteorological Agency, and downloads past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0976] Data preprocessing and analysis

[0977] The server preprocesses the received data and converts it into an appropriate format. First, it performs missing value imputation and outlier detection and removal to prepare the data for analysis. For example, preprocessing involves generating a histogram of the velocity data and identifying and removing anomalous data points.

[0978] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model uses an algorithm trained on a large amount of past data. The model integrates driving data, road condition data, and driver emotion data to calculate a risk score. For example, if the driver is experiencing high stress, the model will assess the risk level as high.

[0979] Real-time alerts and feedback

[0980] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are multiple ways to warn the driver, including providing audio alerts, visual alerts, and haptic feedback depending on the driver's emotional state. For example, if the driver is nervous, a gentle audio alert will be played saying, "There is a high-accident zone ahead. Please be careful," and a warning message will be displayed on the dashboard.

[0981] The user receives the warning and is encouraged to review their driving behavior and drive safely. After receiving multiple warnings, the user is also given feedback on the impact of the warning and areas for improvement. The feedback data includes information on the impact of the warning and changes in driving behavior, and is sent to the server via the device.

[0982] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[0983] Examples and prompts

[0984] As an example of data collection, a server obtains road congestion data for Tokyo in real time from a traffic management center at 5:00 PM on a certain day. A device also collects driving data from Driver A's smartphone for the past week and emotion data obtained from a facial recognition camera, and sends these data to the server. Specific examples of prompts include a voice alert such as "There have been many accidents at the next intersection over the past three years. Please be careful," or a visual message such as "Be careful when making the next left turn."

[0985] In this way, the present invention provides a system that predicts and warns about traffic accident risks by analyzing individual driver characteristics and nationwide road conditions in real time, taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents.

[0986] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0987] Step 1: Data collection

[0988] The device collects driving data from the driver using on-board sensors and smartphones. Specifically, it obtains speed data from GPS and the frequency of sudden braking from the brake pedal sensor. It also detects facial expressions with a camera and records voice data with a microphone for emotion recognition. This data is sent to a server in real time.

[0989] Input: Driving data (speed, frequency of sudden braking, driving time), emotional data (facial expressions, voice)

[0990] Output: Driving and emotion data sent to a server in real time

[0991] Step 2: External Data Acquisition

[0992] The server receives nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Specifically, it receives road congestion data from the Traffic Management Center, weather information from the Japan Meteorological Agency, and past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[0993] Input: Road condition data (congestion, weather, accident history)

[0994] Output: Road condition data stored on the server

[0995] Step 3: Data Preprocessing

[0996] The server preprocesses the received driving data, emotion data, and road condition data, including data format conversion, missing value imputation, and outlier removal. Specifically, it removes outliers in speed data and imputes missing emotion data using statistical methods.

[0997] Input: Received driving data, emotion data, road condition data

[0998] Output: Preprocessed data

[0999] Step 4: Risk assessment

[1000] The server inputs the preprocessed data into a generative AI model to assess traffic accident risk. The generative AI model calculates a risk score by integrating driving data, road condition data, and emotional data. For example, if the driver is experiencing high stress, the model will assess the risk level higher.

[1001] Input: Preprocessed data

[1002] Output: Risk score

[1003] Step 5: Send an alert

[1004] If the risk score exceeds a certain threshold as a result of the risk assessment, the device will issue a real-time warning to the driver. The warning method is a voice alert, playing a gentle message saying, "There is a high-accident zone ahead. Please be careful." The visual alert is a warning message displayed on the dashboard.

[1005] Input: Risk Score

[1006] Output: Real-time warning (audio alert, visual alert)

[1007] Step 6: Gather feedback

[1008] After receiving the warning, the user reviews their driving behavior and provides feedback, including the impact of the warning and changes in their driving behavior, which is sent to the server via the device.

[1009] Input: Feedback data on driving behavior revision

[1010] Output: Feedback data sent to the server

[1011] Step 7: Update the model

[1012] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, thereby improving the accuracy of risk assessment. Once retrained, the generative AI model is applied to the system and used for subsequent risk assessments.

[1013] Input: Feedback data

[1014] Output: An updated generative AI model

[1015] Through the above steps, the system of the present invention takes into account the driver's driving characteristics, national road conditions, and the driver's emotional state, predicts traffic accident risks, and issues warnings, thereby promoting safe driving.

[1016] (Application example 2)

[1017] 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."

[1018] Conventional traffic accident prevention systems assessed risk using only driver driving data and road condition data, making it impossible to accurately assess risk by taking the driver's emotional state into account. Furthermore, the warning format was uniform, making it difficult to issue a warning appropriate to the driver's emotional state. Therefore, improving the effectiveness of traffic accident risk prediction and warnings remains a challenge.

[1019] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1020] In this invention, the server includes a means for collecting driving data and emotional data of the driver, a means for acquiring nationwide road condition data and weather data, a means for analyzing these data and using a generative AI model to evaluate the risk level of a traffic accident, and a means for issuing an audio or visual alert according to the emotional state of the driver when the risk level increases. This makes it possible to integrate the driving data and emotional data of the driver to evaluate the risk of a traffic accident with high accuracy and to issue a warning appropriate to the emotional state of the driver.

[1021] "Driving data" refers to information such as the driver's speed, frequency of sudden braking, and driving time.

[1022] "Emotion data" refers to information about the emotional state of the driver as recognized from their facial expressions and voice.

[1023] "Road condition data" refers to information such as road congestion, weather, and accident history obtained from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[1024] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of past data to assess the risk of traffic accidents.

[1025] "Audio alert" refers to a means of issuing an audio warning to the driver.

[1026] "Visual alert" refers to a means of visually displaying a warning to the driver.

[1027] "Haptic feedback" refers to a means of providing tactile feedback such as vibration to the driver.

[1028] "Risk level" refers to an index that indicates the risk of a traffic accident occurring.

[1029] "Driving behavior patterns" refers to information that indicates a driver's driving habits and characteristic driving behavior.

[1030] The present invention is a system that collects driving data and emotional data of drivers and evaluates and warns about traffic accident risks in real time. An embodiment of this system will be described below with specific examples.

[1031] Data collection and analysis program

[1032] The server obtains road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Road condition data includes congestion levels, weather, and accident history. The terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice.

[1033] The collected data is sent to a server in real time. The server preprocesses the received data, converting it into an appropriate format, and also complements missing values ​​and removes outliers. The preprocessed data is then input into a generative AI model to assess traffic accident risk. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[1034] Real-time alert program

[1035] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, while if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[1036] Feedback and continuous improvement of the system

[1037] Upon receiving the warning, the user will review their driving behavior and strive for safer driving. If the user receives multiple warnings, the system will provide feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device. The server analyzes the received feedback data and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and enables the system to provide more effective warnings.

