System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting and storing a daily behavior pattern of a user; means for receiving disaster information in real time; means for specifying a user in a target area based on the received disaster information; means for collecting latest data (position information, last update time, surrounding user information, and radio wave reception status) of the user when a disaster occurs; and means for issuing a safety confirmation alert to a notification destination registered in advance when the degree of risk exceeds a predetermined threshold.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] Conventional disaster safety confirmation systems require victims to report their own safety, making it difficult to quickly and accurately confirm their safety. Furthermore, if a victim is unconscious or unable to move, there is a problem that safety confirmation cannot be performed at all. Furthermore, due to a lack of information on the surrounding situation and radio wave reception, the risk assessment is insufficient, which can delay life-saving efforts. [Means for solving the problem]

[0005] The present invention solves the aforementioned problems with a system that includes: means for collecting and saving users' daily behavior patterns; means for receiving disaster information in real time; means for identifying users in a target area based on the received disaster information; means for collecting the latest user data (location information, last update time, information about nearby users, and radio wave reception status) when a disaster occurs; means for calculating a risk level by comparing the collected latest data with the daily behavior patterns; and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This system can automatically check the user's safety and notify relevant parties in a prompt and accurate manner. Furthermore, by assessing the risk level taking into account the situation during a disaster, it can support the early implementation of life-saving activities.

[0006] "User" refers to an individual who uses the system and whose safety is to be confirmed.

[0007] "Daily behavior patterns" are data that indicate the user's usual behavior and movement tendencies.

[0008] "Disaster information" refers to information about emergencies such as earthquakes, weather warnings, fires, and large-scale accidents.

[0009] "Real-time" refers to a time frame in which data or information is updated immediately.

[0010] "Receiving" is the act of taking in information or data from outside.

[0011] "Target area" means an area where a disaster has occurred or is likely to be affected by a disaster.

[0012] "Location information" is data that indicates the current location of the device.

[0013] "Last updated time" is the time when the latest data or information was last updated.

[0014] "Nearby user information" is data relating to other users who are near the user.

[0015] "Radio wave reception status" is data indicating the radio wave reception status of a communication device.

[0016] "Collection" is the act of acquiring and storing data.

[0017] "Comparison" is the act of comparing two or more elements and evaluating the differences between them.

[0018] "Danger level" is a number or index that indicates the degree of danger in a particular situation.

[0019] A "threshold" is a limit value set for a particular load or condition.

[0020] The "notification destination" is a contact point registered in advance to receive information about the user's safety.

[0021] A "safety confirmation alert" is a warning or notification issued to confirm the user's safety status. [Brief explanation of the drawings]

[0022] [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

[0023] 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.

[0024] First, the terms used in the following description will be explained.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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."

[0030] [First embodiment]

[0031] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0032] 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.

[0033] 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).

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0039] 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.

[0040] 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.

[0041] 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.

[0042] 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."

[0043] This invention is an automatic safety confirmation system for use in disasters, which allows users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs.

[0044] Overall system configuration

[0045] This system mainly consists of the following elements:

[0046] 1. A means of collecting and storing users' daily behavior patterns

[0047] 2. Means of receiving disaster information in real time

[0048] 3. A method for identifying users in target areas based on disaster information

[0049] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[0050] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[0051] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[0052] Specific embodiments of each means will be described in detail below.

[0053] Collecting and storing users' daily behavior patterns

[0054] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0055] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[0056] Learning daily behavior patterns

[0057] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[0058] Receiving disaster information and identifying affected areas

[0059] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[0060] Data collection in the event of a disaster

[0061] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[0062] Comparison of the latest data and calculation of risk

[0063] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[0064] Issuance of safety confirmation alerts

[0065] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, and signal reception status.

[0066] Specific examples

[0067] For example, if user A uses a smartphone on a daily basis, location information and activity data are sent from the device to the server. The server receives and stores this data, and AI learns user A's daily behavior patterns.

[0068] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User A is in that area.

[0069] User A's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, and radio wave reception status to the server. The server analyzes this data and calculates the risk level based on any differences with User A's behavioral patterns.

[0070] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User A's registered contacts (e.g., family and workplace). This is expected to enable User A's safety to be confirmed quickly and accurately, and for necessary rescue operations to be carried out promptly.

[0071] The processing flow will be explained below.

[0072] Step 1:

[0073] The device collects data on the user's daily behavior (location information, activity information, app usage, health data, etc.).

[0074] Step 2:

[0075] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[0076] Step 3:

[0077] The server receives the data sent from the terminal and stores it in a database.

[0078] Step 4:

[0079] The server inputs the stored data into an AI model that learns the user's daily behavior patterns.

[0080] Step 5:

[0081] The server generates and updates each user's behavioral pattern profile based on the learning results.

[0082] Step 6:

[0083] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from the disaster information service.

[0084] Step 7:

[0085] The server identifies the disaster-affected area based on the disaster information received, and extracts users in that area from the database.

[0086] Step 8:

[0087] The terminal receives disaster information and instantly collects current location information, last update time, information on nearby users, and radio wave reception status.

[0088] Step 9:

[0089] The latest data collected by the device is sent to the server.

[0090] Step 10:

[0091] The server compares the latest data received from the device with the daily behavior pattern profile, and the AI ​​calculates the level of risk.

[0092] Step 11:

[0093] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[0094] Step 12:

[0095] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, and radio wave reception status.

[0096] Example 1

[0097] 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."

[0098] In the event of a disaster, prompt and accurate safety confirmation is extremely important for saving many lives. However, manual safety confirmation requires time and effort, and in the event of a disaster, communication infrastructure may be destroyed or disrupted, making it difficult to respond quickly. To address these issues, the present invention aims to provide a system that uses AI to automatically confirm the safety of users and quickly send necessary notifications.

[0099] 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.

[0100] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the latest collected data with daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for transmitting the generated data to the server and analyzing it with an AI model, and means for updating the behavior pattern profile based on the daily behavior patterns. This makes it possible to quickly and accurately confirm the safety of users when a disaster occurs.

[0101] "User's daily behavioral patterns" is a general term for data related to the user's daily actions, movements, and activities.

[0102] "Disaster information" refers to real-time information about disasters such as earthquakes, weather warnings, fires, and large-scale accidents.

[0103] The "target area" refers to the area where a disaster has occurred, as identified based on disaster information.

[0104] "Latest user data" refers to data such as the user's location information at the time of the disaster, the last update time, information about nearby users, and radio wave reception conditions.

[0105] The "risk level" is a value that indicates the degree of risk to the user, calculated by comparing the latest collected data with daily behavior patterns.

[0106] A "safety confirmation alert" is a warning notification issued to a notification destination registered in advance when the user's risk level exceeds a certain threshold.

[0107] "AI model" refers to artificial intelligence algorithms and systems that analyze collected data and learn users' daily behavioral patterns.

[0108] A "behavioral pattern profile" is a data set that shows the characteristics of a user's unique daily behavior, generated based on the results of learning by an AI model.

[0109] The "Disaster Information Service" is a system that distributes real-time disaster-related information provided by organizations such as the Japan Meteorological Agency.

[0110] The present invention is an automatic safety confirmation system for use in the event of a disaster, allowing users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs. Each element of the present invention is described in detail below.

[0111] This system mainly consists of the following elements:

[0112] Collecting and storing users' daily behavior patterns

[0113] As the user uses their smartphone on a daily basis, the device collects location information, app and web usage, activity information (walking, driving, standing still), and health data (number of steps, heart rate, etc.). This data is sent to a server at regular intervals (e.g., every 15 minutes). The hardware used is assumed to be a smartphone (e.g., Android device or iPhone). The collected data is encrypted by the device using HTTPS and securely sent to the server.

[0114] Sending data to the server

[0115] The server receives the data sent from the device and stores it in a database. This allows the user's daily behavioral patterns to be accumulated and prepared for later AI analysis. The software used is a database management system (e.g., MySQL or PostgreSQL) and an AI framework for data analysis (e.g., TensorFlow or PyTorch).

[0116] Learning daily behavior patterns

[0117] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated. The Python programming language and related libraries (e.g., NumPy, Pandas) are used to train the AI ​​model.

[0118] Receiving disaster information

[0119] The server works in conjunction with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, and large-scale accidents. For example, disaster information is sent to the server via an API, and the data is analyzed immediately. A real-time data processing framework (e.g., Apache Kafka) is used for disaster information management.

[0120] Identifying target areas and collecting data

[0121] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database.The device then receives the disaster information and immediately collects the current location information, the last update time, information about nearby users (via short-range communication), and the radio wave reception status, and promptly transmits this data to the server.

[0122] Comparison of the latest data and calculation of risk

[0123] The server compares the latest data sent from the device with the daily behavioral pattern profile stored in the database. The AI ​​evaluates the differences between these data and calculates the level of risk. If the level of risk exceeds a certain threshold, it detects it as an abnormal pattern.

[0124] Issuance of safety confirmation alerts

[0125] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status. Notifications are sent via email, SMS, push notifications, etc.

[0126] Specific examples

[0127] For example, user A uses a smartphone on a daily basis, and location and activity data is sent from the device to a server. The server stores the received data, and AI learns user A's daily behavioral patterns based on data such as commuting route, activity times, and places of stay.

[0128] One day, a magnitude 5.0 earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and checks the database to see if User A is in that area. User A's device receives the earthquake information and immediately collects and sends this information to the server, along with the user's current location, last update time, information about nearby users, and signal reception status. The server uses this data to compare it with normal activity patterns and calculates User A's risk level. If the risk level exceeds a threshold, the server automatically sends a safety alert and detailed information to User A's registered contacts (e.g., family and workplace).

[0129] Example prompts for generative AI models

[0130] Generate code to program an automatic disaster safety confirmation system described in natural language. The system collects users' daily behavior from their smartphones and sends the data to a server. The server uses AI to learn daily behavior patterns and receives and analyzes disaster information in real time. When a disaster occurs, it collects the latest user data, calculates the risk level, and automatically sends safety confirmation alerts as needed.

[0131] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0132] Step 1:

[0133] Collecting users' daily behavior patterns

[0134] The device collects data on the user's daily activities. Specifically, it uses the smartphone's GPS function, app usage, activity sensors (accelerometer and gyroscope), and health data (pedometer and heart rate monitor) to collect location information, app usage history, movement status (walking, driving, stationary, etc.), and vital signs at regular intervals (e.g., every 15 minutes).

[0135] Input: Data from various sensors and apps on the user's smartphone

[0136] Output: A set of user's daily behavior data (location information, activity information, health data, etc.)

[0137] Step 2:

[0138] Sending data to the server

[0139] The device sends the collected daily behavior data to a server at regular intervals. The data is encrypted using the HTTPS protocol and sent securely to the server.

[0140] Input: Collected daily behavior data

[0141] Output: Encrypted daily activity data packet sent to the server

[0142] Step 3:

[0143] Saving daily behavior patterns

[0144] The server receives the daily behavior data sent from the device and stores it in a database, which is later used for AI analysis.

[0145] Input: Encrypted daily activity data sent from the device

[0146] Output: Daily behavior data stored in a database

[0147] Step 4:

[0148] Learning daily behavior patterns

[0149] The server periodically inputs the daily behavior data stored in the database into the AI ​​model for analysis. The AI ​​model analyzes this data and learns the user's usual behavior patterns. Based on the learning results, a behavioral pattern profile for each user is generated and regularly updated.

[0150] Input: Daily behavior data stored in a database

[0151] Output: Generated and updated behavioral pattern profile

[0152] Step 5:

[0153] Receiving disaster information

[0154] The server works with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. For example, it receives earthquake occurrence information from disaster information services via API.

[0155] Input: Disaster information from various disaster information services

[0156] Output: Disaster information data immediately stored on the server

[0157] Step 6:

[0158] Identifying the target area

[0159] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database. For example, it analyzes the latitude and longitude information of the earthquake area and identifies users within that range.

[0160] Input: Received disaster information, user location information stored in the database

[0161] Output: List of users in the identified disaster area

[0162] Step 7:

[0163] Collection of the latest data

[0164] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information about nearby users (via short-range communication such as Bluetooth), and radio wave reception status, and quickly transmits this data to the server.

[0165] Input: Disaster information, data from various sensors and communication devices on the user's smartphone

[0166] Output: Latest location information, last update time, information on nearby users, data packets of signal reception status

[0167] Step 8:

[0168] Calculating the risk

[0169] The server compares the latest data sent from the device with the stored daily behavioral pattern profile, and the AI ​​calculates the risk level. For example, if the user's location does not change for a long period of time even though it is during a normal activity time, it will determine this as an abnormality.

[0170] Input: Latest data sent from the device, saved behavioral pattern profile

[0171] Output: Calculated risk

[0172] Step 9:

[0173] Issuance of safety confirmation alerts

[0174] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status.

[0175] Input: Calculated risk level, user's registered contact information

[0176] Output: Safety confirmation alert sent and detailed information

[0177] (Application example 1)

[0178] 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."

[0179] In modern society, the occurrence of natural disasters and large-scale accidents is an unavoidable problem. Quickly and accurately confirming the safety of people when a disaster occurs is an important issue that directly relates to ensuring the safety of victims and streamlining rescue efforts. However, current safety confirmation methods involve a lot of manual confirmation, making it difficult to grasp safety in a timely manner. Furthermore, they are insufficient in proposing appropriate evacuation routes in the event of a disaster. To solve these issues, a system is needed that can more efficiently and automatically confirm safety and also propose appropriate evacuation routes.

[0180] 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.

[0181] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the collected latest data with the daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for automatically suggesting evacuation routes for high-risk users, and means for calculating the level of risk and confirming the safety of users based on prompt text from the generation AI model. This makes it possible to quickly and automatically confirm the safety of users and suggest appropriate evacuation routes when a disaster occurs.

[0182] "User's daily behavioral patterns" refer to the user's daily patterns of movement and activities, and include location information, app and web usage, activity information, health care data, etc.

[0183] "Means for receiving disaster information in real time" refers to a system that works in conjunction with disaster information services provided by the Japan Meteorological Agency and other organizations, and has the function of receiving disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. in real time.

[0184] The "means for identifying users in the target area" has a function of extracting users who are in the disaster-affected area from a database based on the received disaster information.

[0185] The "means for collecting the latest data" has the function of collecting information such as the user's location, the last update time, information about nearby users, and radio wave reception conditions when a disaster occurs.

[0186] The "means for calculating the risk level" has a function of comparing the latest collected data with the saved daily behavior patterns, evaluating the difference, and calculating the risk level of the user.

[0187] The "means for issuing safety confirmation alerts to notification recipients" has the function of automatically sending safety confirmation alerts to notification recipients registered in advance by the user when the risk level exceeds a certain threshold.

[0188] The "means for automatically suggesting evacuation routes" is a function that automatically suggests appropriate evacuation routes based on AI analysis to users at high risk.

[0189] The means for checking safety based on the "prompt sentence of the generative AI model" has the function of inputting the prompt sentence into the AI ​​model, calculating the user's level of risk, and checking the user's safety.

[0190] This invention is a system that automatically checks the user's safety during a disaster and notifies emergency contacts. The system uses AI to learn the user's daily behavior patterns, and when a disaster occurs, it compares them with the user's latest data to calculate the level of risk. In addition, if the risk is high, the system has the function of suggesting an appropriate evacuation route.

[0191] Hardware and software configuration

[0192] The hardware and software used to realize the system of the present invention are as follows.

[0193] Hardware: Smartphones, servers

[0194] Software: Python, requests (library for calling APIs), geopy (geographic calculation), smtplib (for sending emails)

[0195] External API: Real-time disaster information API for disaster information provision services

[0196] Overall system configuration

[0197] Safety confirmation system configuration

[0198] 1. Data Collection:

[0199] The server collects the user's daily behavioral patterns. The smartphone periodically collects data such as the user's location, activity, and signal reception status, and sends it to the server. The server stores this data and later analyzes it using AI.

[0200] 2. Receiving disaster information:

[0201] The server works in conjunction with disaster information services to receive disaster information in real time. Information on earthquakes, weather warnings, fires, large-scale accidents, etc. is sent to the server. Based on this information, the area where the disaster occurred is identified.

[0202] 3. Identifying users in the target area:

[0203] The server extracts users who are in the disaster area from the database, thereby identifying users who may be affected by the disaster.

[0204] 4. Collecting the latest data:

[0205] When a disaster occurs, the smartphone collects the user's latest location information, activity information, and signal reception status, and immediately sends it to a server, making it possible to check the safety of users in real time.

[0206] 5. Data comparison and risk calculation:

[0207] The server uses AI to compare the latest data with daily behavior patterns. AI analysis evaluates the differences between the collected data and normal behavior patterns and calculates the user's risk level.