[1038] Specific examples

[1039] 1. Data Collection Examples

[1040] The server obtains real-time data on road congestion in Tokyo at 5:00 PM on a given day from the traffic management center. The device collects driving data from the past week from the driver's smartphone and emotional data obtained from a facial recognition camera (e.g., stress level and changes in facial expression), and sends the data to the server.

[1041] 2. Data analysis examples

[1042] The server checks that a particular intersection in Tokyo has had a high number of accidents over the past three years. The emotion engine evaluates the driver's stress level as "high," and the generative AI model determines the risk level as "high."

[1043] 3. Examples of warnings

[1044] Just before the user (driver) approaches a specific intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful." At the same time, a warning light on the dashboard flashes to draw attention.

[1045] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[1046] As described above, this invention analyzes the individual driving characteristics of drivers and nationwide road conditions in real time, and also takes into account the emotional state of the driver, thereby predicting and warning about the risk of traffic accidents, thereby promoting safe driving and preventing traffic accidents from occurring.

[1047] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1048] Step 1:

[1049] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, and accident history. As input, it receives real-time road condition information from each data source, and as output, it integrates and stores this data. Specifically, it sends API requests, analyzes the response data, and extracts the necessary information.

[1050] Step 2:

[1051] The device collects driving data from the driver using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, etc. It also uses a camera and microphone to collect emotional data from the driver's facial expressions and voice. As input, it receives real-time driving data and emotional data from the sensors and devices, and as output, it sends this data to a server. Specifically, it acquires the data stream from the device, unifies the format, and transmits it.

[1052] Step 3:

[1053] The server preprocesses the received driving data and emotion data. This preprocessing includes converting the data format, filling in missing values, and removing outliers. It receives the driving data and emotion data sent from the device as input, and generates preprocessed data as output. Specifically, it stores the data in a database while processing missing values ​​and outliers within the program.

[1054] Step 4:

[1055] The server inputs the preprocessed data into a generative AI model to evaluate the risk of traffic accidents. The generative AI model learns from a huge amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It receives preprocessed driving data and emotion data as input, and generates risk assessment results as output. Specifically, it inputs data into the generative AI model, analyzes the prediction results, and calculates the risk level.

[1056] Step 5:

[1057] The server also reflects the driver's emotional data in the generative AI model and determines the risk level based on the risk assessment results. It receives the emotional assessment data from the emotion engine and the risk assessment results from the generative AI model as input, and generates the final risk level as output. Specifically, it integrates the emotional assessment data with the output of the generative AI model and modifies the risk level.

[1058] Step 6:

[1059] If the risk assessment determines that there is a high risk of a traffic accident, the device issues a real-time warning to the driver. The format of the warning is adjusted according to the driver's emotional state. As input, the device receives the risk assessment results and emotional evaluation data from the server, and as output, issues an appropriate warning to the driver. Specifically, the device selects and executes the method of issuing an audio or visual alert based on the driver's emotional state.

[1060] Step 7:

[1061] When the user receives the warning, they will review their driving behavior and strive for safe driving. If they receive multiple warnings, they will be given feedback on the effectiveness of the warnings and areas for improvement. The system receives driver feedback data as input and sends this data to the server as output. Specific operations include providing an interface for collecting feedback and sending the input information to the server.

[1062] Step 8:

[1063] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and provides more effective warnings. It receives user feedback and emotion data as input and creates improved versions of the generative AI model and emotion engine as output. Specifically, it uses the collected data to retrain the algorithm and update and apply the new model.

[1064] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[1065] 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.

[1066] 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.

[1067] 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.

[1068] [Fourth embodiment]

[1069] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1070] 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.

[1071] 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).

[1072] 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.

[1073] 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.

[1074] 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).

[1075] 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.

[1076] 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.

[1077] 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.

[1078] 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.

[1079] 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.

[1080] 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.

[1081] 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."

[1082] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. Below, we will explain in detail the program processing of this system in natural language.

[1083] Data collection and analysis program

[1084] The server first periodically acquires nationwide road condition data. This data includes congestion levels, weather, accident history, and other information provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, and other organizations. Meanwhile, the device uses on-board sensors and a smartphone app to collect driver data. This data includes speed, frequency of sudden braking, driving time, and other information. The collected data is then sent to the server in real time.

[1085] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[1086] Warning program

[1087] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are several ways to issue the warning, mainly audio alerts, visual alerts, haptic feedback, or a combination of these. For example, a voice prompt saying, "There is a high-accident zone ahead. Please be careful," will be played, and a warning message will be displayed on the dashboard.

[1088] Feedback and continuous improvement of the system

[1089] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[1090] Specific examples

[1091] 1. Data Collection Examples

[1092] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[1093] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[1094] 2. Data analysis examples

[1095] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[1096] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[1097] 3. Examples of warnings

[1098] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[1099] At the same time, a warning light on the dashboard will flash to alert you.

[1100] In this way, this invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about the risk of traffic accidents.

[1101] The processing flow will be explained below.

[1102] Step 1: Data collection

[1103] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[1104] The device collects the driver's driving data (speed, frequency of sudden braking, driving time, etc.) through on-board sensors and a smartphone app and transmits it to a server in real time.

[1105] Step 2: Data Preprocessing

[1106] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, complements missing values, detects and removes outliers, and organizes the data appropriately as time series data.

[1107] Step 3: Data analysis

[1108] The server inputs the preprocessed data into a generative AI model. The generative AI model learns from a large amount of past data and integrates and analyzes road condition data and driving data. Specifically, it evaluates the current driving situation based on factors such as traffic volume, weather conditions, and past accident history, and calculates the risk level of a traffic accident.

[1109] Step 4: Risk assessment

[1110] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model, which are then classified into categories such as "low risk," "medium risk," and "high risk."

[1111] Step 5: Send an alert

[1112] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is given by audio alert, visual alert, haptic feedback, or a combination of these. For example, a voice prompt saying "There is a high-accident zone ahead. Please be careful" is played, and a warning message is displayed on the dashboard at the same time.

[1113] Step 6: Adjust your driving behavior

[1114] The user receives a warning and is encouraged to review their driving behavior and take steps to drive safely, such as slowing down and paying attention to their surroundings.

[1115] Step 7: Gather feedback

[1116] If the user receives multiple warnings, they provide feedback on the effectiveness and areas for improvement. This feedback includes information on changes in driving behavior and the effectiveness of the system. The feedback data is sent to the server via the device.

[1117] Step 8: Update the Model

[1118] The server analyzes the received feedback data and uses it to retrain the generative AI model, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[1119] Example 1

[1120] 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."

[1121] In order to reduce the risk of traffic accidents, it is essential to provide drivers with accurate warnings in real time. However, current systems have difficulty effectively integrating data on road conditions and driver behavior nationwide to make highly accurate risk predictions. To address this issue, there is a need for a system that can accurately predict the risk of traffic accidents and issue warnings to drivers in real time.

[1122] 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.

[1123] In this invention, the server includes a means for acquiring nationwide road condition data, a means for collecting driver driving data, a means for preprocessing the acquired and collected data, a means including a generative AI model for analyzing the preprocessed data to evaluate the risk level of a traffic accident, and a means for issuing a warning to the driver in real time when the risk level increases. This enables highly accurate risk prediction based on individual driving characteristics and nationwide road conditions, and allows drivers to receive real-time warnings to prevent the risk of a traffic accident.

[1124] "National road condition data" refers to data including road congestion, weather conditions, and past accident history obtained from data providers such as the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[1125] "Driver driving data" refers to data related to driving behavior, such as vehicle speed, number of sudden braking attempts, driving time, and location information, collected using in-vehicle sensors and smartphone apps.

[1126] "Preprocessing" refers to the process of filling in missing values ​​in collected data, removing outliers, and converting the data into a format suitable for analysis.

[1127] A "generative AI model" is an artificial intelligence model built on a large dataset, and is a machine learning or deep learning model used to assess the risk of traffic accidents.

[1128] A "warning" is a real-time notification to the driver when the risk of a traffic accident increases, and may be provided via audio alerts, visual alerts, and / or haptic feedback.

[1129] This invention relates to a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time. This system is mainly composed of a server, a terminal, and a user, and specific embodiments are described below.

[1130] Data collection methods

[1131] The server periodically obtains nationwide road condition data provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. To obtain this data, it receives data in JSON format from multiple API endpoints.

[1132] Meanwhile, the device collects driving data from the driver using on-board sensors and a smartphone app. Collected data includes the vehicle's speed, number of sudden braking attempts, driving time, and location information. This data is sent to a server in real time. The smartphone app on the device periodically collects data from the sensors in the background and transfers it to the server every few seconds.

[1133] Data analysis format

[1134] The server preprocesses the received data. This preprocessing involves filling in missing values, removing outliers, and converting the data into a format suitable for analysis. Specifically, data cleaning is performed using a Python program and the Pandas library. Missing values ​​are filled in using the average of the preceding and following data, and abnormal values ​​(e.g., speed data that suddenly reaches several hundred km / h) are removed as outliers.

[1135] The preprocessed data is then fed into a generative AI model to assess the risk level of a traffic accident. This generative AI model is a deep learning model built using TensorFlow or PyTorch and trained on a large dataset. The model's output includes a risk score and specific warnings.

[1136] Warning issuance format

[1137] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. This warning can be provided by audio alerts, visual alerts, and / or haptic feedback. For example, a voice prompt such as "There is a high-accident zone ahead. Please be careful" may be played from the smartphone, while a warning message is simultaneously displayed on the in-car display.

[1138] The device uses push notifications or Bluetooth connectivity to display information on the in-car display or smartphone, allowing the driver to quickly recognize important information on the spot.

[1139] Forms of feedback and system improvement

[1140] When users receive a warning from the system, they can review their driving behavior and strive for safer driving. If they receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback is given in the form of a survey via a smartphone app. For example, questions such as "Was the warning helpful?" and "Has it changed your driving behavior?" are displayed.

[1141] The feedback data is sent to a server and used to retrain the generative AI model. The server retrains the model based on new data, improving the accuracy of the system. This process is repeated periodically to ensure continuous improvement of the system.

[1142] Examples of specific examples and prompts

[1143] 1. Data Collection Example

[1144] The server obtains real-time road congestion data for Tokyo at 5 p.m. from the traffic management center.

[1145] The device collects driving data from the driver's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to a server.

[1146] 2. Data Analysis Example

[1147] The server checks that a particular intersection in Tokyo has had many accidents over the past three years.

[1148] The analysis results show that the risk level is high because the particular driver frequently brakes suddenly during evening traffic jams.

[1149] 3. Examples of warnings

[1150] Just before the user (driver) approaches a specific intersection, the device issues a voice alert saying, "There is a high-accident area ahead. Please slow down and be careful."

[1151] At the same time, a warning light on the dashboard will flash to alert you.

[1152] Example prompt sentence:

[1153] "Please explain how you collect and analyze driver behavior data and nationwide road condition data to predict the risk of traffic accidents, and how you issue warnings in real time."

[1154] The above is a specific embodiment for carrying out the present invention. This system enables drivers to accurately predict risks based on their individual driving characteristics and nationwide road conditions, and provides real-time warnings to prevent the risk of traffic accidents.

[1155] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1156] Step 1:

[1157] The server periodically obtains road condition data from across the country. As input, it receives data in JSON format from API endpoints provided by the Traffic Management Center, the Japan Meteorological Agency, the Ministry of Land, Infrastructure, Transport and Tourism, etc. The data includes road congestion status, weather conditions, and past accident history. As output, the received data is stored in the server's database.

[1158] Specifically, the server periodically sends an API request like this:

[1159] GET / road_conditions?region=tokyo&datetime=2023-10-12T17:00:00

[1160] GET / weather?region=tokyo&datetime=2023-10-12T17:00:00

[1161] GET / accident_history?region=tokyo&datetime=2023-10-12

[1162] This allows the latest road condition data to be acquired.

[1163] Step 2:

[1164] The device collects the driver's driving data using on-board sensors and a smartphone app and transmits it to a server in real time. The inputs include speed, number of sudden braking attempts, driving time, location information, etc. As an output, the collected data is transferred to the server and added to a database.

[1165] Specifically, the device app periodically sends sensor data to the server in the background:

[1166] POST / upload_driving_data

[1167] Content-Type: application / json

[1168] {

[1169] "speed": 60,

[1170] "brake_count": 3,

[1171] "driving_time": "30 minutes",

[1172] "location": "35.6895, 139.6917"

[1173] }

[1174] Step 3:

[1175] The server preprocesses the received data. The input is the road condition data and driving data collected in Step 1 and Step 2. This data may contain missing values ​​and outliers, so these are imputed and removed. The output is the preprocessed, clean data.

[1176] Specifically, the server runs a Python script that uses the Pandas library to clean the data:

[1177] python

[1178] import pandas as pd

[1179] Data loading

[1180] data = pd.read_json("driving_data.json")

[1181] Missing value imputation

[1182] data['speed'].fillna(data['speed'].mean(), inplace=True)

[1183] Outlier removal

[1184] data = data[data['speed'] < 150]

[1185] Step 4:

[1186] The server inputs the preprocessed data into the generative AI model to perform risk assessment. The input includes preprocessed road condition data and driving data. The generative AI model integrates and analyzes this data to assess the risk level of traffic accidents. The output is a risk score and specific warning content.