[0208] 6. Issuance of Safety Confirmation Alerts:

[0209] If the specified risk threshold is exceeded, a safety confirmation alert is automatically sent to the user's emergency contacts, allowing the necessary response to be taken promptly.

[0210] 7. Suggested evacuation routes:

[0211] For users at high risk, evacuation routes are automatically suggested based on AI analysis. By providing appropriate evacuation routes, users' safety is supported.

[0212] Specific example explanation

[0213] Assume that user A travels daily in central Tokyo and its suburbs. The invention collects user A's location information and activity data, and the server learns his daily behavior patterns based on this data. One day, an earthquake occurs, and data from a disaster information service is sent to the server. If it is confirmed that user A is included in the area affected by the earthquake, the server collects user A's latest location information and activity data at that time and compares it with his daily behavior patterns. If the analysis determines that there is a high risk, the server issues a safety confirmation alert to user A's family and workplace. Furthermore, evacuation routes are automatically suggested to user A's smartphone.

[0214] Prompt Sentence Examples

[0215] Use the following prompt for the generative AI model:

[0216] Please calculate the risk level based on the daily behavior pattern data of user ID: user123, comparing it with the latest information (location information, activity data, etc.) in the event of a disaster. Provide the following data:

[0217] Daily behavior pattern data: [Detailed data]

[0218] Latest data: [Detailed data]

[0219] In this way, the present invention automatically checks the user's safety during a disaster, quickly notifies emergency contacts, and supports necessary responses, thereby ensuring the safety of victims and improving the efficiency of rescue operations.

[0220] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0221] Step 1:

[0222] Collect user location and activity data.

[0223] Input: Sensor data such as location information, activity data, and signal reception status from the user's smartphone.

[0224] Processing: The smartphone collects this data at regular intervals, converts it into a data format, and sends it to the server.

[0225] Output: A set of user location and activity data sent to the server.

[0226] Step 2:

[0227] The server stores the user's daily behavior patterns.

[0228] Input: Location, activity data, and other sensor data sent from your smartphone.

[0229] Processing: The server stores the received data in a database and integrates it with past data, thereby accumulating daily behavior patterns.

[0230] Output: User's daily behavior pattern data stored in a database.

[0231] Step 3:

[0232] Receive disaster information in real time.

[0233] Input: Real-time data from disaster information services (earthquake occurrence information, weather warnings, fire information, large-scale accident information, etc.).

[0234] Processing: The server calls the API and obtains disaster information.

[0235] Output: Disaster information data stored on the server.

[0236] Step 4:

[0237] Identify users in your target area.

[0238] Input: Disaster information (data on the location and scale of the disaster), user's daily behavior pattern data (location information).

[0239] Processing: The server analyzes the disaster area and extracts users who are in that area from the database.

[0240] Output: A list of users in the disaster area.

[0241] Step 5:

[0242] Collect the latest data when a disaster occurs.

[0243] Input: Disaster notification, latest location information of the user's smartphone, activity data, and signal reception status.

[0244] Processing: Upon receiving disaster information, the smartphone immediately collects the latest sensor data and sends it to the server.

[0245] Output: The latest sensor data sent to the server.

[0246] Step 6:

[0247] The latest data is compared with daily behavior patterns to calculate the level of risk.

[0248] Input: Latest sensor data, daily behavior pattern data.

[0249] Processing: The server uses the generated AI model to compare the latest data with daily behavior patterns and calculate the user's risk level. The prompt sentence is input into the model and the analysis results are obtained.

[0250] Output: Calculated user risk level (levels such as normal, caution, danger, etc.).

[0251] Step 7:

[0252] If the risk level exceeds a certain threshold, an alert will be issued to check the safety of the person.

[0253] Input: Risk data, user's emergency contact information.

[0254] Processing: If the server detects a risk level that exceeds the threshold, it will automatically send an email or SMS to the user's emergency contacts to confirm their safety.

[0255] Output: Safety alert sent to emergency contacts.

[0256] Step 8:

[0257] Automatically suggests evacuation routes to users at high risk.

[0258] Input: User's current location, risk level data, and surrounding environment information.

[0259] Processing: The server uses AI to analyze surrounding geographical information and disaster information to generate an appropriate evacuation route, which is then sent to the user's smartphone.

[0260] Output: Evacuation route displayed on the user's smartphone.

[0261] 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.

[0262] This invention is an automatic safety confirmation system for use in disaster situations, and by combining it with an emotion engine that recognizes the user's emotions, it is possible to perform more accurate risk assessments. This system uses AI to analyze smartphone usage data, and aims to automatically confirm and notify the user's safety when a disaster occurs.

[0263] Overall system configuration

[0264] The system consists of the following main elements:

[0265] 1. A means of collecting and storing users' daily behavior patterns

[0266] 2. Means of receiving disaster information in real time

[0267] 3. A method for identifying users in target areas based on disaster information

[0268] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[0269] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[0270] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[0271] 7. Emotion engine that recognizes user emotions

[0272] Specific embodiments of each means will be described in detail below.

[0273] Collecting and storing users' daily behavior patterns

[0274] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0275] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[0276] Learning daily behavior patterns

[0277] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[0278] Receiving disaster information and identifying affected areas

[0279] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[0280] Data collection in the event of a disaster

[0281] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[0282] Comparison of the latest data and calculation of risk

[0283] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[0284] Use of emotion engine

[0285] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions, voice, and actions, and sends it to the server.

[0286] The server uses the emotion data sent by the user to assess the risk level. The emotion data indicates the user's stress level and tension, and the accuracy of the risk level is improved based on this. For example, if a user is in an abnormally stressed state when a disaster occurs, this information is reflected in the risk level calculation.

[0287] Issuance of safety confirmation alerts

[0288] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[0289] Specific examples

[0290] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives and stores this data, and the AI ​​learns user B's daily behavioral patterns.

[0291] One day, an earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area, and extracts from the database whether User B is in that area.

[0292] User B's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, signal reception status, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[0293] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family or workplace). This is expected to enable User B's safety to be confirmed quickly and accurately, and to enable necessary rescue operations to be carried out promptly.

[0294] The processing flow will be explained below.

[0295] Step 1:

[0296] The device collects data on the user's daily activities (location information, activity information, app usage, health data, and emotional data). Emotional data is collected by an emotion engine that recognizes the user's facial expressions, voice, and behavior.

[0297] Step 2:

[0298] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[0299] Step 3:

[0300] The server receives the data sent from the terminal and stores it in a database.

[0301] Step 4:

[0302] The server inputs the stored data into the AI ​​model, which learns the user's daily behavioral patterns and emotional data. Based on the learning results, a behavioral pattern profile for each user is generated and updated.

[0303] Step 5:

[0304] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from various disaster information services.

[0305] Step 6:

[0306] The server identifies the disaster-affected area based on the disaster information received, and extracts users who are in that area from the database.

[0307] Step 7:

[0308] When the device receives disaster information, it immediately collects the current location information, the last update time, information about nearby users, radio wave reception conditions, and emotional data.

[0309] Step 8:

[0310] The latest data collected by the device is sent to the server.

[0311] Step 9:

[0312] The server compares the latest data received from the device with the daily behavioral pattern profile and emotional data, and the AI ​​calculates the risk level, taking into account the user's emotional data (e.g., stress level and tension).

[0313] Step 10:

[0314] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[0315] Step 11:

[0316] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, radio wave reception status, and emotional data.

[0317] Example 2

[0318] 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."

[0319] In the event of a disaster, it is extremely important to quickly and accurately confirm the safety of users and take appropriate action. However, conventional systems do not take into account the user's daily behavioral patterns or emotional state in risk assessment, resulting in low accuracy in safety confirmation. The present invention aims to solve these problems.

[0320] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest data of users (location information, last update time, information on surrounding users, and communication environment) when a disaster occurs, means for comparing the collected latest data with daily behavior patterns to calculate a risk level, means for recognizing user emotion data and incorporating it into risk assessment, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables highly accurate risk assessment that takes into account the user's behavior patterns and emotional state.

[0321] "Means for collecting and storing users' daily behavioral patterns" refers to means for collecting users' daily activities, location information, app usage, and activity information, and storing this in a database.

[0322] "Means for receiving disaster information in real time" refers to means for obtaining disaster information such as earthquakes, weather, fires, and large-scale accidents in real time and transmitting it to a server.

[0323] The "means for identifying users in a target area based on received disaster information" is a means for identifying a disaster-affected area from received disaster information and extracting users who are in that area from a database.

[0324] "Means for collecting the latest user data (location information, last update time, surrounding user information, communication environment) when a disaster occurs" refers to means for collecting the user's current location information, last update time, user information via surrounding communications, and the status of radio waves and the communication environment when a disaster occurs.

[0325] The "means for calculating the risk by comparing the latest collected data with daily behavior patterns" is a means for comparing the latest collected data with the user's saved daily behavior pattern profile, evaluating the difference, and calculating the risk.

[0326] "Means for recognizing the user's emotional data and incorporating it into risk assessment" refers to means for collecting and analyzing data on the user's emotional state (e.g., facial expressions and voice) and reflecting it in risk assessment.

[0327] "Means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold" refers to a means for automatically sending a safety confirmation alert to the user's registered emergency contacts (family, workplace, etc.) when the calculated risk level exceeds a pre-set threshold.

[0328] This invention is an automatic safety confirmation system for use in the event of a disaster, which performs more accurate risk assessment based on the user's daily behavioral patterns and emotional data. The system uses AI to analyze smartphone usage data and automatically confirms and notifies the user's safety in the event of a disaster.

[0329] System Configuration

[0330] The system consists of the following main elements:

[0331] 1. A means of collecting and storing users' daily behavior patterns

[0332] The device collects data on the user's daily behavior using the smartphone's GPS, health sensors, app usage logs, camera, microphone, etc. This data includes location information, activity data (walking, driving, and standing still), heart rate, and app usage.

[0333] The collected data is sent to the server at regular intervals (for example, every 15 minutes) and stored on the server.

[0334] 2. A means of learning daily behavior patterns

[0335] The server periodically retrieves daily behavior data stored in the database and analyzes it using a generative AI model, primarily using deep learning frameworks such as TensorFlow and PyTorch.

[0336] The AI ​​model learns the user's usual behavioral patterns (for example, daily commuting route and activity times) and generates and updates a behavioral pattern profile based on this.

[0337] 3. Means of receiving disaster information and identifying affected areas

[0338] The server works in conjunction with various disaster information services (e.g., the Japan Meteorological Agency's API) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents.

[0339] When disaster information is received, the area where the disaster occurred is identified based on that information, and users who are in that area are extracted from the database.

[0340] 4. How to collect the latest user data when a disaster occurs

[0341] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information on nearby users (via short-range communication), and the communication environment.

[0342] This data is quickly transmitted to the server.

[0343] 5. Comparing the latest data with daily behavior patterns and calculating the risk level

[0344] The server compares the latest data sent from the device with the stored daily behavior pattern profile.

[0345] The generative AI evaluates the differences in this data and calculates the risk level, detecting anomalies based on deviations from normal behavioral patterns.

[0346] 6. Methods for Recognizing User Emotional Data and Incorporating It into Risk Assessment

[0347] The device collects emotional data from the user's facial expressions, voice, and actions, and sends it to a server using a camera and microphone.

[0348] The server performs a risk assessment based on the emotional data analyzed by the emotion engine, taking into account the user's stress level and tension.

[0349] 7. How to issue safety confirmation alerts

[0350] If the calculated risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace).

[0351] The alert includes the user's last location, last update time, information about nearby users, communication environment, emotional data, etc.

[0352] Specific examples

[0353] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives this data and stores it in a database, and the generative AI model learns user B's daily behavioral patterns.

[0354] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User B is in that area.

[0355] User B's device receives the earthquake information and immediately sends its current location, last update time, information about nearby users, communication environment, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[0356] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family, workplace). This process is expected to quickly and accurately confirm User B's safety, and enable necessary rescue operations to be carried out promptly.

[0357] Prompt Sentence Examples

[0358] "Please explain in detail your automated safety confirmation system in the event of a disaster. Please tell us the specific processing steps, from data collection to risk assessment and safety confirmation alerts. Please also include examples of application in real-life scenarios."

[0359] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0360] Processing Steps

[0361] Step 1: Data collection

[0362] The device collects information about the user's daily activities, including location information using GPS, activity information obtained from smartphone sensors (e.g., walking, driving, standing still), app usage, and heart rate.

[0363] Input: User's smartphone usage, sensor information

[0364] Output: Collected daily behavior data (e.g., location information, heart rate)

[0365] Specific operation: The device activates the sensor every 15 minutes and packages the acquired data in JSON format.

[0366] Step 2: Send data

[0367] The device sends the collected daily behavior data to the server at regular intervals (e.g., every 15 minutes).

[0368] Input: Collected daily behavior data

[0369] Output: Data sent to the server

[0370] Specific operation: The device sends JSON formatted data to the server's API endpoint using HTTPS.

[0371] Step 3: Save Data

[0372] The server receives the daily activity data sent from the terminal and stores it in a database.

[0373] Input: Daily activity data sent from the device

[0374] Output: Behavioral data stored in a database

[0375] Specific behavior: The server parses the received JSON data and stores it in the database in the appropriate format.

[0376] Step 4: Learning behavioral patterns

[0377] The server periodically retrieves the behavioral data stored in the database, inputs it into the generative AI model, and analyzes it.

[0378] Input: Saved daily activity data

[0379] Output: Learned behavioral pattern profile

[0380] Specific operation: The server uses TensorFlow and PyTorch to run a behavior prediction model and learn user behavior patterns.

[0381] Step 5: Receiving disaster information

[0382] The server receives information in real time from the disaster information service.

[0383] Input: Data from disaster information service

[0384] Output: Disaster information received by the server

[0385] Specific operation: The server periodically checks the API endpoint to obtain the latest disaster information.

[0386] Step 6: Identify users in your target area

[0387] The server identifies the disaster-affected area from the received disaster information and extracts users within that area.

[0388] Input: Disaster information, user location information in the database

[0389] Output: A list of users within the identified target area

[0390] Specific operation: The server uses GIS to map the disaster area and extracts users in that area from the database.

[0391] Step 7: Collect the latest data

[0392] When the terminal receives disaster information, it collects the current location information, the last update time, information on nearby users, and the communication environment.

[0393] Input: Disaster information, device sensor information

[0394] Output: Latest collected data (location information, communication environment, etc.)

[0395] What it does: The device performs GPS and Bluetooth scans to collect and send the latest data.

[0396] Step 8: Data comparison and risk calculation

[0397] The server compares the latest data sent from the device with the saved daily behavior patterns and calculates the level of risk.

[0398] Input: Latest data, daily behavior pattern profile

[0399] Output: Calculated risk

[0400] Specific operation: The server uses a generative AI model to detect abnormalities in behavioral patterns and evaluates the level of risk based on the results.

[0401] Step 9: Collect and evaluate emotion data

[0402] The terminal collects the user's emotional data (facial expressions, voice, etc.) and sends it to the server.

[0403] Input: User's facial expression and voice data

[0404] Output: Parsed emotion data

[0405] Specific operation: The device uses the camera and microphone to capture emotion data and sends it to the server.

[0406] Step 10: Finalize your risk assessment

[0407] The server recalculates the risk assessment based on the emotion data.

[0408] Input: Emotion data, initial risk assessment

[0409] Output: Adjusted risk assessment

[0410] Specific operation: The server takes in the emotional data, reassess the danger level, and generates the final result.

[0411] Step 11: Issue an alert

[0412] If the calculated risk level exceeds a threshold, the server sends a safety confirmation alert to the registered destination of the target user.

[0413] Input: Adjusted risk assessment, user registration information

[0414] Output: Safety confirmation alert issued

[0415] Specific operation: The server generates an alert message containing the necessary information (e.g., last location, nearby user information, etc.) and sends it to the registered recipient via email or SMS.

[0416] In this way, the system takes into account the user's daily behavior and emotional state, enabling quick and highly accurate confirmation of safety during a disaster.

[0417] (Application example 2)

[0418] 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."

[0419] Although systems for confirming users' safety during disasters already exist, they typically rely on users to provide information voluntarily, which often results in inaccurate information being unavailable. Furthermore, these systems assess risk based solely on location and activity data, without taking into account the user's emotional state, limiting the accuracy of risk assessments. Therefore, there is a need for a system that can more quickly and accurately confirm users' safety and provide appropriate notifications.

[0420] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in the target area based on the received disaster information, means for collecting user emotion data and using it for risk assessment, means for calculating the risk level by comparing the latest collected data with daily behavior patterns, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables more accurate risk assessment and faster safety confirmation by analyzing the user emotion data and reflecting the stress level and tension level in the risk assessment.