[1187] Specifically, the server runs models built using TensorFlow or PyTorch:

[1188] python

[1189] import tensorflow as tf

[1190] Model Loading

[1191] model = tf.keras.models.load_model('risk_model.h5')

[1192] prediction

[1193] predictions = model.predict(preprocessed_data)

[1194] risk_score = predictions[0]

[1195] Step 5:

[1196] As a result of the risk assessment, the device issues a warning to the driver. The inputs include a risk score and warning content obtained from the generative AI model. The output is real-time notification to the driver in the form of audio alerts, visual alerts, and haptic feedback.

[1197] Specifically, the device will display a warning message on the car's display via Bluetooth:

[1198] python

[1199] Audio alerts

[1200] play_sound("warning_audio.mp3")

[1201] Visual Alerts

[1202] show_message("Accident-prone area ahead. Please slow down and take care.")

[1203] Haptic Feedback

[1204] vibrate_device()

[1205] Step 6:

[1206] Users receive warnings from the system and are encouraged to review their driving behavior and drive safely. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. Feedback data in the form of a questionnaire is included as input. The feedback data is sent to the server as output and used to improve the system.

[1207] Specifically, users answer a survey via a smartphone app:

[1208] POST / feedback

[1209] Content-Type: application / json

[1210] {

[1211] "useful": true,

[1212] "behavior_change": "Yes"

[1213] }

[1214] Through these processing steps, the system can use driver behavior data and nationwide road condition data to predict the risk of traffic accidents and issue warnings in real time.

[1215] (Application example 1)

[1216] 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."

[1217] There is a challenge to promote safe driving and prevent traffic accidents by providing a system that predicts the risk of traffic accidents in real time and issues effective warnings to drivers and autonomous vehicles. In addition, with conventional systems, it is difficult for autonomous vehicles to properly assess risks and automatically take appropriate control, so we would like to improve this aspect as well.

[1218] 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.

[1219] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for analyzing this data and assessing the risk level of traffic accidents, means for issuing a warning to the driver when the risk level increases, and means for intervening in the automatic control system of the autonomous vehicle, thereby enabling the driver and the autonomous driving system to receive risk information in real time and take appropriate measures.

[1220] "Driving data" refers to information relating to the driver's driving behavior, including data such as speed, frequency of sudden braking, and driving time.

[1221] "Road condition data" is information about roads across the country, including congestion levels, weather, accident history, and so on.

[1222] "Analysis" refers to the process of processing and analyzing collected data to assess the risk of traffic accidents.

[1223] The "risk level" is an evaluation index that indicates the possibility of a traffic accident occurring.

[1224] "Warnings" are messages or signals that alert the driver when the risk level increases, and can take the form of audio alerts, visual alerts, or haptic feedback.

[1225] An "automatic control system" is a system that automatically controls the driving operations of an autonomous vehicle. It receives information on traffic risks and adjusts vehicle speed and changes course.

[1226] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and conduct risk assessments.

[1227] A "prompt sentence" is an instruction sentence used as input to a generative AI model.

[1228] This invention is a system that collects and analyzes driver's driving behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers and self-driving vehicles in real time. This system includes the following means.

[1229] Data collection and analysis

[1230] The server periodically collects nationwide road condition data. This data includes congestion levels, weather, and accident history provided by the Traffic Management Center and the Japan Meteorological Agency. The device also collects driver driving data using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, and more. The collected data is sent to the server in real time.

[1231] The server then analyzes the received data. First, it preprocesses the received data and converts it into an appropriate format. This step also involves filling in missing values ​​and removing outliers. Next, the data is input into a generative AI model to perform risk assessment. At this time, highly accurate risk assessment is possible by integrating and analyzing road condition data and driving data.

[1232] Warning

[1233] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver or the autonomous vehicle's control system. The warning can be issued in multiple ways, primarily via audio alerts, visual alerts, or haptic feedback, or a combination of these. For example, a voice prompt saying, "Be careful. Accident hotspot ahead," may be played, along with a warning message displayed on the dashboard. The device may also instruct the autonomous vehicle to adjust speed or change course based on pre-set prompts.

[1234] Feedback and continuous improvement of the system

[1235] Users can receive warnings from the system and review their driving behavior to ensure safe driving. If users receive multiple warnings, they can provide feedback to the system about the effectiveness of the warnings and areas for improvement. This feedback data is sent to the server and used to retrain the generative AI model. This allows the system to continuously improve its accuracy, enabling more effective risk assessment and warnings.

[1236] Specific examples

[1237] Data Collection Examples

[1238] The server obtains real-time data on road congestion in urban areas at 5 p.m. on a given day from the traffic management center.

[1239] The device collects driving data from Driver A's smartphone over the past week (e.g., number of sudden braking attempts, frequency of speeding) and sends it to the server.

[1240] Data analysis examples

[1241] The server checks that a particular intersection has had a high number of accidents over the past three years.

[1242] As a result of the analysis, Driver A is assessed as having a high risk level because he frequently brakes suddenly during evening traffic jams.

[1243] Examples of warnings issued

[1244] Just before the user, Driver A, approaches that particular intersection, the device issues a voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[1245] At the same time, a warning light on the dashboard will flash to alert you.

[1246] In the case of self-driving vehicles, the terminal sends instructions to the automatic control system to adjust speed and change course.

[1247] Example prompts to input to the generative AI model

[1248] Enter your driving and road condition data to assess your risk of traffic accidents.

[1249] Driving data: speed, frequency of hard braking, driving time

[1250] Road condition data: congestion, weather, accident history

[1251] This invention is a system that promotes safe driving and prevents traffic accidents by analyzing individual driver characteristics and nationwide road conditions in real time and predicting and warning about traffic accident risks. This enables not only drivers but also automated vehicles to take appropriate action in situations with a high risk of accidents.

[1252] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1253] Step 1:

[1254] The server acquires road condition data from across the country. Specifically, it periodically collects data such as congestion levels, weather, and accident history from public institutions such as traffic management centers and the Japan Meteorological Agency via API. The acquired data is in formats such as JSON or XML, which it then parses appropriately and stores in an internal data store. The input for this step is road condition data from the external API, and the output is internal data ready for analysis.

[1255] Step 2:

[1256] The terminal collects the driver's driving data. This is done using on-board sensors and a smartphone app. Collected data items include speed, frequency of sudden braking, and driving time. The collected data is sent to the server in real time. The input for this step is driving data from the on-board sensors and smartphone, and the output is the data before analysis that is sent to the server.