[0421] "User's daily behavior patterns" refers to the totality of data such as the user's daily actions, movements, and app usage.

[0422] "Disaster information" refers to information about emergencies such as earthquakes, fires, weather warnings, and large-scale accidents.

[0423] "Target area" refers to the geographical area identified as being affected by a disaster based on disaster information.

[0424] The "latest data" is real-time data including the user's location information, the last update time, information about nearby users, radio wave reception conditions, and emotion data at the time of the disaster.

[0425] The "risk level" is a numerical value or classification of the degree of risk in the situation in which the user is placed.

[0426] "Emotion data" is data that indicates the emotional state of the user as recognized from facial expressions, tone of voice, and the like.

[0427] "Risk assessment" is the process of calculating a user's risk level based on collected data.

[0428] A "safety confirmation alert" is a message or notification sent to a contact to confirm whether the user is safe.

[0429] The present invention is an automatic safety confirmation system for use in the event of a disaster, which collects data on users' daily behavior patterns and emotions and performs highly accurate risk assessments in the event of a disaster. Hereinafter, embodiments of the present invention will be described in detail.

[0430] Overall system configuration

[0431] This system is mainly composed of three elements: a server, a terminal, and a user.

[0432] Collecting and storing daily behavior patterns

[0433] As users use their smartphones on a daily basis, the device collects location information, activity data (e.g., walking, driving, and standing still), app usage, and emotional data (e.g., facial expressions and tone of voice). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0434] Learning daily behavior patterns

[0435] The server periodically retrieves daily behavior data stored in the database and inputs it into an AI model (e.g., TensorFlow, PyTorch). The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route, activity times). Based on the learning results, a behavioral pattern profile is generated and continuously updated.

[0436] Receiving disaster information and identifying affected areas

[0437] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the server identifies the area where the disaster occurred and extracts users who are in that area from the database.

[0438] Data collection in the event of a disaster

[0439] When a device receives disaster information, it immediately collects its current location, the last update time, information about nearby users (e.g., device detection via short-range communication), and the status of radio wave reception. This data is then promptly sent to the server. The device also collects emotion data and sends it to the server.

[0440] Comparison of the latest data and calculation of risk

[0441] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change even during normal activity times, it will be detected as an abnormal pattern.

[0442] Use of emotion engine

[0443] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions and tone of voice and sends it to a server. The server analyzes the emotion data and reflects stress levels and tension in the risk assessment, improving the accuracy of the risk assessment.

[0444] Issuance of safety confirmation alerts

[0445] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[0446] Specific examples

[0447] For example, when a user uses a smartphone on a daily basis, location information, app usage status, and emotional data are sent from the device to a server. The server receives and stores this information, and AI learns the user's daily behavioral patterns. One day, an earthquake occurs. The server receives disaster information, confirms that the disaster occurred in a specific area, and extracts from the database whether the user is in that area. The device receives the earthquake information and immediately collects current location information, last update time, information about surrounding users, signal reception status, and emotional data, and sends them to the server. The server analyzes this data and calculates the level of risk based on the difference between the behavioral pattern and the emotional data. If the level of risk exceeds a threshold, the server automatically sends a safety confirmation alert and detailed information to the user's registered contacts.

[0448] Example prompts for generative AI models

[0449] Enter the user's face photo and voice data:

[0450] Facial photo data:[path_to_image]\

[0451] Audio data:[path_to_audio]\

[0452] \

[0453] Parse the following data attributes:

[0454] 1. Emotion recognition (e.g., joy, anxiety, anger, sadness)

[0455] 2. Tone of voice (e.g., nervous, stressed, calm)

[0456] \

[0457] As output, return the user's emotional state and its associated numerical data (e.g., stress level 0-100).

[0458] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0459] Step 1:

[0460] Data collection

[0461] The device collects the user's daily behavioral patterns. Specifically, it obtains location information, activity data, app usage, and emotional data (facial expressions, tone of voice) using sensors, microphones, and cameras. For example, it uses APIs such as LocationServices.requestLocationUpdates(), ActivityRecognitionClient.requestActivityUpdates(), Camera.captureImage(), and Microphone.recordAudio(). The input data is various types of user behavioral information, and the output data is the collected behavioral data.

[0462] Step 2:

[0463] Data transmission

[0464] The collected data is sent to the server at regular intervals (for example, every 15 minutes). This involves sending a POST request using an HTTP client. For example, HttpClient.sendPostRequest(data). The input data is the behavioral data collected on the device, and the output data is the behavioral data sent to the server.

[0465] Step 3:

[0466] Data storage

[0467] The server stores the received data in a database. Specifically, it executes an INSERT statement in the database using SQL or similar. Example: Database.insertData(data). The input data is the behavioral data sent, and the output data is the behavioral data stored in the database.

[0468] Step 4:

[0469] Learning behavioral patterns

[0470] The server periodically retrieves the behavioral data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's daily behavioral patterns. Example: aiModel.fit(trainingData). The input data is past behavioral data, and the output data is the user's daily behavioral pattern profile.

[0471] Step 5:

[0472] Receiving disaster information

[0473] The server works with the disaster information service to receive disaster information in real time. Example: disasterAlertService.receiveAlert(). The input data is disaster information, and the output data is the received disaster information.

[0474] Step 6:

[0475] Identifying target areas

[0476] The server identifies the disaster-affected area based on the received disaster information and extracts users who are in that area from the database. The input data is the disaster information and the user's location information, and the output data is a list of users in the identified target area.

[0477] Step 7:

[0478] Collection of the latest data

[0479] When disaster information is received, the device immediately collects the current location information, the last update time, information about nearby users, signal reception status, and emotion data. (Example: LocationServices.requestLocationUpdate(), HeartbeatSensor.getLatestReading()) The input data is the disaster information, and the output data is the latest collected user data.

[0480] Step 8:

[0481] Sending data

[0482] Quickly send the latest data from the device to the server. Example: HttpClient.sendPostRequest(latestData). The input data is the latest user data collected, and the output data is the latest data sent to the server.

[0483] Step 9:

[0484] Calculating the risk

[0485] The server compares the latest data sent from the device with the stored daily behavior pattern profile. The AI ​​evaluates the differences between these data and calculates the risk level. Example: riskLevel = aiModel.predict(latestData). The input data is the latest user data and daily behavior pattern profile, and the output data is the risk assessment result.

[0486] Step 10:

[0487] Emotional Data Analysis

[0488] The server analyzes the emotion data and reflects the stress level and tension in the risk assessment. Example: emotionAnalysisEngine.analyze(emotionData). The input data is emotion data, and the output data is the emotion data analysis result.

[0489] Step 11:

[0490] Issuance of safety confirmation alerts

[0491] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts. Example: NotificationService.sendAlert(contactInfo, alertDetails). The input data is the user's risk assessment result and contact information, and the output data is the sent alert.

[0492] At each step, the process flow is clear, detailing how the input data is processed or calculated and what the output is, making it easy to understand how the entire system works.

[0493] 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.

[0494] 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.

[0495] 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.

[0496] [Second embodiment]

[0497] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0498] 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.

[0499] 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).

[0500] 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.

[0501] 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.

[0502] 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).

[0503] 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.

[0504] 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.

[0505] 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.

[0506] 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.

[0507] 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.

[0508] 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."

[0509] This invention is an automatic safety confirmation system for use in disasters, which allows users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs.

[0510] Overall system configuration

[0511] This system mainly consists of the following elements:

[0512] 1. A means of collecting and storing users' daily behavior patterns

[0513] 2. Means of receiving disaster information in real time

[0514] 3. A method for identifying users in target areas based on disaster information

[0515] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[0516] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[0517] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[0518] Specific embodiments of each means will be described in detail below.

[0519] Collecting and storing users' daily behavior patterns

[0520] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0521] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[0522] Learning daily behavior patterns

[0523] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[0524] Receiving disaster information and identifying affected areas

[0525] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[0526] Data collection in the event of a disaster

[0527] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[0528] Comparison of the latest data and calculation of risk

[0529] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[0530] Issuance of safety confirmation alerts

[0531] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, and signal reception status.

[0532] Specific examples

[0533] For example, if user A uses a smartphone on a daily basis, location information and activity data are sent from the device to the server. The server receives and stores this data, and AI learns user A's daily behavior patterns.

[0534] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User A is in that area.

[0535] User A's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, and radio wave reception status to the server. The server analyzes this data and calculates the risk level based on any differences with User A's behavioral patterns.

[0536] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User A's registered contacts (e.g., family and workplace). This is expected to enable User A's safety to be confirmed quickly and accurately, and for necessary rescue operations to be carried out promptly.

[0537] The processing flow will be explained below.

[0538] Step 1:

[0539] The device collects data on the user's daily behavior (location information, activity information, app usage, health data, etc.).

[0540] Step 2:

[0541] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[0542] Step 3:

[0543] The server receives the data sent from the terminal and stores it in a database.

[0544] Step 4:

[0545] The server inputs the stored data into an AI model that learns the user's daily behavior patterns.

[0546] Step 5:

[0547] The server generates and updates each user's behavioral pattern profile based on the learning results.

[0548] Step 6:

[0549] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from the disaster information service.

[0550] Step 7:

[0551] The server identifies the disaster-affected area based on the disaster information received, and extracts users in that area from the database.

[0552] Step 8:

[0553] The terminal receives disaster information and instantly collects current location information, last update time, information on nearby users, and radio wave reception status.

[0554] Step 9:

[0555] The latest data collected by the device is sent to the server.

[0556] Step 10:

[0557] The server compares the latest data received from the device with the daily behavior pattern profile, and the AI ​​calculates the level of risk.

[0558] Step 11:

[0559] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[0560] Step 12:

[0561] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, and radio wave reception status.

[0562] Example 1

[0563] 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."

[0564] In the event of a disaster, prompt and accurate safety confirmation is extremely important for saving many lives. However, manual safety confirmation requires time and effort, and in the event of a disaster, communication infrastructure may be destroyed or disrupted, making it difficult to respond quickly. To address these issues, the present invention aims to provide a system that uses AI to automatically confirm the safety of users and quickly send necessary notifications.

[0565] 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.

[0566] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the latest collected data with daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for transmitting the generated data to the server and analyzing it with an AI model, and means for updating the behavior pattern profile based on the daily behavior patterns. This makes it possible to quickly and accurately confirm the safety of users when a disaster occurs.

[0567] "User's daily behavioral patterns" is a general term for data related to the user's daily actions, movements, and activities.

[0568] "Disaster information" refers to real-time information about disasters such as earthquakes, weather warnings, fires, and large-scale accidents.

[0569] The "target area" refers to the area where a disaster has occurred, as identified based on disaster information.

[0570] "Latest user data" refers to data such as the user's location information at the time of the disaster, the last update time, information about nearby users, and radio wave reception conditions.

[0571] The "risk level" is a value that indicates the degree of risk to the user, calculated by comparing the latest collected data with daily behavior patterns.

[0572] A "safety confirmation alert" is a warning notification issued to a notification destination registered in advance when the user's risk level exceeds a certain threshold.

[0573] "AI model" refers to artificial intelligence algorithms and systems that analyze collected data and learn users' daily behavioral patterns.

[0574] A "behavioral pattern profile" is a data set that shows the characteristics of a user's unique daily behavior, generated based on the results of learning by an AI model.

[0575] The "Disaster Information Service" is a system that distributes real-time disaster-related information provided by organizations such as the Japan Meteorological Agency.

[0576] The present invention is an automatic safety confirmation system for use in the event of a disaster, allowing users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs. Each element of the present invention is described in detail below.

[0577] This system mainly consists of the following elements:

[0578] Collecting and storing users' daily behavior patterns

[0579] As the user uses their smartphone on a daily basis, the device collects location information, app and web usage, activity information (walking, driving, standing still), and health data (number of steps, heart rate, etc.). This data is sent to a server at regular intervals (e.g., every 15 minutes). The hardware used is assumed to be a smartphone (e.g., Android device or iPhone). The collected data is encrypted by the device using HTTPS and securely sent to the server.

[0580] Sending data to the server

[0581] The server receives the data sent from the device and stores it in a database. This allows the user's daily behavioral patterns to be accumulated and prepared for later AI analysis. The software used is a database management system (e.g., MySQL or PostgreSQL) and an AI framework for data analysis (e.g., TensorFlow or PyTorch).

[0582] Learning daily behavior patterns

[0583] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated. The Python programming language and related libraries (e.g., NumPy, Pandas) are used to train the AI ​​model.

[0584] Receiving disaster information

[0585] The server works in conjunction with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, and large-scale accidents. For example, disaster information is sent to the server via an API, and the data is analyzed immediately. A real-time data processing framework (e.g., Apache Kafka) is used for disaster information management.

[0586] Identifying target areas and collecting data

[0587] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database.The device then receives the disaster information and immediately collects the current location information, the last update time, information about nearby users (via short-range communication), and the radio wave reception status, and promptly transmits this data to the server.

[0588] Comparison of the latest data and calculation of risk

[0589] The server compares the latest data sent from the device with the daily behavioral pattern profile stored in the database. The AI ​​evaluates the differences between these data and calculates the level of risk. If the level of risk exceeds a certain threshold, it detects it as an abnormal pattern.

[0590] Issuance of safety confirmation alerts

[0591] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status. Notifications are sent via email, SMS, push notifications, etc.

[0592] Specific examples

[0593] For example, user A uses a smartphone on a daily basis, and location and activity data is sent from the device to a server. The server stores the received data, and AI learns user A's daily behavioral patterns based on data such as commuting route, activity times, and places of stay.

[0594] One day, a magnitude 5.0 earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and checks the database to see if User A is in that area. User A's device receives the earthquake information and immediately collects and sends this information to the server, along with the user's current location, last update time, information about nearby users, and signal reception status. The server uses this data to compare it with normal activity patterns and calculates User A's risk level. If the risk level exceeds a threshold, the server automatically sends a safety alert and detailed information to User A's registered contacts (e.g., family and workplace).

[0595] Example prompts for generative AI models

[0596] Generate code to program an automatic disaster safety confirmation system described in natural language. The system collects users' daily behavior from their smartphones and sends the data to a server. The server uses AI to learn daily behavior patterns and receives and analyzes disaster information in real time. When a disaster occurs, it collects the latest user data, calculates the risk level, and automatically sends safety confirmation alerts as needed.

[0597] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0598] Step 1:

[0599] Collecting users' daily behavior patterns

[0600] The device collects data on the user's daily activities. Specifically, it uses the smartphone's GPS function, app usage, activity sensors (accelerometer and gyroscope), and health data (pedometer and heart rate monitor) to collect location information, app usage history, movement status (walking, driving, stationary, etc.), and vital signs at regular intervals (e.g., every 15 minutes).

[0601] Input: Data from various sensors and apps on the user's smartphone

[0602] Output: A set of user's daily behavior data (location information, activity information, health data, etc.)

[0603] Step 2:

[0604] Sending data to the server

[0605] The device sends the collected daily behavior data to a server at regular intervals. The data is encrypted using the HTTPS protocol and sent securely to the server.

[0606] Input: Collected daily behavior data

[0607] Output: Encrypted daily activity data packet sent to the server

[0608] Step 3:

[0609] Saving daily behavior patterns

[0610] The server receives the daily behavior data sent from the device and stores it in a database, which is later used for AI analysis.

[0611] Input: Encrypted daily activity data sent from the device

[0612] Output: Daily behavior data stored in a database

[0613] Step 4:

[0614] Learning daily behavior patterns

[0615] The server periodically inputs the daily behavior data stored in the database into the AI ​​model for analysis. The AI ​​model analyzes this data and learns the user's usual behavior patterns. Based on the learning results, a behavioral pattern profile for each user is generated and regularly updated.

[0616] Input: Daily behavior data stored in a database

[0617] Output: Generated and updated behavioral pattern profile

[0618] Step 5:

[0619] Receiving disaster information

[0620] The server works with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. For example, it receives earthquake occurrence information from disaster information services via API.

[0621] Input: Disaster information from various disaster information services

[0622] Output: Disaster information data immediately stored on the server

[0623] Step 6:

[0624] Identifying the target area

[0625] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database. For example, it analyzes the latitude and longitude information of the earthquake area and identifies users within that range.

[0626] Input: Received disaster information, user location information stored in the database

[0627] Output: List of users in the identified disaster area

[0628] Step 7:

[0629] Collection of the latest data

[0630] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information about nearby users (via short-range communication such as Bluetooth), and radio wave reception status, and quickly transmits this data to the server.

[0631] Input: Disaster information, data from various sensors and communication devices on the user's smartphone

[0632] Output: Latest location information, last update time, information on nearby users, data packets of signal reception status

[0633] Step 8:

[0634] Calculating the risk

[0635] The server compares the latest data sent from the device with the stored daily behavioral pattern profile, and the AI ​​calculates the risk level. For example, if the user's location does not change for a long period of time even though it is during a normal activity time, it will determine this as an abnormality.

[0636] Input: Latest data sent from the device, saved behavioral pattern profile

[0637] Output: Calculated risk

[0638] Step 9:

[0639] Issuance of safety confirmation alerts

[0640] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status.

[0641] Input: Calculated risk level, user's registered contact information

[0642] Output: Safety confirmation alert sent and detailed information

[0643] (Application example 1)

[0644] 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."

[0645] In modern society, the occurrence of natural disasters and large-scale accidents is an unavoidable problem. Quickly and accurately confirming the safety of people when a disaster occurs is an important issue that directly relates to ensuring the safety of victims and streamlining rescue efforts. However, current safety confirmation methods involve a lot of manual confirmation, making it difficult to grasp safety in a timely manner. Furthermore, they are insufficient in proposing appropriate evacuation routes in the event of a disaster. To solve these issues, a system is needed that can more efficiently and automatically confirm safety and also propose appropriate evacuation routes.

[0646] 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.

[0647] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the collected latest data with the daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for automatically suggesting evacuation routes for high-risk users, and means for calculating the level of risk and confirming the safety of users based on prompt text from the generation AI model. This makes it possible to quickly and automatically confirm the safety of users and suggest appropriate evacuation routes when a disaster occurs.

[0648] "User's daily behavioral patterns" refer to the user's daily patterns of movement and activities, and include location information, app and web usage, activity information, health care data, etc.

[0649] "Means for receiving disaster information in real time" refers to a system that works in conjunction with disaster information services provided by the Japan Meteorological Agency and other organizations, and has the function of receiving disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. in real time.

[0650] The "means for identifying users in the target area" has a function of extracting users who are in the disaster-affected area from a database based on the received disaster information.

[0651] The "means for collecting the latest data" has the function of collecting information such as the user's location, the last update time, information about nearby users, and radio wave reception conditions when a disaster occurs.

[0652] The "means for calculating the risk level" has a function of comparing the latest collected data with the saved daily behavior patterns, evaluating the difference, and calculating the risk level of the user.

[0653] The "means for issuing safety confirmation alerts to notification recipients" has the function of automatically sending safety confirmation alerts to notification recipients registered in advance by the user when the risk level exceeds a certain threshold.

[0654] The "means for automatically suggesting evacuation routes" is a function that automatically suggests appropriate evacuation routes based on AI analysis to users at high risk.

[0655] The means for checking safety based on the "prompt sentence of the generative AI model" has the function of inputting the prompt sentence into the AI ​​model, calculating the user's level of risk, and checking the user's safety.

[0656] This invention is a system that automatically checks the user's safety during a disaster and notifies emergency contacts. The system uses AI to learn the user's daily behavior patterns, and when a disaster occurs, it compares them with the user's latest data to calculate the level of risk. In addition, if the risk is high, the system has the function of suggesting an appropriate evacuation route.

[0657] Hardware and software configuration

[0658] The hardware and software used to realize the system of the present invention are as follows.

[0659] Hardware: Smartphones, servers

[0660] Software: Python, requests (library for calling APIs), geopy (geographic calculation), smtplib (for sending emails)

[0661] External API: Real-time disaster information API for disaster information provision services

[0662] Overall system configuration

[0663] Safety confirmation system configuration

[0664] 1. Data Collection:

[0665] The server collects the user's daily behavioral patterns. The smartphone periodically collects data such as the user's location, activity, and signal reception status, and sends it to the server. The server stores this data and later analyzes it using AI.

[0666] 2. Receiving disaster information:

[0667] The server works in conjunction with disaster information services to receive disaster information in real time. Information on earthquakes, weather warnings, fires, large-scale accidents, etc. is sent to the server. Based on this information, the area where the disaster occurred is identified.

[0668] 3. Identifying users in the target area:

[0669] The server extracts users who are in the disaster area from the database, thereby identifying users who may be affected by the disaster.

[0670] 4. Collecting the latest data:

[0671] When a disaster occurs, the smartphone collects the user's latest location information, activity information, and signal reception status, and immediately sends it to a server, making it possible to check the safety of users in real time.

[0672] 5. Data comparison and risk calculation:

[0673] The server uses AI to compare the latest data with daily behavior patterns. AI analysis evaluates the differences between the collected data and normal behavior patterns and calculates the user's risk level.

[0674] 6. Issuance of Safety Confirmation Alerts:

[0675] If the specified risk threshold is exceeded, a safety confirmation alert is automatically sent to the user's emergency contacts, allowing the necessary response to be taken promptly.

[0676] 7. Suggested evacuation routes:

[0677] For users at high risk, evacuation routes are automatically suggested based on AI analysis. By providing appropriate evacuation routes, users' safety is supported.

[0678] Specific example explanation

[0679] Assume that user A travels daily in central Tokyo and its suburbs. The invention collects user A's location information and activity data, and the server learns his daily behavior patterns based on this data. One day, an earthquake occurs, and data from a disaster information service is sent to the server. If it is confirmed that user A is included in the area affected by the earthquake, the server collects user A's latest location information and activity data at that time and compares it with his daily behavior patterns. If the analysis determines that there is a high risk, the server issues a safety confirmation alert to user A's family and workplace. Furthermore, evacuation routes are automatically suggested to user A's smartphone.

[0680] Prompt Sentence Examples

[0681] Use the following prompt for the generative AI model:

[0682] Please calculate the risk level based on the daily behavior pattern data of user ID: user123, comparing it with the latest information (location information, activity data, etc.) in the event of a disaster. Provide the following data:

[0683] Daily behavior pattern data: [Detailed data]

[0684] Latest data: [Detailed data]

[0685] In this way, the present invention automatically checks the user's safety during a disaster, quickly notifies emergency contacts, and supports necessary responses, thereby ensuring the safety of victims and improving the efficiency of rescue operations.

[0686] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0687] Step 1:

[0688] Collect user location and activity data.

[0689] Input: Sensor data such as location information, activity data, and signal reception status from the user's smartphone.

[0690] Processing: The smartphone collects this data at regular intervals, converts it into a data format, and sends it to the server.

[0691] Output: A set of user location and activity data sent to the server.

[0692] Step 2:

[0693] The server stores the user's daily behavior patterns.

[0694] Input: Location, activity data, and other sensor data sent from your smartphone.

[0695] Processing: The server stores the received data in a database and integrates it with past data, thereby accumulating daily behavior patterns.

[0696] Output: User's daily behavior pattern data stored in a database.

[0697] Step 3:

[0698] Receive disaster information in real time.

[0699] Input: Real-time data from disaster information services (earthquake occurrence information, weather warnings, fire information, large-scale accident information, etc.).

[0700] Processing: The server calls the API and obtains disaster information.

[0701] Output: Disaster information data stored on the server.

[0702] Step 4:

[0703] Identify users in your target area.

[0704] Input: Disaster information (data on the location and scale of the disaster), user's daily behavior pattern data (location information).

[0705] Processing: The server analyzes the disaster area and extracts users who are in that area from the database.

[0706] Output: A list of users in the disaster area.

[0707] Step 5:

[0708] Collect the latest data when a disaster occurs.

[0709] Input: Disaster notification, latest location information of the user's smartphone, activity data, and signal reception status.

[0710] Processing: Upon receiving disaster information, the smartphone immediately collects the latest sensor data and sends it to the server.

[0711] Output: The latest sensor data sent to the server.

[0712] Step 6:

[0713] The latest data is compared with daily behavior patterns to calculate the level of risk.

[0714] Input: Latest sensor data, daily behavior pattern data.

[0715] Processing: The server uses the generated AI model to compare the latest data with daily behavior patterns and calculate the user's risk level. The prompt sentence is input into the model and the analysis results are obtained.

[0716] Output: Calculated user risk level (levels such as normal, caution, danger, etc.).

[0717] Step 7:

[0718] If the risk level exceeds a certain threshold, an alert will be issued to check the safety of the person.

[0719] Input: Risk data, user's emergency contact information.

[0720] Processing: If the server detects a risk level that exceeds the threshold, it will automatically send an email or SMS to the user's emergency contacts to confirm their safety.

[0721] Output: Safety alert sent to emergency contacts.

[0722] Step 8:

[0723] Automatically suggests evacuation routes to users at high risk.

[0724] Input: User's current location, risk level data, and surrounding environment information.

[0725] Processing: The server uses AI to analyze surrounding geographical information and disaster information to generate an appropriate evacuation route, which is then sent to the user's smartphone.

[0726] Output: Evacuation route displayed on the user's smartphone.

[0727] 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.

[0728] This invention is an automatic safety confirmation system for use in disaster situations, and by combining it with an emotion engine that recognizes the user's emotions, it is possible to perform more accurate risk assessments. This system uses AI to analyze smartphone usage data, and aims to automatically confirm and notify the user's safety when a disaster occurs.

[0729] Overall system configuration

[0730] The system consists of the following main elements:

[0731] 1. A means of collecting and storing users' daily behavior patterns

[0732] 2. Means of receiving disaster information in real time

[0733] 3. A method for identifying users in target areas based on disaster information

[0734] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[0735] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[0736] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[0737] 7. Emotion engine that recognizes user emotions

[0738] Specific embodiments of each means will be described in detail below.

[0739] Collecting and storing users' daily behavior patterns

[0740] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0741] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[0742] Learning daily behavior patterns

[0743] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[0744] Receiving disaster information and identifying affected areas

[0745] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[0746] Data collection in the event of a disaster

[0747] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[0748] Comparison of the latest data and calculation of risk

[0749] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[0750] Use of emotion engine

[0751] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions, voice, and actions, and sends it to the server.

[0752] The server uses the emotion data sent by the user to assess the risk level. The emotion data indicates the user's stress level and tension, and the accuracy of the risk level is improved based on this. For example, if a user is in an abnormally stressed state when a disaster occurs, this information is reflected in the risk level calculation.

[0753] Issuance of safety confirmation alerts

[0754] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[0755] Specific examples

[0756] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives and stores this data, and the AI ​​learns user B's daily behavioral patterns.

[0757] One day, an earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area, and extracts from the database whether User B is in that area.

[0758] User B's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, signal reception status, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[0759] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family or workplace). This is expected to enable User B's safety to be confirmed quickly and accurately, and to enable necessary rescue operations to be carried out promptly.

[0760] The processing flow will be explained below.

[0761] Step 1:

[0762] The device collects data on the user's daily activities (location information, activity information, app usage, health data, and emotional data). Emotional data is collected by an emotion engine that recognizes the user's facial expressions, voice, and behavior.

[0763] Step 2:

[0764] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[0765] Step 3:

[0766] The server receives the data sent from the terminal and stores it in a database.

[0767] Step 4:

[0768] The server inputs the stored data into the AI ​​model, which learns the user's daily behavioral patterns and emotional data. Based on the learning results, a behavioral pattern profile for each user is generated and updated.

[0769] Step 5:

[0770] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from various disaster information services.

[0771] Step 6:

[0772] The server identifies the disaster-affected area based on the disaster information received, and extracts users who are in that area from the database.

[0773] Step 7:

[0774] When the device receives disaster information, it immediately collects the current location information, the last update time, information about nearby users, radio wave reception conditions, and emotional data.

[0775] Step 8:

[0776] The latest data collected by the device is sent to the server.

[0777] Step 9:

[0778] The server compares the latest data received from the device with the daily behavioral pattern profile and emotional data, and the AI ​​calculates the risk level, taking into account the user's emotional data (e.g., stress level and tension).

[0779] Step 10:

[0780] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[0781] Step 11:

[0782] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, radio wave reception status, and emotional data.

[0783] Example 2

[0784] 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."

[0785] In the event of a disaster, it is extremely important to quickly and accurately confirm the safety of users and take appropriate action. However, conventional systems do not take into account the user's daily behavioral patterns or emotional state in risk assessment, resulting in low accuracy in safety confirmation. The present invention aims to solve these problems.

[0786] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest data of users (location information, last update time, information on surrounding users, and communication environment) when a disaster occurs, means for comparing the collected latest data with daily behavior patterns to calculate a risk level, means for recognizing user emotion data and incorporating it into risk assessment, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables highly accurate risk assessment that takes into account the user's behavior patterns and emotional state.

[0787] "Means for collecting and storing users' daily behavioral patterns" refers to means for collecting users' daily activities, location information, app usage, and activity information, and storing this in a database.

[0788] "Means for receiving disaster information in real time" refers to means for obtaining disaster information such as earthquakes, weather, fires, and large-scale accidents in real time and transmitting it to a server.

[0789] The "means for identifying users in a target area based on received disaster information" is a means for identifying a disaster-affected area from received disaster information and extracting users who are in that area from a database.

[0790] "Means for collecting the latest user data (location information, last update time, surrounding user information, communication environment) when a disaster occurs" refers to means for collecting the user's current location information, last update time, user information via surrounding communications, and the status of radio waves and the communication environment when a disaster occurs.

[0791] The "means for calculating the risk by comparing the latest collected data with daily behavior patterns" is a means for comparing the latest collected data with the user's saved daily behavior pattern profile, evaluating the difference, and calculating the risk.

[0792] "Means for recognizing the user's emotional data and incorporating it into risk assessment" refers to means for collecting and analyzing data on the user's emotional state (e.g., facial expressions and voice) and reflecting it in risk assessment.

[0793] "Means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold" refers to a means for automatically sending a safety confirmation alert to the user's registered emergency contacts (family, workplace, etc.) when the calculated risk level exceeds a pre-set threshold.

[0794] This invention is an automatic safety confirmation system for use in the event of a disaster, which performs more accurate risk assessment based on the user's daily behavioral patterns and emotional data. The system uses AI to analyze smartphone usage data and automatically confirms and notifies the user's safety in the event of a disaster.

[0795] System Configuration

[0796] The system consists of the following main elements:

[0797] 1. A means of collecting and storing users' daily behavior patterns

[0798] The device collects data on the user's daily behavior using the smartphone's GPS, health sensors, app usage logs, camera, microphone, etc. This data includes location information, activity data (walking, driving, and standing still), heart rate, and app usage.

[0799] The collected data is sent to the server at regular intervals (for example, every 15 minutes) and stored on the server.

[0800] 2. A means of learning daily behavior patterns

[0801] The server periodically retrieves daily behavior data stored in the database and analyzes it using a generative AI model, primarily using deep learning frameworks such as TensorFlow and PyTorch.

[0802] The AI ​​model learns the user's usual behavioral patterns (for example, daily commuting route and activity times) and generates and updates a behavioral pattern profile based on this.

[0803] 3. Means of receiving disaster information and identifying affected areas

[0804] The server works in conjunction with various disaster information services (e.g., the Japan Meteorological Agency's API) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents.

[0805] When disaster information is received, the area where the disaster occurred is identified based on that information, and users who are in that area are extracted from the database.

[0806] 4. How to collect the latest user data when a disaster occurs

[0807] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information on nearby users (via short-range communication), and the communication environment.

[0808] This data is quickly transmitted to the server.

[0809] 5. Comparing the latest data with daily behavior patterns and calculating the risk level

[0810] The server compares the latest data sent from the device with the stored daily behavior pattern profile.

[0811] The generative AI evaluates the differences in this data and calculates the risk level, detecting anomalies based on deviations from normal behavioral patterns.

[0812] 6. Methods for Recognizing User Emotional Data and Incorporating It into Risk Assessment

[0813] The device collects emotional data from the user's facial expressions, voice, and actions, and sends it to a server using a camera and microphone.

[0814] The server performs a risk assessment based on the emotional data analyzed by the emotion engine, taking into account the user's stress level and tension.

[0815] 7. How to issue safety confirmation alerts

[0816] If the calculated risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace).

[0817] The alert includes the user's last location, last update time, information about nearby users, communication environment, emotional data, etc.

[0818] Specific examples

[0819] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives this data and stores it in a database, and the generative AI model learns user B's daily behavioral patterns.

[0820] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User B is in that area.

[0821] User B's device receives the earthquake information and immediately sends its current location, last update time, information about nearby users, communication environment, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[0822] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family, workplace). This process is expected to quickly and accurately confirm User B's safety, and enable necessary rescue operations to be carried out promptly.

[0823] Prompt Sentence Examples

[0824] "Please explain in detail your automated safety confirmation system in the event of a disaster. Please tell us the specific processing steps, from data collection to risk assessment and safety confirmation alerts. Please also include examples of application in real-life scenarios."