[1257] Step 3:

[1258] The server integrates the received driving data and road condition data and performs preprocessing. Specifically, it complements missing values ​​in the data and removes abnormal values ​​(outliers). The preprocessed data is then standardized and normalized, and converted into a format suitable for the generative AI model. The input for this step is the driving data and road condition data, and the output is the preprocessed data.

[1259] Step 4:

[1260] The server inputs preprocessed data into a generative AI model to evaluate the risk level of a traffic accident. This model uses a machine learning algorithm to perform risk assessment based on the preprocessed data. As a specific example, the risk assessment algorithm calculates a risk score using a neural network. If the risk score exceeds a certain threshold, the risk is determined to be high. The input to this step is preprocessed data, and the output is the risk level assessment result (risk score).

[1261] Step 5:

[1262] The device issues a warning to the driver based on the risk level evaluation results. Warning methods include audio alerts, visual alerts, and haptic feedback, and may be a combination of these. For example, the driver may hear a voice prompt saying, "There is an accident-prone area ahead. Please be careful," and a warning message may be displayed on the dashboard. The input of this step is the risk level evaluation result, and the output is a warning to the driver.

[1263] Step 6:

[1264] The terminal intervenes in the automated control system of the autonomous vehicle and instructs it to take appropriate action. Specifically, if it determines that the risk is high, it sends instructions to adjust the vehicle's speed or change course. The input to this step is the risk level assessment result, and the output is an operational instruction to the automated control system.

[1265] Step 7:

[1266] After receiving the warning, the user reviews their driving behavior and strives for safe driving. They also provide information about the effectiveness of the warning and areas for improvement to the system through the feedback function. This feedback data is sent to the server and used to retrain the generative AI model. The input of this step is user feedback, and the output is data used for retraining.

[1267] Step 8:

[1268] The server receives feedback data from users and retrains the generative AI model. This allows the system to continuously improve its accuracy and enable more effective risk assessment and warning. The input of this step is the feedback data from users, and the output is the retrained generative AI model.

[1269] 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.

[1270] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. Below, we will explain in detail the program processing of this system in natural language.

[1271] Data collection and analysis program

[1272] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, and more. Meanwhile, the terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. The collected data is sent to the server in real time.

[1273] The server preprocesses the received data, converting it into an appropriate format, imputing missing values, and detecting and removing outliers, making the data suitable for analysis.

[1274] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[1275] Real-time alert program

[1276] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, and conversely, if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[1277] Feedback and continuous improvement of the system

[1278] The user is encouraged to review their driving behavior and strive for safe driving after receiving the warning. If the user receives multiple warnings, the system provides feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device.

[1279] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[1280] Specific examples

[1281] 1. Data Collection Examples

[1282] The server obtains real-time road congestion data for Tokyo at 5 p.m. on a given day from the traffic management center.

[1283] The device collects driving data from the past week from Driver A's smartphone and emotional data (e.g., stress levels and changes in facial expressions) obtained from a facial recognition camera, and sends the data to a server.

[1284] 2. Data analysis examples

[1285] The server checks that a particular intersection in Tokyo has seen a high number of accidents over the past three years.

[1286] The emotion engine assesses Driver A's stress level as "high," and the generative AI model determines the risk level as "high."

[1287] 3. Examples of warnings

[1288] Just before the user, Driver A, approaches that particular intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful."

[1289] At the same time, a warning light on the dashboard will flash to alert you.

[1290] In this way, the present invention predicts and warns about the risk of traffic accidents by analyzing the individual driving characteristics of drivers and nationwide road conditions in real time, and also taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents from occurring.

[1291] The processing flow will be explained below.

[1292] Step 1: Data collection

[1293] The server receives real-time data on road conditions across the country from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism, including congestion levels, weather, accident history, and more.

[1294] The device uses in-vehicle sensors, a smartphone app, and emotion recognition devices (e.g., cameras and microphones) to collect the driver's driving and emotional data, including speed, frequency of sudden braking, driving time, facial expressions, tone of voice, etc. The collected data is sent to a server in real time.

[1295] Step 2: Data Preprocessing

[1296] The server performs preprocessing on the received road condition data and driving data. Specifically, it converts the data format, fills in missing values, and detects and removes outliers. It also organizes the data appropriately as time-series data. Meanwhile, it also performs format conversion and preprocessing on emotion data.

[1297] Step 3: Data analysis

[1298] The server inputs the preprocessed data into the generative AI model and emotion engine. The generative AI model learns from a large amount of past data and integrates road condition data and driving data for analysis. The emotion engine analyzes the driver's emotional state from their facial expressions and voice and feeds the results back to the generative AI model. For example, if the driver is under stress, the engine will reflect this and adjust the risk assessment.

[1299] Step 4: Risk assessment

[1300] The server evaluates the driver's current risk level based on the analysis results obtained from the generative AI model and emotion engine. The evaluation results are classified into categories such as "low risk," "medium risk," and "high risk." If the emotion data indicates high stress, the risk level is adjusted higher.

[1301] Step 5: Send an alert

[1302] The device receives the risk assessment results sent from the server. If the risk level is assessed as "medium risk" or "high risk," the device issues a real-time warning to the driver. The warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used saying, "There is an accident hotspot ahead. Please be careful," and a warning message will be displayed on the dashboard.

[1303] Step 6: Adjust your driving behavior

[1304] Users are encouraged to review their driving behavior and focus on safe driving by receiving the warning. This includes slowing down and paying attention to their surroundings. The system also includes voice guidance and changes to the alert content to help drivers relax.

[1305] Step 7: Gather feedback

[1306] If the user receives multiple warnings, they are asked to provide feedback on the effectiveness of the warnings and areas for improvement. This feedback includes information on the impact of the warnings, changes in driving behavior, and changes in emotions. The feedback data is sent to the server via the device.

[1307] Step 8: Update the Model

[1308] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine, improving the accuracy of risk assessment and providing more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[1309] Example 2

[1310] 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."

[1311] Conventional driver assistance systems not only collect and analyze driver data and nationwide road condition data in real time, but also have the problem of being unable to perform risk assessments that take into account the driver's emotional state. Furthermore, data pre-processing during risk assessments is often insufficient, potentially reducing the accuracy of the analysis. Furthermore, warning methods for increased risk are limited, making them unable to adapt to a variety of situations.

[1312] 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.

[1313] In this invention, the server includes means for collecting driver driving data, means for acquiring nationwide road condition data, means for preprocessing this data and converting it into an appropriate format, means for completing missing values ​​and detecting and removing outliers, means for inputting the preprocessed data into a generative AI model to evaluate the risk level of a traffic accident, means for collecting driver emotion data and reflecting it in risk assessment, and means for issuing a warning to the driver when the risk level increases. This enables highly accurate risk assessment and diverse warning methods that take into account the driver's emotional state while improving data accuracy.