[0825] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0826] Processing Steps

[0827] Step 1: Data collection

[0828] The device collects information about the user's daily activities, including location information using GPS, activity information obtained from smartphone sensors (e.g., walking, driving, standing still), app usage, and heart rate.

[0829] Input: User's smartphone usage, sensor information

[0830] Output: Collected daily behavior data (e.g., location information, heart rate)

[0831] Specific operation: The device activates the sensor every 15 minutes and packages the acquired data in JSON format.

[0832] Step 2: Send data

[0833] The device sends the collected daily behavior data to the server at regular intervals (e.g., every 15 minutes).

[0834] Input: Collected daily behavior data

[0835] Output: Data sent to the server

[0836] Specific operation: The device sends JSON formatted data to the server's API endpoint using HTTPS.

[0837] Step 3: Save Data

[0838] The server receives the daily activity data sent from the terminal and stores it in a database.

[0839] Input: Daily activity data sent from the device

[0840] Output: Behavioral data stored in a database

[0841] Specific behavior: The server parses the received JSON data and stores it in the database in the appropriate format.

[0842] Step 4: Learning behavioral patterns

[0843] The server periodically retrieves the behavioral data stored in the database, inputs it into the generative AI model, and analyzes it.

[0844] Input: Saved daily activity data

[0845] Output: Learned behavioral pattern profile

[0846] Specific operation: The server uses TensorFlow and PyTorch to run a behavior prediction model and learn user behavior patterns.

[0847] Step 5: Receiving disaster information

[0848] The server receives information in real time from the disaster information service.

[0849] Input: Data from disaster information service

[0850] Output: Disaster information received by the server

[0851] Specific operation: The server periodically checks the API endpoint to obtain the latest disaster information.

[0852] Step 6: Identify users in your target area

[0853] The server identifies the disaster-affected area from the received disaster information and extracts users within that area.

[0854] Input: Disaster information, user location information in the database

[0855] Output: A list of users within the identified target area

[0856] Specific operation: The server uses GIS to map the disaster area and extracts users in that area from the database.

[0857] Step 7: Collect the latest data

[0858] When the terminal receives disaster information, it collects the current location information, the last update time, information on nearby users, and the communication environment.

[0859] Input: Disaster information, device sensor information

[0860] Output: Latest collected data (location information, communication environment, etc.)

[0861] What it does: The device performs GPS and Bluetooth scans to collect and send the latest data.

[0862] Step 8: Data comparison and risk calculation

[0863] The server compares the latest data sent from the device with the saved daily behavior patterns and calculates the level of risk.

[0864] Input: Latest data, daily behavior pattern profile

[0865] Output: Calculated risk

[0866] Specific operation: The server uses a generative AI model to detect abnormalities in behavioral patterns and evaluates the level of risk based on the results.

[0867] Step 9: Collect and evaluate emotion data

[0868] The terminal collects the user's emotional data (facial expressions, voice, etc.) and sends it to the server.

[0869] Input: User's facial expression and voice data

[0870] Output: Parsed emotion data

[0871] Specific operation: The device uses the camera and microphone to capture emotion data and sends it to the server.

[0872] Step 10: Finalize your risk assessment

[0873] The server recalculates the risk assessment based on the emotion data.

[0874] Input: Emotion data, initial risk assessment

[0875] Output: Adjusted risk assessment

[0876] Specific operation: The server takes in the emotional data, reassess the danger level, and generates the final result.

[0877] Step 11: Issue an alert

[0878] If the calculated risk level exceeds a threshold, the server sends a safety confirmation alert to the registered destination of the target user.

[0879] Input: Adjusted risk assessment, user registration information

[0880] Output: Safety confirmation alert issued

[0881] Specific operation: The server generates an alert message containing the necessary information (e.g., last location, nearby user information, etc.) and sends it to the registered recipient via email or SMS.

[0882] In this way, the system takes into account the user's daily behavior and emotional state, enabling quick and highly accurate confirmation of safety during a disaster.

[0883] (Application example 2)

[0884] 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."

[0885] Although systems for confirming users' safety during disasters already exist, they typically rely on users to provide information voluntarily, which often results in inaccurate information being unavailable. Furthermore, these systems assess risk based solely on location and activity data, without taking into account the user's emotional state, limiting the accuracy of risk assessments. Therefore, there is a need for a system that can more quickly and accurately confirm users' safety and provide appropriate notifications.

[0886] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in the target area based on the received disaster information, means for collecting user emotion data and using it for risk assessment, means for calculating the risk level by comparing the latest collected data with daily behavior patterns, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables more accurate risk assessment and faster safety confirmation by analyzing the user emotion data and reflecting the stress level and tension level in the risk assessment.

[0887] "User's daily behavior patterns" refers to the totality of data such as the user's daily actions, movements, and app usage.

[0888] "Disaster information" refers to information about emergencies such as earthquakes, fires, weather warnings, and large-scale accidents.

[0889] "Target area" refers to the geographical area identified as being affected by a disaster based on disaster information.

[0890] The "latest data" is real-time data including the user's location information, the last update time, information about nearby users, radio wave reception conditions, and emotion data at the time of the disaster.

[0891] The "risk level" is a numerical value or classification of the degree of risk in the situation in which the user is placed.

[0892] "Emotion data" is data that indicates the emotional state of the user as recognized from facial expressions, tone of voice, and the like.

[0893] "Risk assessment" is the process of calculating a user's risk level based on collected data.

[0894] A "safety confirmation alert" is a message or notification sent to a contact to confirm whether the user is safe.

[0895] The present invention is an automatic safety confirmation system for use in the event of a disaster, which collects data on users' daily behavior patterns and emotions and performs highly accurate risk assessments in the event of a disaster. Hereinafter, embodiments of the present invention will be described in detail.

[0896] Overall system configuration

[0897] This system is mainly composed of three elements: a server, a terminal, and a user.

[0898] Collecting and storing daily behavior patterns

[0899] As users use their smartphones on a daily basis, the device collects location information, activity data (e.g., walking, driving, and standing still), app usage, and emotional data (e.g., facial expressions and tone of voice). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0900] Learning daily behavior patterns

[0901] The server periodically retrieves daily behavior data stored in the database and inputs it into an AI model (e.g., TensorFlow, PyTorch). The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route, activity times). Based on the learning results, a behavioral pattern profile is generated and continuously updated.

[0902] Receiving disaster information and identifying affected areas

[0903] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the server identifies the area where the disaster occurred and extracts users who are in that area from the database.

[0904] Data collection in the event of a disaster

[0905] When a device receives disaster information, it immediately collects its current location, the last update time, information about nearby users (e.g., device detection via short-range communication), and the status of radio wave reception. This data is then promptly sent to the server. The device also collects emotion data and sends it to the server.

[0906] Comparison of the latest data and calculation of risk

[0907] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change even during normal activity times, it will be detected as an abnormal pattern.

[0908] Use of emotion engine

[0909] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions and tone of voice and sends it to a server. The server analyzes the emotion data and reflects stress levels and tension in the risk assessment, improving the accuracy of the risk assessment.

[0910] Issuance of safety confirmation alerts

[0911] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[0912] Specific examples

[0913] For example, when a user uses a smartphone on a daily basis, location information, app usage status, and emotional data are sent from the device to a server. The server receives and stores this information, and AI learns the user's daily behavioral patterns. One day, an earthquake occurs. The server receives disaster information, confirms that the disaster occurred in a specific area, and extracts from the database whether the user is in that area. The device receives the earthquake information and immediately collects current location information, last update time, information about surrounding users, signal reception status, and emotional data, and sends them to the server. The server analyzes this data and calculates the level of risk based on the difference between the behavioral pattern and the emotional data. If the level of risk exceeds a threshold, the server automatically sends a safety confirmation alert and detailed information to the user's registered contacts.

[0914] Example prompts for generative AI models

[0915] Enter the user's face photo and voice data:

[0916] Facial photo data:[path_to_image]\

[0917] Audio data:[path_to_audio]\

[0918] \

[0919] Parse the following data attributes:

[0920] 1. Emotion recognition (e.g., joy, anxiety, anger, sadness)

[0921] 2. Tone of voice (e.g., nervous, stressed, calm)

[0922] \

[0923] As output, return the user's emotional state and its associated numerical data (e.g., stress level 0-100).

[0924] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0925] Step 1:

[0926] Data collection

[0927] The device collects the user's daily behavioral patterns. Specifically, it obtains location information, activity data, app usage, and emotional data (facial expressions, tone of voice) using sensors, microphones, and cameras. For example, it uses APIs such as LocationServices.requestLocationUpdates(), ActivityRecognitionClient.requestActivityUpdates(), Camera.captureImage(), and Microphone.recordAudio(). The input data is various types of user behavioral information, and the output data is the collected behavioral data.

[0928] Step 2:

[0929] Data transmission

[0930] The collected data is sent to the server at regular intervals (for example, every 15 minutes). This involves sending a POST request using an HTTP client. For example, HttpClient.sendPostRequest(data). The input data is the behavioral data collected on the device, and the output data is the behavioral data sent to the server.

[0931] Step 3:

[0932] Data storage

[0933] The server stores the received data in a database. Specifically, it executes an INSERT statement in the database using SQL or similar. Example: Database.insertData(data). The input data is the behavioral data sent, and the output data is the behavioral data stored in the database.

[0934] Step 4:

[0935] Learning behavioral patterns

[0936] The server periodically retrieves the behavioral data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's daily behavioral patterns. Example: aiModel.fit(trainingData). The input data is past behavioral data, and the output data is the user's daily behavioral pattern profile.

[0937] Step 5:

[0938] Receiving disaster information

[0939] The server works with the disaster information service to receive disaster information in real time. Example: disasterAlertService.receiveAlert(). The input data is disaster information, and the output data is the received disaster information.

[0940] Step 6:

[0941] Identifying target areas

[0942] The server identifies the disaster-affected area based on the received disaster information and extracts users who are in that area from the database. The input data is the disaster information and the user's location information, and the output data is a list of users in the identified target area.

[0943] Step 7:

[0944] Collection of the latest data

[0945] When disaster information is received, the device immediately collects the current location information, the last update time, information about nearby users, signal reception status, and emotion data. (Example: LocationServices.requestLocationUpdate(), HeartbeatSensor.getLatestReading()) The input data is the disaster information, and the output data is the latest collected user data.

[0946] Step 8:

[0947] Sending data

[0948] Quickly send the latest data from the device to the server. Example: HttpClient.sendPostRequest(latestData). The input data is the latest user data collected, and the output data is the latest data sent to the server.

[0949] Step 9:

[0950] Calculating the risk

[0951] The server compares the latest data sent from the device with the stored daily behavior pattern profile. The AI ​​evaluates the differences between these data and calculates the risk level. Example: riskLevel = aiModel.predict(latestData). The input data is the latest user data and daily behavior pattern profile, and the output data is the risk assessment result.

[0952] Step 10:

[0953] Emotional Data Analysis

[0954] The server analyzes the emotion data and reflects the stress level and tension in the risk assessment. Example: emotionAnalysisEngine.analyze(emotionData). The input data is emotion data, and the output data is the emotion data analysis result.

[0955] Step 11:

[0956] Issuance of safety confirmation alerts

[0957] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts. Example: NotificationService.sendAlert(contactInfo, alertDetails). The input data is the user's risk assessment result and contact information, and the output data is the sent alert.

[0958] At each step, the process flow is clear, detailing how the input data is processed or calculated and what the output is, making it easy to understand how the entire system works.

[0959] 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.

[0960] 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.

[0961] 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.

[0962] [Third embodiment]

[0963] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0964] 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.

[0965] 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).

[0966] 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.

[0967] 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.

[0968] 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).

[0969] 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.

[0970] 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.

[0971] 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.

[0972] 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.

[0973] 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.

[0974] 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."

[0975] This invention is an automatic safety confirmation system for use in disasters, which allows users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs.

[0976] Overall system configuration

[0977] This system mainly consists of the following elements:

[0978] 1. A means of collecting and storing users' daily behavior patterns

[0979] 2. Means of receiving disaster information in real time

[0980] 3. A method for identifying users in target areas based on disaster information

[0981] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[0982] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[0983] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[0984] Specific embodiments of each means will be described in detail below.

[0985] Collecting and storing users' daily behavior patterns

[0986] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[0987] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[0988] Learning daily behavior patterns

[0989] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[0990] Receiving disaster information and identifying affected areas

[0991] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[0992] Data collection in the event of a disaster

[0993] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[0994] Comparison of the latest data and calculation of risk

[0995] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[0996] Issuance of safety confirmation alerts

[0997] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, and signal reception status.

[0998] Specific examples

[0999] For example, if user A uses a smartphone on a daily basis, location information and activity data are sent from the device to the server. The server receives and stores this data, and AI learns user A's daily behavior patterns.

[1000] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User A is in that area.

[1001] User A's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, and radio wave reception status to the server. The server analyzes this data and calculates the risk level based on any differences with User A's behavioral patterns.

[1002] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User A's registered contacts (e.g., family and workplace). This is expected to enable User A's safety to be confirmed quickly and accurately, and for necessary rescue operations to be carried out promptly.

[1003] The processing flow will be explained below.

[1004] Step 1:

[1005] The device collects data on the user's daily behavior (location information, activity information, app usage, health data, etc.).

[1006] Step 2:

[1007] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[1008] Step 3:

[1009] The server receives the data sent from the terminal and stores it in a database.

[1010] Step 4:

[1011] The server inputs the stored data into an AI model that learns the user's daily behavior patterns.

[1012] Step 5:

[1013] The server generates and updates each user's behavioral pattern profile based on the learning results.

[1014] Step 6:

[1015] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from the disaster information service.

[1016] Step 7:

[1017] The server identifies the disaster-affected area based on the disaster information received, and extracts users in that area from the database.

[1018] Step 8:

[1019] The terminal receives disaster information and instantly collects current location information, last update time, information on nearby users, and radio wave reception status.

[1020] Step 9:

[1021] The latest data collected by the device is sent to the server.

[1022] Step 10:

[1023] The server compares the latest data received from the device with the daily behavior pattern profile, and the AI ​​calculates the level of risk.

[1024] Step 11:

[1025] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[1026] Step 12:

[1027] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, and radio wave reception status.

[1028] Example 1

[1029] 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."

[1030] In the event of a disaster, prompt and accurate safety confirmation is extremely important for saving many lives. However, manual safety confirmation requires time and effort, and in the event of a disaster, communication infrastructure may be destroyed or disrupted, making it difficult to respond quickly. To address these issues, the present invention aims to provide a system that uses AI to automatically confirm the safety of users and quickly send necessary notifications.

[1031] 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.

[1032] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the latest collected data with daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for transmitting the generated data to the server and analyzing it with an AI model, and means for updating the behavior pattern profile based on the daily behavior patterns. This makes it possible to quickly and accurately confirm the safety of users when a disaster occurs.

[1033] "User's daily behavioral patterns" is a general term for data related to the user's daily actions, movements, and activities.

[1034] "Disaster information" refers to real-time information about disasters such as earthquakes, weather warnings, fires, and large-scale accidents.

[1035] The "target area" refers to the area where a disaster has occurred, as identified based on disaster information.

[1036] "Latest user data" refers to data such as the user's location information at the time of the disaster, the last update time, information about nearby users, and radio wave reception conditions.

[1037] The "risk level" is a value that indicates the degree of risk to the user, calculated by comparing the latest collected data with daily behavior patterns.

[1038] A "safety confirmation alert" is a warning notification issued to a notification destination registered in advance when the user's risk level exceeds a certain threshold.

[1039] "AI model" refers to artificial intelligence algorithms and systems that analyze collected data and learn users' daily behavioral patterns.

[1040] A "behavioral pattern profile" is a data set that shows the characteristics of a user's unique daily behavior, generated based on the results of learning by an AI model.

[1041] The "Disaster Information Service" is a system that distributes real-time disaster-related information provided by organizations such as the Japan Meteorological Agency.

[1042] The present invention is an automatic safety confirmation system for use in the event of a disaster, allowing users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs. Each element of the present invention is described in detail below.

[1043] This system mainly consists of the following elements:

[1044] Collecting and storing users' daily behavior patterns

[1045] As the user uses their smartphone on a daily basis, the device collects location information, app and web usage, activity information (walking, driving, standing still), and health data (number of steps, heart rate, etc.). This data is sent to a server at regular intervals (e.g., every 15 minutes). The hardware used is assumed to be a smartphone (e.g., Android device or iPhone). The collected data is encrypted by the device using HTTPS and securely sent to the server.

[1046] Sending data to the server

[1047] The server receives the data sent from the device and stores it in a database. This allows the user's daily behavioral patterns to be accumulated and prepared for later AI analysis. The software used is a database management system (e.g., MySQL or PostgreSQL) and an AI framework for data analysis (e.g., TensorFlow or PyTorch).

[1048] Learning daily behavior patterns

[1049] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated. The Python programming language and related libraries (e.g., NumPy, Pandas) are used to train the AI ​​model.

[1050] Receiving disaster information

[1051] The server works in conjunction with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, and large-scale accidents. For example, disaster information is sent to the server via an API, and the data is analyzed immediately. A real-time data processing framework (e.g., Apache Kafka) is used for disaster information management.

[1052] Identifying target areas and collecting data

[1053] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database.The device then receives the disaster information and immediately collects the current location information, the last update time, information about nearby users (via short-range communication), and the radio wave reception status, and promptly transmits this data to the server.

[1054] Comparison of the latest data and calculation of risk

[1055] The server compares the latest data sent from the device with the daily behavioral pattern profile stored in the database. The AI ​​evaluates the differences between these data and calculates the level of risk. If the level of risk exceeds a certain threshold, it detects it as an abnormal pattern.

[1056] Issuance of safety confirmation alerts

[1057] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status. Notifications are sent via email, SMS, push notifications, etc.

[1058] Specific examples

[1059] For example, user A uses a smartphone on a daily basis, and location and activity data is sent from the device to a server. The server stores the received data, and AI learns user A's daily behavioral patterns based on data such as commuting route, activity times, and places of stay.

[1060] One day, a magnitude 5.0 earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and checks the database to see if User A is in that area. User A's device receives the earthquake information and immediately collects and sends this information to the server, along with the user's current location, last update time, information about nearby users, and signal reception status. The server uses this data to compare it with normal activity patterns and calculates User A's risk level. If the risk level exceeds a threshold, the server automatically sends a safety alert and detailed information to User A's registered contacts (e.g., family and workplace).

[1061] Example prompts for generative AI models

[1062] Generate code to program an automatic disaster safety confirmation system described in natural language. The system collects users' daily behavior from their smartphones and sends the data to a server. The server uses AI to learn daily behavior patterns and receives and analyzes disaster information in real time. When a disaster occurs, it collects the latest user data, calculates the risk level, and automatically sends safety confirmation alerts as needed.

[1063] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1064] Step 1:

[1065] Collecting users' daily behavior patterns

[1066] The device collects data on the user's daily activities. Specifically, it uses the smartphone's GPS function, app usage, activity sensors (accelerometer and gyroscope), and health data (pedometer and heart rate monitor) to collect location information, app usage history, movement status (walking, driving, stationary, etc.), and vital signs at regular intervals (e.g., every 15 minutes).

[1067] Input: Data from various sensors and apps on the user's smartphone

[1068] Output: A set of user's daily behavior data (location information, activity information, health data, etc.)

[1069] Step 2:

[1070] Sending data to the server

[1071] The device sends the collected daily behavior data to a server at regular intervals. The data is encrypted using the HTTPS protocol and sent securely to the server.

[1072] Input: Collected daily behavior data

[1073] Output: Encrypted daily activity data packet sent to the server

[1074] Step 3:

[1075] Saving daily behavior patterns

[1076] The server receives the daily behavior data sent from the device and stores it in a database, which is later used for AI analysis.

[1077] Input: Encrypted daily activity data sent from the device

[1078] Output: Daily behavior data stored in a database

[1079] Step 4:

[1080] Learning daily behavior patterns

[1081] The server periodically inputs the daily behavior data stored in the database into the AI ​​model for analysis. The AI ​​model analyzes this data and learns the user's usual behavior patterns. Based on the learning results, a behavioral pattern profile for each user is generated and regularly updated.

[1082] Input: Daily behavior data stored in a database

[1083] Output: Generated and updated behavioral pattern profile

[1084] Step 5:

[1085] Receiving disaster information

[1086] The server works with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. For example, it receives earthquake occurrence information from disaster information services via API.

[1087] Input: Disaster information from various disaster information services

[1088] Output: Disaster information data immediately stored on the server

[1089] Step 6:

[1090] Identifying the target area

[1091] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database. For example, it analyzes the latitude and longitude information of the earthquake area and identifies users within that range.

[1092] Input: Received disaster information, user location information stored in the database

[1093] Output: List of users in the identified disaster area

[1094] Step 7:

[1095] Collection of the latest data

[1096] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information about nearby users (via short-range communication such as Bluetooth), and radio wave reception status, and quickly transmits this data to the server.

[1097] Input: Disaster information, data from various sensors and communication devices on the user's smartphone

[1098] Output: Latest location information, last update time, information on nearby users, data packets of signal reception status

[1099] Step 8:

[1100] Calculating the risk

[1101] The server compares the latest data sent from the device with the stored daily behavioral pattern profile, and the AI ​​calculates the risk level. For example, if the user's location does not change for a long period of time even though it is during a normal activity time, it will determine this as an abnormality.

[1102] Input: Latest data sent from the device, saved behavioral pattern profile

[1103] Output: Calculated risk

[1104] Step 9:

[1105] Issuance of safety confirmation alerts

[1106] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status.

[1107] Input: Calculated risk level, user's registered contact information

[1108] Output: Safety confirmation alert sent and detailed information

[1109] (Application example 1)

[1110] 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."

[1111] In modern society, the occurrence of natural disasters and large-scale accidents is an unavoidable problem. Quickly and accurately confirming the safety of people when a disaster occurs is an important issue that directly relates to ensuring the safety of victims and streamlining rescue efforts. However, current safety confirmation methods involve a lot of manual confirmation, making it difficult to grasp safety in a timely manner. Furthermore, they are insufficient in proposing appropriate evacuation routes in the event of a disaster. To solve these issues, a system is needed that can more efficiently and automatically confirm safety and also propose appropriate evacuation routes.

[1112] 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.

[1113] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the collected latest data with the daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for automatically suggesting evacuation routes for high-risk users, and means for calculating the level of risk and confirming the safety of users based on prompt text from the generation AI model. This makes it possible to quickly and automatically confirm the safety of users and suggest appropriate evacuation routes when a disaster occurs.

[1114] "User's daily behavioral patterns" refer to the user's daily patterns of movement and activities, and include location information, app and web usage, activity information, health care data, etc.

[1115] "Means for receiving disaster information in real time" refers to a system that works in conjunction with disaster information services provided by the Japan Meteorological Agency and other organizations, and has the function of receiving disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. in real time.

[1116] The "means for identifying users in the target area" has a function of extracting users who are in the disaster-affected area from a database based on the received disaster information.

[1117] The "means for collecting the latest data" has the function of collecting information such as the user's location, the last update time, information about nearby users, and radio wave reception conditions when a disaster occurs.

[1118] The "means for calculating the risk level" has a function of comparing the latest collected data with the saved daily behavior patterns, evaluating the difference, and calculating the risk level of the user.

[1119] The "means for issuing safety confirmation alerts to notification recipients" has the function of automatically sending safety confirmation alerts to notification recipients registered in advance by the user when the risk level exceeds a certain threshold.

[1120] The "means for automatically suggesting evacuation routes" is a function that automatically suggests appropriate evacuation routes based on AI analysis to users at high risk.

[1121] The means for checking safety based on the "prompt sentence of the generative AI model" has the function of inputting the prompt sentence into the AI ​​model, calculating the user's level of risk, and checking the user's safety.

[1122] This invention is a system that automatically checks the user's safety during a disaster and notifies emergency contacts. The system uses AI to learn the user's daily behavior patterns, and when a disaster occurs, it compares them with the user's latest data to calculate the level of risk. In addition, if the risk is high, the system has the function of suggesting an appropriate evacuation route.

[1123] Hardware and software configuration

[1124] The hardware and software used to realize the system of the present invention are as follows.

[1125] Hardware: Smartphones, servers

[1126] Software: Python, requests (library for calling APIs), geopy (geographic calculation), smtplib (for sending emails)

[1127] External API: Real-time disaster information API for disaster information provision services

[1128] Overall system configuration

[1129] Safety confirmation system configuration

[1130] 1. Data Collection:

[1131] The server collects the user's daily behavioral patterns. The smartphone periodically collects data such as the user's location, activity, and signal reception status, and sends it to the server. The server stores this data and later analyzes it using AI.

[1132] 2. Receiving disaster information:

[1133] The server works in conjunction with disaster information services to receive disaster information in real time. Information on earthquakes, weather warnings, fires, large-scale accidents, etc. is sent to the server. Based on this information, the area where the disaster occurred is identified.

[1134] 3. Identifying users in the target area:

[1135] The server extracts users who are in the disaster area from the database, thereby identifying users who may be affected by the disaster.

[1136] 4. Collecting the latest data:

[1137] When a disaster occurs, the smartphone collects the user's latest location information, activity information, and signal reception status, and immediately sends it to a server, making it possible to check the safety of users in real time.

[1138] 5. Data comparison and risk calculation:

[1139] The server uses AI to compare the latest data with daily behavior patterns. AI analysis evaluates the differences between the collected data and normal behavior patterns and calculates the user's risk level.

[1140] 6. Issuance of Safety Confirmation Alerts:

[1141] If the specified risk threshold is exceeded, a safety confirmation alert is automatically sent to the user's emergency contacts, allowing the necessary response to be taken promptly.

[1142] 7. Suggested evacuation routes:

[1143] For users at high risk, evacuation routes are automatically suggested based on AI analysis. By providing appropriate evacuation routes, users' safety is supported.

[1144] Specific example explanation

[1145] Assume that user A travels daily in central Tokyo and its suburbs. The invention collects user A's location information and activity data, and the server learns his daily behavior patterns based on this data. One day, an earthquake occurs, and data from a disaster information service is sent to the server. If it is confirmed that user A is included in the area affected by the earthquake, the server collects user A's latest location information and activity data at that time and compares it with his daily behavior patterns. If the analysis determines that there is a high risk, the server issues a safety confirmation alert to user A's family and workplace. Furthermore, evacuation routes are automatically suggested to user A's smartphone.

[1146] Prompt Sentence Examples

[1147] Use the following prompt for the generative AI model:

[1148] Please calculate the risk level based on the daily behavior pattern data of user ID: user123, comparing it with the latest information (location information, activity data, etc.) in the event of a disaster. Provide the following data:

[1149] Daily behavior pattern data: [Detailed data]

[1150] Latest data: [Detailed data]

[1151] In this way, the present invention automatically checks the user's safety during a disaster, quickly notifies emergency contacts, and supports necessary responses, thereby ensuring the safety of victims and improving the efficiency of rescue operations.

[1152] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1153] Step 1:

[1154] Collect user location and activity data.

[1155] Input: Sensor data such as location information, activity data, and signal reception status from the user's smartphone.

[1156] Processing: The smartphone collects this data at regular intervals, converts it into a data format, and sends it to the server.

[1157] Output: A set of user location and activity data sent to the server.

[1158] Step 2:

[1159] The server stores the user's daily behavior patterns.

[1160] Input: Location, activity data, and other sensor data sent from your smartphone.

[1161] Processing: The server stores the received data in a database and integrates it with past data, thereby accumulating daily behavior patterns.

[1162] Output: User's daily behavior pattern data stored in a database.

[1163] Step 3:

[1164] Receive disaster information in real time.

[1165] Input: Real-time data from disaster information services (earthquake occurrence information, weather warnings, fire information, large-scale accident information, etc.).

[1166] Processing: The server calls the API and obtains disaster information.

[1167] Output: Disaster information data stored on the server.

[1168] Step 4:

[1169] Identify users in your target area.

[1170] Input: Disaster information (data on the location and scale of the disaster), user's daily behavior pattern data (location information).

[1171] Processing: The server analyzes the disaster area and extracts users who are in that area from the database.

[1172] Output: A list of users in the disaster area.

[1173] Step 5:

[1174] Collect the latest data when a disaster occurs.

[1175] Input: Disaster notification, latest location information of the user's smartphone, activity data, and signal reception status.

[1176] Processing: Upon receiving disaster information, the smartphone immediately collects the latest sensor data and sends it to the server.

[1177] Output: The latest sensor data sent to the server.

[1178] Step 6:

[1179] The latest data is compared with daily behavior patterns to calculate the level of risk.

[1180] Input: Latest sensor data, daily behavior pattern data.

[1181] Processing: The server uses the generated AI model to compare the latest data with daily behavior patterns and calculate the user's risk level. The prompt sentence is input into the model and the analysis results are obtained.

[1182] Output: Calculated user risk level (levels such as normal, caution, danger, etc.).

[1183] Step 7:

[1184] If the risk level exceeds a certain threshold, an alert will be issued to check the safety of the person.

[1185] Input: Risk data, user's emergency contact information.

[1186] Processing: If the server detects a risk level that exceeds the threshold, it will automatically send an email or SMS to the user's emergency contacts to confirm their safety.

[1187] Output: Safety alert sent to emergency contacts.

[1188] Step 8:

[1189] Automatically suggests evacuation routes to users at high risk.

[1190] Input: User's current location, risk level data, and surrounding environment information.

[1191] Processing: The server uses AI to analyze surrounding geographical information and disaster information to generate an appropriate evacuation route, which is then sent to the user's smartphone.

[1192] Output: Evacuation route displayed on the user's smartphone.

[1193] 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.

[1194] This invention is an automatic safety confirmation system for use in disaster situations, and by combining it with an emotion engine that recognizes the user's emotions, it is possible to perform more accurate risk assessments. This system uses AI to analyze smartphone usage data, and aims to automatically confirm and notify the user's safety when a disaster occurs.

[1195] Overall system configuration

[1196] The system consists of the following main elements:

[1197] 1. A means of collecting and storing users' daily behavior patterns

[1198] 2. Means of receiving disaster information in real time

[1199] 3. A method for identifying users in target areas based on disaster information

[1200] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[1201] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[1202] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[1203] 7. Emotion engine that recognizes user emotions

[1204] Specific embodiments of each means will be described in detail below.

[1205] Collecting and storing users' daily behavior patterns

[1206] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[1207] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[1208] Learning daily behavior patterns

[1209] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[1210] Receiving disaster information and identifying affected areas

[1211] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[1212] Data collection in the event of a disaster

[1213] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[1214] Comparison of the latest data and calculation of risk

[1215] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[1216] Use of emotion engine

[1217] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions, voice, and actions, and sends it to the server.

[1218] The server uses the emotion data sent by the user to assess the risk level. The emotion data indicates the user's stress level and tension, and the accuracy of the risk level is improved based on this. For example, if a user is in an abnormally stressed state when a disaster occurs, this information is reflected in the risk level calculation.

[1219] Issuance of safety confirmation alerts

[1220] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[1221] Specific examples

[1222] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives and stores this data, and the AI ​​learns user B's daily behavioral patterns.

[1223] One day, an earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area, and extracts from the database whether User B is in that area.

[1224] User B's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, signal reception status, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[1225] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family or workplace). This is expected to enable User B's safety to be confirmed quickly and accurately, and to enable necessary rescue operations to be carried out promptly.

[1226] The processing flow will be explained below.

[1227] Step 1:

[1228] The device collects data on the user's daily activities (location information, activity information, app usage, health data, and emotional data). Emotional data is collected by an emotion engine that recognizes the user's facial expressions, voice, and behavior.

[1229] Step 2:

[1230] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[1231] Step 3:

[1232] The server receives the data sent from the terminal and stores it in a database.

[1233] Step 4:

[1234] The server inputs the stored data into the AI ​​model, which learns the user's daily behavioral patterns and emotional data. Based on the learning results, a behavioral pattern profile for each user is generated and updated.

[1235] Step 5:

[1236] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from various disaster information services.

[1237] Step 6:

[1238] The server identifies the disaster-affected area based on the disaster information received, and extracts users who are in that area from the database.

[1239] Step 7:

[1240] When the device receives disaster information, it immediately collects the current location information, the last update time, information about nearby users, radio wave reception conditions, and emotional data.

[1241] Step 8:

[1242] The latest data collected by the device is sent to the server.

[1243] Step 9:

[1244] The server compares the latest data received from the device with the daily behavioral pattern profile and emotional data, and the AI ​​calculates the risk level, taking into account the user's emotional data (e.g., stress level and tension).

[1245] Step 10:

[1246] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[1247] Step 11:

[1248] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, radio wave reception status, and emotional data.

[1249] Example 2

[1250] 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."

[1251] In the event of a disaster, it is extremely important to quickly and accurately confirm the safety of users and take appropriate action. However, conventional systems do not take into account the user's daily behavioral patterns or emotional state in risk assessment, resulting in low accuracy in safety confirmation. The present invention aims to solve these problems.