[1314] "Driving data" refers to data related to the driver's driving behavior, specifically data including speed, frequency of sudden braking, driving time, and the like.

[1315] "Road condition data" refers to data on traffic conditions across the country, including congestion levels, weather, accident history, and the like.

[1316] "Preprocessing" is the process of converting collected data into a format suitable for analysis, and is a procedure for filling in missing values ​​and detecting and removing outliers.

[1317] A "generative AI model" is an algorithm or software that uses machine learning technology to learn from large amounts of past data and perform risk assessments based on new data.

[1318] "Risk level" is an index that evaluates the probability and degree of danger of a traffic accident, and is a number calculated using a generative AI model.

[1319] "Emotion data" refers to data relating to emotions acquired from the driver's facial expressions and voice, and includes stress levels, emotional states, and the like.

[1320] A "warning" is a notification to a driver that the risk of a traffic accident has increased, and may take the form of an audio alert, visual alert, or haptic feedback.

[1321] "Feedback" is information provided by the driver about changes in driving behavior and the impact of warnings, and is data that is sent to the server and used to improve the system.

[1322] This invention combines a system that collects and analyzes driver behavior data and nationwide road condition data, predicts the risk of traffic accidents, and issues warnings to drivers in real time, with an emotion engine that recognizes the driver's emotions. The implementation method of this system will be described in detail below.

[1323] Data collection methods

[1324] The device collects the driver's driving data using on-board sensors and smartphones. This driving data includes speed, frequency of sudden braking, and driving time. It also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice. This data is sent to a server in real time. As a specific example of collection, speed data is obtained using the smartphone's GPS function, and the frequency of sudden braking is detected using on-board sensors. Facial expression data is collected using a camera, and the driver's voice is recorded using a microphone, and voice tone analysis is performed.

[1325] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, accident history, etc. As a specific example, it obtains real-time road congestion data from the Traffic Management Center, receives current and forecast weather information from the Japan Meteorological Agency, and downloads past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[1326] Data preprocessing and analysis

[1327] The server preprocesses the received data and converts it into an appropriate format. First, it performs missing value imputation and outlier detection and removal to prepare the data for analysis. For example, preprocessing involves generating a histogram of the velocity data and identifying and removing anomalous data points.

[1328] The server then inputs the preprocessed data into a generative AI model to assess the risk of traffic accidents. The generative AI model uses an algorithm trained on a large amount of past data. The model integrates driving data, road condition data, and driver emotion data to calculate a risk score. For example, if the driver is experiencing high stress, the model will assess the risk level as high.

[1329] Real-time alerts and feedback

[1330] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. There are multiple ways to warn the driver, including providing audio alerts, visual alerts, and haptic feedback depending on the driver's emotional state. For example, if the driver is nervous, a gentle audio alert will be played saying, "There is a high-accident zone ahead. Please be careful," and a warning message will be displayed on the dashboard.

[1331] The user receives the warning and is encouraged to review their driving behavior and drive safely. After receiving multiple warnings, the user is also given feedback on the impact of the warning and areas for improvement. The feedback data includes information on the impact of the warning and changes in driving behavior, and is sent to the server via the device.

[1332] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, which improves the accuracy of risk assessment and provides more effective warnings. Once the model update is complete, the new model is applied to the system and used for future risk assessments.

[1333] Examples and prompts

[1334] As an example of data collection, a server obtains road congestion data for Tokyo in real time from a traffic management center at 5:00 PM on a certain day. A device also collects driving data from Driver A's smartphone for the past week and emotion data obtained from a facial recognition camera, and sends these data to the server. Specific examples of prompts include a voice alert such as "There have been many accidents at the next intersection over the past three years. Please be careful," or a visual message such as "Be careful when making the next left turn."

[1335] In this way, the present invention provides a system that predicts and warns about traffic accident risks by analyzing individual driver characteristics and nationwide road conditions in real time, taking into account the driver's emotional state, thereby promoting safe driving and preventing traffic accidents.

[1336] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1337] Step 1: Data collection

[1338] The device collects driving data from the driver using on-board sensors and smartphones. Specifically, it obtains speed data from GPS and the frequency of sudden braking from the brake pedal sensor. It also detects facial expressions with a camera and records voice data with a microphone for emotion recognition. This data is sent to a server in real time.

[1339] Input: Driving data (speed, frequency of sudden braking, driving time), emotional data (facial expressions, voice)

[1340] Output: Driving and emotion data sent to a server in real time

[1341] Step 2: External Data Acquisition

[1342] The server receives nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Specifically, it receives road congestion data from the Traffic Management Center, weather information from the Japan Meteorological Agency, and past accident history data from the Ministry of Land, Infrastructure, Transport and Tourism.

[1343] Input: Road condition data (congestion, weather, accident history)

[1344] Output: Road condition data stored on the server

[1345] Step 3: Data Preprocessing

[1346] The server preprocesses the received driving data, emotion data, and road condition data, including data format conversion, missing value imputation, and outlier removal. Specifically, it removes outliers in speed data and imputes missing emotion data using statistical methods.

[1347] Input: Received driving data, emotion data, road condition data

[1348] Output: Preprocessed data

[1349] Step 4: Risk assessment

[1350] The server inputs the preprocessed data into a generative AI model to assess traffic accident risk. The generative AI model calculates a risk score by integrating driving data, road condition data, and emotional data. For example, if the driver is experiencing high stress, the model will assess the risk level higher.

[1351] Input: Preprocessed data

[1352] Output: Risk score

[1353] Step 5: Send an alert

[1354] If the risk score exceeds a certain threshold as a result of the risk assessment, the device will issue a real-time warning to the driver. The warning method is a voice alert, playing a gentle message saying, "There is a high-accident zone ahead. Please be careful." The visual alert is a warning message displayed on the dashboard.

[1355] Input: Risk Score

[1356] Output: Real-time warning (audio alert, visual alert)

[1357] Step 6: Gather feedback

[1358] After receiving the warning, the user reviews their driving behavior and provides feedback, including the impact of the warning and changes in their driving behavior, which is sent to the server via the device.

[1359] Input: Feedback data on driving behavior revision

[1360] Output: Feedback data sent to the server

[1361] Step 7: Update the model

[1362] The server analyzes the received feedback data and uses it to retrain the generative AI model and emotion engine, thereby improving the accuracy of risk assessment. Once retrained, the generative AI model is applied to the system and used for subsequent risk assessments.

[1363] Input: Feedback data

[1364] Output: An updated generative AI model

[1365] Through the above steps, the system of the present invention takes into account the driver's driving characteristics, national road conditions, and the driver's emotional state, predicts traffic accident risks, and issues warnings, thereby promoting safe driving.

[1366] (Application example 2)

[1367] 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."

[1368] Conventional traffic accident prevention systems assessed risk using only driver driving data and road condition data, making it impossible to accurately assess risk by taking the driver's emotional state into account. Furthermore, the warning format was uniform, making it difficult to issue a warning appropriate to the driver's emotional state. Therefore, improving the effectiveness of traffic accident risk prediction and warnings remains a challenge.

[1369] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1370] In this invention, the server includes a means for collecting driving data and emotional data of the driver, a means for acquiring nationwide road condition data and weather data, a means for analyzing these data and using a generative AI model to evaluate the risk level of a traffic accident, and a means for issuing an audio or visual alert according to the emotional state of the driver when the risk level increases. This makes it possible to integrate the driving data and emotional data of the driver to evaluate the risk of a traffic accident with high accuracy and to issue a warning appropriate to the emotional state of the driver.

[1371] "Driving data" refers to information such as the driver's speed, frequency of sudden braking, and driving time.

[1372] "Emotion data" refers to information about the emotional state of the driver as recognized from their facial expressions and voice.

[1373] "Road condition data" refers to information such as road congestion, weather, and accident history obtained from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism.

[1374] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of past data to assess the risk of traffic accidents.

[1375] "Audio alert" refers to a means of issuing an audio warning to the driver.

[1376] "Visual alert" refers to a means of visually displaying a warning to the driver.

[1377] "Haptic feedback" refers to a means of providing tactile feedback such as vibration to the driver.

[1378] "Risk level" refers to an index that indicates the risk of a traffic accident occurring.

[1379] "Driving behavior patterns" refers to information that indicates a driver's driving habits and characteristic driving behavior.

[1380] The present invention is a system that collects driving data and emotional data of drivers and evaluates and warns about traffic accident risks in real time. An embodiment of this system will be described below with specific examples.

[1381] Data collection and analysis program

[1382] The server obtains road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. Road condition data includes congestion levels, weather, and accident history. The terminal uses on-board sensors and a smartphone app to collect driver driving data (speed, frequency of sudden braking, driving time, etc.). The terminal also uses emotion recognition devices such as cameras and microphones to collect emotional data from the driver's facial expressions and voice.

[1383] The collected data is sent to a server in real time. The server preprocesses the received data, converting it into an appropriate format, and also complements missing values ​​and removes outliers. The preprocessed data is then input into a generative AI model to assess traffic accident risk. The generative AI model learns from a large amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It also uses an emotion engine to obtain driver emotion data and reflect this in the risk assessment. For example, if the driver is feeling stressed, the risk level will be assessed higher.

[1384] Real-time alert program

[1385] If the risk assessment determines that there is a high risk of a traffic accident, the device will issue a real-time warning to the driver. The form of the warning is adjusted according to the driver's emotional state. For example, if the driver is nervous, a gentle voice alert will be used, while if the driver is not feeling stressed, a simple visual alert will be displayed. Specifically, the device will play a voice guide saying, "There is a high-accident zone ahead. Please be careful," and will also display a warning message on the dashboard.

[1386] Feedback and continuous improvement of the system

[1387] Upon receiving the warning, the user will review their driving behavior and strive for safer driving. If the user receives multiple warnings, the system will provide feedback on the effectiveness of the warnings and areas for improvement. The feedback includes information on the impact of the warnings and changes in driving behavior. The feedback data is sent to the server via the device. The server analyzes the received feedback data and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and enables the system to provide more effective warnings.

[1388] Specific examples

[1389] 1. Data Collection Examples

[1390] The server obtains real-time data on road congestion in Tokyo at 5:00 PM on a given day from the traffic management center. The device collects driving data from the past week from the driver's smartphone and emotional data obtained from a facial recognition camera (e.g., stress level and changes in facial expression), and sends the data to the server.

[1391] 2. Data analysis examples

[1392] The server checks that a particular intersection in Tokyo has had a high number of accidents over the past three years. The emotion engine evaluates the driver's stress level as "high," and the generative AI model determines the risk level as "high."

[1393] 3. Examples of warnings

[1394] Just before the user (driver) approaches a specific intersection, the device issues a gentle voice alert saying, "There is a high-accident zone ahead. Please slow down and be careful." At the same time, a warning light on the dashboard flashes to draw attention.

[1395] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[1396] As described above, this invention analyzes the individual driving characteristics of drivers and nationwide road conditions in real time, and also takes into account the emotional state of the driver, thereby predicting and warning about the risk of traffic accidents, thereby promoting safe driving and preventing traffic accidents from occurring.

[1397] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1398] Step 1:

[1399] The server obtains nationwide road condition data in real time from the Traffic Management Center, the Japan Meteorological Agency, and the Ministry of Land, Infrastructure, Transport and Tourism. This data includes congestion levels, weather, and accident history. As input, it receives real-time road condition information from each data source, and as output, it integrates and stores this data. Specifically, it sends API requests, analyzes the response data, and extracts the necessary information.

[1400] Step 2:

[1401] The device collects driving data from the driver using on-board sensors and a smartphone app. This data includes speed, frequency of sudden braking, driving time, etc. It also uses a camera and microphone to collect emotional data from the driver's facial expressions and voice. As input, it receives real-time driving data and emotional data from the sensors and devices, and as output, it sends this data to a server. Specifically, it acquires the data stream from the device, unifies the format, and transmits it.

[1402] Step 3:

[1403] The server preprocesses the received driving data and emotion data. This preprocessing includes converting the data format, filling in missing values, and removing outliers. It receives the driving data and emotion data sent from the device as input, and generates preprocessed data as output. Specifically, it stores the data in a database while processing missing values ​​and outliers within the program.

[1404] Step 4:

[1405] The server inputs the preprocessed data into a generative AI model to evaluate the risk of traffic accidents. The generative AI model learns from a huge amount of past data and integrates road condition data and driving data to perform highly accurate risk assessments. It receives preprocessed driving data and emotion data as input, and generates risk assessment results as output. Specifically, it inputs data into the generative AI model, analyzes the prediction results, and calculates the risk level.

[1406] Step 5:

[1407] The server also reflects the driver's emotional data in the generative AI model and determines the risk level based on the risk assessment results. It receives the emotional assessment data from the emotion engine and the risk assessment results from the generative AI model as input, and generates the final risk level as output. Specifically, it integrates the emotional assessment data with the output of the generative AI model and modifies the risk level.

[1408] Step 6:

[1409] If the risk assessment determines that there is a high risk of a traffic accident, the device issues a real-time warning to the driver. The format of the warning is adjusted according to the driver's emotional state. As input, the device receives the risk assessment results and emotional evaluation data from the server, and as output, issues an appropriate warning to the driver. Specifically, the device selects and executes the method of issuing an audio or visual alert based on the driver's emotional state.