[1252] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest data of users (location information, last update time, information on surrounding users, and communication environment) when a disaster occurs, means for comparing the collected latest data with daily behavior patterns to calculate a risk level, means for recognizing user emotion data and incorporating it into risk assessment, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables highly accurate risk assessment that takes into account the user's behavior patterns and emotional state.

[1253] "Means for collecting and storing users' daily behavioral patterns" refers to means for collecting users' daily activities, location information, app usage, and activity information, and storing this in a database.

[1254] "Means for receiving disaster information in real time" refers to means for obtaining disaster information such as earthquakes, weather, fires, and large-scale accidents in real time and transmitting it to a server.

[1255] The "means for identifying users in a target area based on received disaster information" is a means for identifying a disaster-affected area from received disaster information and extracting users who are in that area from a database.

[1256] "Means for collecting the latest user data (location information, last update time, surrounding user information, communication environment) when a disaster occurs" refers to means for collecting the user's current location information, last update time, user information via surrounding communications, and the status of radio waves and the communication environment when a disaster occurs.

[1257] The "means for calculating the risk by comparing the latest collected data with daily behavior patterns" is a means for comparing the latest collected data with the user's saved daily behavior pattern profile, evaluating the difference, and calculating the risk.

[1258] "Means for recognizing the user's emotional data and incorporating it into risk assessment" refers to means for collecting and analyzing data on the user's emotional state (e.g., facial expressions and voice) and reflecting it in risk assessment.

[1259] "Means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold" refers to a means for automatically sending a safety confirmation alert to the user's registered emergency contacts (family, workplace, etc.) when the calculated risk level exceeds a pre-set threshold.

[1260] This invention is an automatic safety confirmation system for use in the event of a disaster, which performs more accurate risk assessment based on the user's daily behavioral patterns and emotional data. The system uses AI to analyze smartphone usage data and automatically confirms and notifies the user's safety in the event of a disaster.

[1261] System Configuration

[1262] The system consists of the following main elements:

[1263] 1. A means of collecting and storing users' daily behavior patterns

[1264] The device collects data on the user's daily behavior using the smartphone's GPS, health sensors, app usage logs, camera, microphone, etc. This data includes location information, activity data (walking, driving, and standing still), heart rate, and app usage.

[1265] The collected data is sent to the server at regular intervals (for example, every 15 minutes) and stored on the server.

[1266] 2. A means of learning daily behavior patterns

[1267] The server periodically retrieves daily behavior data stored in the database and analyzes it using a generative AI model, primarily using deep learning frameworks such as TensorFlow and PyTorch.

[1268] The AI ​​model learns the user's usual behavioral patterns (for example, daily commuting route and activity times) and generates and updates a behavioral pattern profile based on this.

[1269] 3. Means of receiving disaster information and identifying affected areas

[1270] The server works in conjunction with various disaster information services (e.g., the Japan Meteorological Agency's API) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents.

[1271] When disaster information is received, the area where the disaster occurred is identified based on that information, and users who are in that area are extracted from the database.

[1272] 4. How to collect the latest user data when a disaster occurs

[1273] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information on nearby users (via short-range communication), and the communication environment.

[1274] This data is quickly transmitted to the server.

[1275] 5. Comparing the latest data with daily behavior patterns and calculating the risk level

[1276] The server compares the latest data sent from the device with the stored daily behavior pattern profile.

[1277] The generative AI evaluates the differences in this data and calculates the risk level, detecting anomalies based on deviations from normal behavioral patterns.

[1278] 6. Methods for Recognizing User Emotional Data and Incorporating It into Risk Assessment

[1279] The device collects emotional data from the user's facial expressions, voice, and actions, and sends it to a server using a camera and microphone.

[1280] The server performs a risk assessment based on the emotional data analyzed by the emotion engine, taking into account the user's stress level and tension.

[1281] 7. How to issue safety confirmation alerts

[1282] If the calculated risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace).

[1283] The alert includes the user's last location, last update time, information about nearby users, communication environment, emotional data, etc.

[1284] Specific examples

[1285] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives this data and stores it in a database, and the generative AI model learns user B's daily behavioral patterns.

[1286] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User B is in that area.

[1287] User B's device receives the earthquake information and immediately sends its current location, last update time, information about nearby users, communication environment, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[1288] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family, workplace). This process is expected to quickly and accurately confirm User B's safety, and enable necessary rescue operations to be carried out promptly.

[1289] Prompt Sentence Examples

[1290] "Please explain in detail your automated safety confirmation system in the event of a disaster. Please tell us the specific processing steps, from data collection to risk assessment and safety confirmation alerts. Please also include examples of application in real-life scenarios."

[1291] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1292] Processing Steps

[1293] Step 1: Data collection

[1294] The device collects information about the user's daily activities, including location information using GPS, activity information obtained from smartphone sensors (e.g., walking, driving, standing still), app usage, and heart rate.

[1295] Input: User's smartphone usage, sensor information

[1296] Output: Collected daily behavior data (e.g., location information, heart rate)

[1297] Specific operation: The device activates the sensor every 15 minutes and packages the acquired data in JSON format.

[1298] Step 2: Send data

[1299] The device sends the collected daily behavior data to the server at regular intervals (e.g., every 15 minutes).

[1300] Input: Collected daily behavior data

[1301] Output: Data sent to the server

[1302] Specific operation: The device sends JSON formatted data to the server's API endpoint using HTTPS.

[1303] Step 3: Save Data

[1304] The server receives the daily activity data sent from the terminal and stores it in a database.

[1305] Input: Daily activity data sent from the device

[1306] Output: Behavioral data stored in a database

[1307] Specific behavior: The server parses the received JSON data and stores it in the database in the appropriate format.

[1308] Step 4: Learning behavioral patterns

[1309] The server periodically retrieves the behavioral data stored in the database, inputs it into the generative AI model, and analyzes it.

[1310] Input: Saved daily activity data

[1311] Output: Learned behavioral pattern profile

[1312] Specific operation: The server uses TensorFlow and PyTorch to run a behavior prediction model and learn user behavior patterns.

[1313] Step 5: Receiving disaster information

[1314] The server receives information in real time from the disaster information service.

[1315] Input: Data from disaster information service

[1316] Output: Disaster information received by the server

[1317] Specific operation: The server periodically checks the API endpoint to obtain the latest disaster information.

[1318] Step 6: Identify users in your target area

[1319] The server identifies the disaster-affected area from the received disaster information and extracts users within that area.

[1320] Input: Disaster information, user location information in the database

[1321] Output: A list of users within the identified target area

[1322] Specific operation: The server uses GIS to map the disaster area and extracts users in that area from the database.

[1323] Step 7: Collect the latest data

[1324] When the terminal receives disaster information, it collects the current location information, the last update time, information on nearby users, and the communication environment.

[1325] Input: Disaster information, device sensor information

[1326] Output: Latest collected data (location information, communication environment, etc.)

[1327] What it does: The device performs GPS and Bluetooth scans to collect and send the latest data.

[1328] Step 8: Data comparison and risk calculation

[1329] The server compares the latest data sent from the device with the saved daily behavior patterns and calculates the level of risk.

[1330] Input: Latest data, daily behavior pattern profile

[1331] Output: Calculated risk

[1332] Specific operation: The server uses a generative AI model to detect abnormalities in behavioral patterns and evaluates the level of risk based on the results.

[1333] Step 9: Collect and evaluate emotion data

[1334] The terminal collects the user's emotional data (facial expressions, voice, etc.) and sends it to the server.

[1335] Input: User's facial expression and voice data

[1336] Output: Parsed emotion data

[1337] Specific operation: The device uses the camera and microphone to capture emotion data and sends it to the server.

[1338] Step 10: Finalize your risk assessment

[1339] The server recalculates the risk assessment based on the emotion data.

[1340] Input: Emotion data, initial risk assessment

[1341] Output: Adjusted risk assessment

[1342] Specific operation: The server takes in the emotional data, reassess the danger level, and generates the final result.

[1343] Step 11: Issue an alert

[1344] If the calculated risk level exceeds a threshold, the server sends a safety confirmation alert to the registered destination of the target user.

[1345] Input: Adjusted risk assessment, user registration information

[1346] Output: Safety confirmation alert issued

[1347] Specific operation: The server generates an alert message containing the necessary information (e.g., last location, nearby user information, etc.) and sends it to the registered recipient via email or SMS.

[1348] In this way, the system takes into account the user's daily behavior and emotional state, enabling quick and highly accurate confirmation of safety during a disaster.

[1349] (Application example 2)

[1350] 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."

[1351] Although systems for confirming users' safety during disasters already exist, they typically rely on users to provide information voluntarily, which often results in inaccurate information being unavailable. Furthermore, these systems assess risk based solely on location and activity data, without taking into account the user's emotional state, limiting the accuracy of risk assessments. Therefore, there is a need for a system that can more quickly and accurately confirm users' safety and provide appropriate notifications.

[1352] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in the target area based on the received disaster information, means for collecting user emotion data and using it for risk assessment, means for calculating the risk level by comparing the latest collected data with daily behavior patterns, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables more accurate risk assessment and faster safety confirmation by analyzing the user emotion data and reflecting the stress level and tension level in the risk assessment.

[1353] "User's daily behavior patterns" refers to the totality of data such as the user's daily actions, movements, and app usage.

[1354] "Disaster information" refers to information about emergencies such as earthquakes, fires, weather warnings, and large-scale accidents.

[1355] "Target area" refers to the geographical area identified as being affected by a disaster based on disaster information.

[1356] The "latest data" is real-time data including the user's location information, the last update time, information about nearby users, radio wave reception conditions, and emotion data at the time of the disaster.

[1357] The "risk level" is a numerical value or classification of the degree of risk in the situation in which the user is placed.

[1358] "Emotion data" is data that indicates the emotional state of the user as recognized from facial expressions, tone of voice, and the like.

[1359] "Risk assessment" is the process of calculating a user's risk level based on collected data.

[1360] A "safety confirmation alert" is a message or notification sent to a contact to confirm whether the user is safe.

[1361] The present invention is an automatic safety confirmation system for use in the event of a disaster, which collects data on users' daily behavior patterns and emotions and performs highly accurate risk assessments in the event of a disaster. Hereinafter, embodiments of the present invention will be described in detail.

[1362] Overall system configuration

[1363] This system is mainly composed of three elements: a server, a terminal, and a user.

[1364] Collecting and storing daily behavior patterns

[1365] As users use their smartphones on a daily basis, the device collects location information, activity data (e.g., walking, driving, and standing still), app usage, and emotional data (e.g., facial expressions and tone of voice). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[1366] Learning daily behavior patterns

[1367] The server periodically retrieves daily behavior data stored in the database and inputs it into an AI model (e.g., TensorFlow, PyTorch). The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route, activity times). Based on the learning results, a behavioral pattern profile is generated and continuously updated.

[1368] Receiving disaster information and identifying affected areas

[1369] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the server identifies the area where the disaster occurred and extracts users who are in that area from the database.

[1370] Data collection in the event of a disaster

[1371] When a device receives disaster information, it immediately collects its current location, the last update time, information about nearby users (e.g., device detection via short-range communication), and the status of radio wave reception. This data is then promptly sent to the server. The device also collects emotion data and sends it to the server.

[1372] Comparison of the latest data and calculation of risk

[1373] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change even during normal activity times, it will be detected as an abnormal pattern.

[1374] Use of emotion engine

[1375] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions and tone of voice and sends it to a server. The server analyzes the emotion data and reflects stress levels and tension in the risk assessment, improving the accuracy of the risk assessment.

[1376] Issuance of safety confirmation alerts

[1377] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[1378] Specific examples

[1379] For example, when a user uses a smartphone on a daily basis, location information, app usage status, and emotional data are sent from the device to a server. The server receives and stores this information, and AI learns the user's daily behavioral patterns. One day, an earthquake occurs. The server receives disaster information, confirms that the disaster occurred in a specific area, and extracts from the database whether the user is in that area. The device receives the earthquake information and immediately collects current location information, last update time, information about surrounding users, signal reception status, and emotional data, and sends them to the server. The server analyzes this data and calculates the level of risk based on the difference between the behavioral pattern and the emotional data. If the level of risk exceeds a threshold, the server automatically sends a safety confirmation alert and detailed information to the user's registered contacts.

[1380] Example prompts for generative AI models

[1381] Enter the user's face photo and voice data:

[1382] Facial photo data:[path_to_image]\

[1383] Audio data:[path_to_audio]\

[1384] \

[1385] Parse the following data attributes:

[1386] 1. Emotion recognition (e.g., joy, anxiety, anger, sadness)

[1387] 2. Tone of voice (e.g., nervous, stressed, calm)

[1388] \

[1389] As output, return the user's emotional state and its associated numerical data (e.g., stress level 0-100).

[1390] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1391] Step 1:

[1392] Data collection

[1393] The device collects the user's daily behavioral patterns. Specifically, it obtains location information, activity data, app usage, and emotional data (facial expressions, tone of voice) using sensors, microphones, and cameras. For example, it uses APIs such as LocationServices.requestLocationUpdates(), ActivityRecognitionClient.requestActivityUpdates(), Camera.captureImage(), and Microphone.recordAudio(). The input data is various types of user behavioral information, and the output data is the collected behavioral data.

[1394] Step 2:

[1395] Data transmission

[1396] The collected data is sent to the server at regular intervals (for example, every 15 minutes). This involves sending a POST request using an HTTP client. For example, HttpClient.sendPostRequest(data). The input data is the behavioral data collected on the device, and the output data is the behavioral data sent to the server.

[1397] Step 3:

[1398] Data storage

[1399] The server stores the received data in a database. Specifically, it executes an INSERT statement in the database using SQL or similar. Example: Database.insertData(data). The input data is the behavioral data sent, and the output data is the behavioral data stored in the database.

[1400] Step 4:

[1401] Learning behavioral patterns

[1402] The server periodically retrieves the behavioral data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's daily behavioral patterns. Example: aiModel.fit(trainingData). The input data is past behavioral data, and the output data is the user's daily behavioral pattern profile.

[1403] Step 5:

[1404] Receiving disaster information

[1405] The server works with the disaster information service to receive disaster information in real time. Example: disasterAlertService.receiveAlert(). The input data is disaster information, and the output data is the received disaster information.

[1406] Step 6:

[1407] Identifying target areas

[1408] The server identifies the disaster-affected area based on the received disaster information and extracts users who are in that area from the database. The input data is the disaster information and the user's location information, and the output data is a list of users in the identified target area.

[1409] Step 7:

[1410] Collection of the latest data

[1411] When disaster information is received, the device immediately collects the current location information, the last update time, information about nearby users, signal reception status, and emotion data. (Example: LocationServices.requestLocationUpdate(), HeartbeatSensor.getLatestReading()) The input data is the disaster information, and the output data is the latest collected user data.

[1412] Step 8:

[1413] Sending data

[1414] Quickly send the latest data from the device to the server. Example: HttpClient.sendPostRequest(latestData). The input data is the latest user data collected, and the output data is the latest data sent to the server.

[1415] Step 9:

[1416] Calculating the risk

[1417] The server compares the latest data sent from the device with the stored daily behavior pattern profile. The AI ​​evaluates the differences between these data and calculates the risk level. Example: riskLevel = aiModel.predict(latestData). The input data is the latest user data and daily behavior pattern profile, and the output data is the risk assessment result.

[1418] Step 10:

[1419] Emotional Data Analysis

[1420] The server analyzes the emotion data and reflects the stress level and tension in the risk assessment. Example: emotionAnalysisEngine.analyze(emotionData). The input data is emotion data, and the output data is the emotion data analysis result.

[1421] Step 11:

[1422] Issuance of safety confirmation alerts

[1423] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts. Example: NotificationService.sendAlert(contactInfo, alertDetails). The input data is the user's risk assessment result and contact information, and the output data is the sent alert.

[1424] At each step, the process flow is clear, detailing how the input data is processed or calculated and what the output is, making it easy to understand how the entire system works.

[1425] 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.

[1426] 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.

[1427] 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.

[1428] [Fourth embodiment]

[1429] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1430] 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.

[1431] 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).

[1432] 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.

[1433] 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.

[1434] 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).

[1435] 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.

[1436] 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.

[1437] 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.

[1438] 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.

[1439] 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.

[1440] 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.

[1441] 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."

[1442] This invention is an automatic safety confirmation system for use in disasters, which allows users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs.

[1443] Overall system configuration

[1444] This system mainly consists of the following elements:

[1445] 1. A means of collecting and storing users' daily behavior patterns

[1446] 2. Means of receiving disaster information in real time

[1447] 3. A method for identifying users in target areas based on disaster information

[1448] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[1449] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[1450] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[1451] Specific embodiments of each means will be described in detail below.