[1410] Step 7:

[1411] When the user receives the warning, they will review their driving behavior and strive for safe driving. If they receive multiple warnings, they will be given feedback on the effectiveness of the warnings and areas for improvement. The system receives driver feedback data as input and sends this data to the server as output. Specific operations include providing an interface for collecting feedback and sending the input information to the server.

[1412] Step 8:

[1413] The server analyzes the received feedback and emotion data and uses it to retrain the generative AI model and emotion engine. This improves the accuracy of risk assessment and provides more effective warnings. It receives user feedback and emotion data as input and creates improved versions of the generative AI model and emotion engine as output. Specifically, it uses the collected data to retrain the algorithm and update and apply the new model.

[1414] An example of a prompt sentence is, "The congestion data obtained from the traffic management center is 0.8, the weather is rainy, the accident history is high, the driver's stress level is 6, and the facial expression change index is 2. Please evaluate the risk of a traffic accident under these conditions and decide whether or not a warning is necessary."

[1415] 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.

[1416] 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.

[1417] 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.

[1418] 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.

[1419] 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.

[1420] 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.

[1421] 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).

[1422] 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, motorcycles, and other devices, 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.

[1423] 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."

[1424] 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.

[1425] 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).

[1426] 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.

[1427] 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.

[1428] 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.

[1429] 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.

[1430] 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.

[1431] 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.

[1432] 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.

[1433] 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.

[1434] 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.

[1435] 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.

[1436] The following is further disclosed regarding the above embodiment.

[1437] (Claim 1)

[1438] a first means for collecting driving data of a driver;

[1439] A second means of obtaining nationwide road condition data;

[1440] A third method is to analyze these data and assess the risk level of traffic accidents.

[1441] a fourth means for alerting the driver if the risk level increases;

[1442] A system including:

[1443] (Claim 2)

[1444] A fifth means for analyzing a driving behavior pattern of a driver based on the driving data;

[1445] A sixth measure is to perform an individual risk assessment according to driving behavior patterns; and

[1446] 10. The system of claim 1, further comprising:

[1447] (Claim 3)

[1448] The fourth means of warning is one of an audio alert, a visual alert, and haptic feedback;

[1449] 10. The system of claim 1.

[1450] (Claim 4)

[1451] A seventh means of collecting data in real time and notifying the driver of the analysis results;

[1452] an eighth means for receiving feedback from the driver and updating the data analysis model based thereon;

[1453] 10. The system of claim 1, further comprising:

[1454] "Example 1"

[1455] (Claim 1)

[1456] A means of obtaining road condition data nationwide;

[1457] A means for collecting driving data of a driver;

[1458] a means for pre-processing the acquired and collected data;

[1459] a means including a generative AI model that analyzes the pre-processed data to assess the risk level of a traffic accident;

[1460] a means of providing real-time warnings to drivers when risk levels increase;

[1461] A system including:

[1462] (Claim 2)

[1463] A means for analyzing a driving behavior pattern of a driver based on driving data;

[1464] A means for conducting individual risk assessments according to driving behavior patterns;

[1465] 10. The system of claim 1.

[1466] (Claim 3)

[1467]

[0020] Including means by which the alert is one of an audio alert, a visual alert, and haptic feedback;

[1468] 10. The system of claim 1.

[1469] "Application Example 1"

[1470] (Claim 1)

[1471] a first means for collecting driving data of a driver;

[1472] A second means of obtaining nationwide road condition data;

[1473] A third method is to analyze these data and assess the risk level of traffic accidents.

[1474] a fourth means for alerting the driver if the risk level increases;

[1475] A system including a fifth means for intervening in the automatic control system of the automated vehicle.

[1476] (Claim 2)

[1477] A sixth means for analyzing a driving behavior pattern of the driver based on the driving data;

[1478] A seventh measure is to carry out an individual risk assessment according to driving behavior patterns; and

[1479] an eighth means for conducting a risk assessment using the generative AI model; and

[1480] 10. The system of claim 1, further comprising:

[1481] (Claim 3)

[1482] the fourth means of warning is one of an audio alert, a visual alert, and haptic feedback;

[1483] 10. The system of claim 1, wherein the fifth means instructs the automated driving system to adjust speed or course.

[1484] "Example 2: Combining Emotion Engines"

[1485] (Claim 1)

[1486] a first means for collecting driving data of a driver;

[1487] A second means of obtaining nationwide road condition data;

[1488] A third means of preprocessing these data to convert them into a suitable format, imputing missing values, and detecting and removing outliers.

[1489] A fourth method is to input the preprocessed data into a generative AI model to assess the risk level of traffic accidents;

[1490] A fifth measure is to collect driver emotion data and reflect it in risk assessments.

[1491] a sixth means for issuing a warning to the driver if the risk level increases;

[1492] A system including:

[1493] (Claim 2)

[1494] A seventh means for analyzing a driving behavior pattern of a driver based on the driving data;

[1495] an eighth means for conducting individual risk assessments according to driving behavior patterns;

[1496] 10. The system of claim 1, further comprising:

[1497] (Claim 3)

[1498] The sixth means of warning is one of an audio alert, a visual alert, and haptic feedback;

[1499] 10. The system of claim 1.

[1500] "Application example 2 when combining emotion engines"

[1501] (Claim 1)

[1502] A first means for collecting driving data and emotion data of a driver;

[1503] a second means for obtaining nationwide road condition data and weather data;

[1504] A third method uses generative AI models to analyze these data and assess the risk level of traffic accidents.

[1505] a fourth means for issuing an audio or visual alert depending on the driver's emotional state when the risk level increases;

[1506] A system including:

[1507] (Claim 2)

[1508] A fifth means for analyzing a driving behavior pattern of the driver based on the driving data and the emotion data;

[1509] A sixth method using a generative AI model that performs individualized risk assessment based on driving behavior patterns and emotion data; and

[1510] 10. The system of claim 1, further comprising:

[1511] (Claim 3)

[1512] The fourth means of warning is one of an audio alert, a visual alert, and haptic feedback depending on the driver's emotional state.

[1513] 10. The system of claim 1. [Explanation of symbols]

[1514] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a first means for collecting driving data of a driver; a second means for acquiring nationwide road condition data; A third means for analyzing these data and assessing the risk level of traffic accidents; a fourth means for issuing a warning to the driver if the risk level increases; A system including:

2. A fifth means for analyzing a driving behavior pattern of a driver based on the driving data; A sixth means for performing an individual risk assessment according to a driving behavior pattern; The system of claim 1 further comprising:

3. the fourth means of warning is one of an audio alert, a visual alert, and haptic feedback; The system of claim 1 .

4. a seventh means for collecting data in real time and notifying the driver of the analysis result; an eighth means for receiving feedback from the driver and updating the data analysis model based thereon; The system of claim 1 further comprising:

Images