[1452] Collecting and storing users' daily behavior patterns

[1453] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[1454] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[1455] Learning daily behavior patterns

[1456] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[1457] Receiving disaster information and identifying affected areas

[1458] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[1459] Data collection in the event of a disaster

[1460] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[1461] Comparison of the latest data and calculation of risk

[1462] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[1463] Issuance of safety confirmation alerts

[1464] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, and signal reception status.

[1465] Specific examples

[1466] For example, if user A uses a smartphone on a daily basis, location information and activity data are sent from the device to the server. The server receives and stores this data, and AI learns user A's daily behavior patterns.

[1467] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User A is in that area.

[1468] User A's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, and radio wave reception status to the server. The server analyzes this data and calculates the risk level based on any differences with User A's behavioral patterns.

[1469] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User A's registered contacts (e.g., family and workplace). This is expected to enable User A's safety to be confirmed quickly and accurately, and for necessary rescue operations to be carried out promptly.

[1470] The processing flow will be explained below.

[1471] Step 1:

[1472] The device collects data on the user's daily behavior (location information, activity information, app usage, health data, etc.).

[1473] Step 2:

[1474] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[1475] Step 3:

[1476] The server receives the data sent from the terminal and stores it in a database.

[1477] Step 4:

[1478] The server inputs the stored data into an AI model that learns the user's daily behavior patterns.

[1479] Step 5:

[1480] The server generates and updates each user's behavioral pattern profile based on the learning results.

[1481] Step 6:

[1482] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from the disaster information service.

[1483] Step 7:

[1484] The server identifies the disaster-affected area based on the disaster information received, and extracts users in that area from the database.

[1485] Step 8:

[1486] The terminal receives disaster information and instantly collects current location information, last update time, information on nearby users, and radio wave reception status.

[1487] Step 9:

[1488] The latest data collected by the device is sent to the server.

[1489] Step 10:

[1490] The server compares the latest data received from the device with the daily behavior pattern profile, and the AI ​​calculates the level of risk.

[1491] Step 11:

[1492] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[1493] Step 12:

[1494] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, and radio wave reception status.

[1495] Example 1

[1496] 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."

[1497] In the event of a disaster, prompt and accurate safety confirmation is extremely important for saving many lives. However, manual safety confirmation requires time and effort, and in the event of a disaster, communication infrastructure may be destroyed or disrupted, making it difficult to respond quickly. To address these issues, the present invention aims to provide a system that uses AI to automatically confirm the safety of users and quickly send necessary notifications.

[1498] 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.

[1499] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the latest collected data with daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for transmitting the generated data to the server and analyzing it with an AI model, and means for updating the behavior pattern profile based on the daily behavior patterns. This makes it possible to quickly and accurately confirm the safety of users when a disaster occurs.

[1500] "User's daily behavioral patterns" is a general term for data related to the user's daily actions, movements, and activities.

[1501] "Disaster information" refers to real-time information about disasters such as earthquakes, weather warnings, fires, and large-scale accidents.

[1502] The "target area" refers to the area where a disaster has occurred, as identified based on disaster information.

[1503] "Latest user data" refers to data such as the user's location information at the time of the disaster, the last update time, information about nearby users, and radio wave reception conditions.

[1504] The "risk level" is a value that indicates the degree of risk to the user, calculated by comparing the latest collected data with daily behavior patterns.

[1505] A "safety confirmation alert" is a warning notification issued to a notification destination registered in advance when the user's risk level exceeds a certain threshold.

[1506] "AI model" refers to artificial intelligence algorithms and systems that analyze collected data and learn users' daily behavioral patterns.

[1507] A "behavioral pattern profile" is a data set that shows the characteristics of a user's unique daily behavior, generated based on the results of learning by an AI model.

[1508] The "Disaster Information Service" is a system that distributes real-time disaster-related information provided by organizations such as the Japan Meteorological Agency.

[1509] The present invention is an automatic safety confirmation system for use in the event of a disaster, allowing users to automatically notify others of their safety status. The system uses AI to analyze smartphone usage data and automatically confirms and notifies users of their safety status when a disaster occurs. Each element of the present invention is described in detail below.

[1510] This system mainly consists of the following elements:

[1511] Collecting and storing users' daily behavior patterns

[1512] As the user uses their smartphone on a daily basis, the device collects location information, app and web usage, activity information (walking, driving, standing still), and health data (number of steps, heart rate, etc.). This data is sent to a server at regular intervals (e.g., every 15 minutes). The hardware used is assumed to be a smartphone (e.g., Android device or iPhone). The collected data is encrypted by the device using HTTPS and securely sent to the server.

[1513] Sending data to the server

[1514] The server receives the data sent from the device and stores it in a database. This allows the user's daily behavioral patterns to be accumulated and prepared for later AI analysis. The software used is a database management system (e.g., MySQL or PostgreSQL) and an AI framework for data analysis (e.g., TensorFlow or PyTorch).

[1515] Learning daily behavior patterns

[1516] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated. The Python programming language and related libraries (e.g., NumPy, Pandas) are used to train the AI ​​model.

[1517] Receiving disaster information

[1518] The server works in conjunction with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, and large-scale accidents. For example, disaster information is sent to the server via an API, and the data is analyzed immediately. A real-time data processing framework (e.g., Apache Kafka) is used for disaster information management.

[1519] Identifying target areas and collecting data

[1520] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database.The device then receives the disaster information and immediately collects the current location information, the last update time, information about nearby users (via short-range communication), and the radio wave reception status, and promptly transmits this data to the server.

[1521] Comparison of the latest data and calculation of risk

[1522] The server compares the latest data sent from the device with the daily behavioral pattern profile stored in the database. The AI ​​evaluates the differences between these data and calculates the level of risk. If the level of risk exceeds a certain threshold, it detects it as an abnormal pattern.

[1523] Issuance of safety confirmation alerts

[1524] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status. Notifications are sent via email, SMS, push notifications, etc.

[1525] Specific examples

[1526] For example, user A uses a smartphone on a daily basis, and location and activity data is sent from the device to a server. The server stores the received data, and AI learns user A's daily behavioral patterns based on data such as commuting route, activity times, and places of stay.

[1527] One day, a magnitude 5.0 earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and checks the database to see if User A is in that area. User A's device receives the earthquake information and immediately collects and sends this information to the server, along with the user's current location, last update time, information about nearby users, and signal reception status. The server uses this data to compare it with normal activity patterns and calculates User A's risk level. If the risk level exceeds a threshold, the server automatically sends a safety alert and detailed information to User A's registered contacts (e.g., family and workplace).

[1528] Example prompts for generative AI models

[1529] Generate code to program an automatic disaster safety confirmation system described in natural language. The system collects users' daily behavior from their smartphones and sends the data to a server. The server uses AI to learn daily behavior patterns and receives and analyzes disaster information in real time. When a disaster occurs, it collects the latest user data, calculates the risk level, and automatically sends safety confirmation alerts as needed.

[1530] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1531] Step 1:

[1532] Collecting users' daily behavior patterns

[1533] The device collects data on the user's daily activities. Specifically, it uses the smartphone's GPS function, app usage, activity sensors (accelerometer and gyroscope), and health data (pedometer and heart rate monitor) to collect location information, app usage history, movement status (walking, driving, stationary, etc.), and vital signs at regular intervals (e.g., every 15 minutes).

[1534] Input: Data from various sensors and apps on the user's smartphone

[1535] Output: A set of user's daily behavior data (location information, activity information, health data, etc.)

[1536] Step 2:

[1537] Sending data to the server

[1538] The device sends the collected daily behavior data to a server at regular intervals. The data is encrypted using the HTTPS protocol and sent securely to the server.

[1539] Input: Collected daily behavior data

[1540] Output: Encrypted daily activity data packet sent to the server

[1541] Step 3:

[1542] Saving daily behavior patterns

[1543] The server receives the daily behavior data sent from the device and stores it in a database, which is later used for AI analysis.

[1544] Input: Encrypted daily activity data sent from the device

[1545] Output: Daily behavior data stored in a database

[1546] Step 4:

[1547] Learning daily behavior patterns

[1548] The server periodically inputs the daily behavior data stored in the database into the AI ​​model for analysis. The AI ​​model analyzes this data and learns the user's usual behavior patterns. Based on the learning results, a behavioral pattern profile for each user is generated and regularly updated.

[1549] Input: Daily behavior data stored in a database

[1550] Output: Generated and updated behavioral pattern profile

[1551] Step 5:

[1552] Receiving disaster information

[1553] The server works with various disaster information services, such as the Japan Meteorological Agency, to receive real-time disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. For example, it receives earthquake occurrence information from disaster information services via API.

[1554] Input: Disaster information from various disaster information services

[1555] Output: Disaster information data immediately stored on the server

[1556] Step 6:

[1557] Identifying the target area

[1558] The server identifies the disaster area based on the received disaster information and extracts users in that area from the database. For example, it analyzes the latitude and longitude information of the earthquake area and identifies users within that range.

[1559] Input: Received disaster information, user location information stored in the database

[1560] Output: List of users in the identified disaster area

[1561] Step 7:

[1562] Collection of the latest data

[1563] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information about nearby users (via short-range communication such as Bluetooth), and radio wave reception status, and quickly transmits this data to the server.

[1564] Input: Disaster information, data from various sensors and communication devices on the user's smartphone

[1565] Output: Latest location information, last update time, information on nearby users, data packets of signal reception status

[1566] Step 8:

[1567] Calculating the risk

[1568] The server compares the latest data sent from the device with the stored daily behavioral pattern profile, and the AI ​​calculates the risk level. For example, if the user's location does not change for a long period of time even though it is during a normal activity time, it will determine this as an abnormality.

[1569] Input: Latest data sent from the device, saved behavioral pattern profile

[1570] Output: Calculated risk

[1571] Step 9:

[1572] Issuance of safety confirmation alerts

[1573] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace). The notification includes the user's last location, last update time, information about nearby users, and signal reception status.

[1574] Input: Calculated risk level, user's registered contact information

[1575] Output: Safety confirmation alert sent and detailed information

[1576] (Application example 1)

[1577] 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."

[1578] In modern society, the occurrence of natural disasters and large-scale accidents is an unavoidable problem. Quickly and accurately confirming the safety of people when a disaster occurs is an important issue that directly relates to ensuring the safety of victims and streamlining rescue efforts. However, current safety confirmation methods involve a lot of manual confirmation, making it difficult to grasp safety in a timely manner. Furthermore, they are insufficient in proposing appropriate evacuation routes in the event of a disaster. To solve these issues, a system is needed that can more efficiently and automatically confirm safety and also propose appropriate evacuation routes.

[1579] 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.

[1580] In this invention, the server includes means for collecting and saving users' daily behavior patterns, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs, means for calculating the level of risk by comparing the collected latest data with the daily behavior patterns, means for issuing a safety confirmation alert to pre-registered notification destinations when the level of risk exceeds a certain threshold, means for automatically suggesting evacuation routes for high-risk users, and means for calculating the level of risk and confirming the safety of users based on prompt text from the generation AI model. This makes it possible to quickly and automatically confirm the safety of users and suggest appropriate evacuation routes when a disaster occurs.

[1581] "User's daily behavioral patterns" refer to the user's daily patterns of movement and activities, and include location information, app and web usage, activity information, health care data, etc.

[1582] "Means for receiving disaster information in real time" refers to a system that works in conjunction with disaster information services provided by the Japan Meteorological Agency and other organizations, and has the function of receiving disaster information on earthquakes, weather warnings, fires, large-scale accidents, etc. in real time.

[1583] The "means for identifying users in the target area" has a function of extracting users who are in the disaster-affected area from a database based on the received disaster information.

[1584] The "means for collecting the latest data" has the function of collecting information such as the user's location, the last update time, information about nearby users, and radio wave reception conditions when a disaster occurs.

[1585] The "means for calculating the risk level" has a function of comparing the latest collected data with the saved daily behavior patterns, evaluating the difference, and calculating the risk level of the user.

[1586] The "means for issuing safety confirmation alerts to notification recipients" has the function of automatically sending safety confirmation alerts to notification recipients registered in advance by the user when the risk level exceeds a certain threshold.

[1587] The "means for automatically suggesting evacuation routes" is a function that automatically suggests appropriate evacuation routes based on AI analysis to users at high risk.

[1588] The means for checking safety based on the "prompt sentence of the generative AI model" has the function of inputting the prompt sentence into the AI ​​model, calculating the user's level of risk, and checking the user's safety.

[1589] This invention is a system that automatically checks the user's safety during a disaster and notifies emergency contacts. The system uses AI to learn the user's daily behavior patterns, and when a disaster occurs, it compares them with the user's latest data to calculate the level of risk. In addition, if the risk is high, the system has the function of suggesting an appropriate evacuation route.

[1590] Hardware and software configuration

[1591] The hardware and software used to realize the system of the present invention are as follows.

[1592] Hardware: Smartphones, servers

[1593] Software: Python, requests (library for calling APIs), geopy (geographic calculation), smtplib (for sending emails)

[1594] External API: Real-time disaster information API for disaster information provision services

[1595] Overall system configuration

[1596] Safety confirmation system configuration

[1597] 1. Data Collection:

[1598] The server collects the user's daily behavioral patterns. The smartphone periodically collects data such as the user's location, activity, and signal reception status, and sends it to the server. The server stores this data and later analyzes it using AI.

[1599] 2. Receiving disaster information:

[1600] The server works in conjunction with disaster information services to receive disaster information in real time. Information on earthquakes, weather warnings, fires, large-scale accidents, etc. is sent to the server. Based on this information, the area where the disaster occurred is identified.

[1601] 3. Identifying users in the target area:

[1602] The server extracts users who are in the disaster area from the database, thereby identifying users who may be affected by the disaster.

[1603] 4. Collecting the latest data:

[1604] When a disaster occurs, the smartphone collects the user's latest location information, activity information, and signal reception status, and immediately sends it to a server, making it possible to check the safety of users in real time.

[1605] 5. Data comparison and risk calculation:

[1606] The server uses AI to compare the latest data with daily behavior patterns. AI analysis evaluates the differences between the collected data and normal behavior patterns and calculates the user's risk level.

[1607] 6. Issuance of Safety Confirmation Alerts:

[1608] If the specified risk threshold is exceeded, a safety confirmation alert is automatically sent to the user's emergency contacts, allowing the necessary response to be taken promptly.

[1609] 7. Suggested evacuation routes:

[1610] For users at high risk, evacuation routes are automatically suggested based on AI analysis. By providing appropriate evacuation routes, users' safety is supported.

[1611] Specific example explanation

[1612] Assume that user A travels daily in central Tokyo and its suburbs. The invention collects user A's location information and activity data, and the server learns his daily behavior patterns based on this data. One day, an earthquake occurs, and data from a disaster information service is sent to the server. If it is confirmed that user A is included in the area affected by the earthquake, the server collects user A's latest location information and activity data at that time and compares it with his daily behavior patterns. If the analysis determines that there is a high risk, the server issues a safety confirmation alert to user A's family and workplace. Furthermore, evacuation routes are automatically suggested to user A's smartphone.

[1613] Prompt Sentence Examples

[1614] Use the following prompt for the generative AI model:

[1615] Please calculate the risk level based on the daily behavior pattern data of user ID: user123, comparing it with the latest information (location information, activity data, etc.) in the event of a disaster. Provide the following data:

[1616] Daily behavior pattern data: [Detailed data]

[1617] Latest data: [Detailed data]

[1618] In this way, the present invention automatically checks the user's safety during a disaster, quickly notifies emergency contacts, and supports necessary responses, thereby ensuring the safety of victims and improving the efficiency of rescue operations.

[1619] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1620] Step 1:

[1621] Collect user location and activity data.

[1622] Input: Sensor data such as location information, activity data, and signal reception status from the user's smartphone.

[1623] Processing: The smartphone collects this data at regular intervals, converts it into a data format, and sends it to the server.

[1624] Output: A set of user location and activity data sent to the server.

[1625] Step 2:

[1626] The server stores the user's daily behavior patterns.

[1627] Input: Location, activity data, and other sensor data sent from your smartphone.

[1628] Processing: The server stores the received data in a database and integrates it with past data, thereby accumulating daily behavior patterns.

[1629] Output: User's daily behavior pattern data stored in a database.

[1630] Step 3:

[1631] Receive disaster information in real time.

[1632] Input: Real-time data from disaster information services (earthquake occurrence information, weather warnings, fire information, large-scale accident information, etc.).

[1633] Processing: The server calls the API and obtains disaster information.

[1634] Output: Disaster information data stored on the server.

[1635] Step 4:

[1636] Identify users in your target area.

[1637] Input: Disaster information (data on the location and scale of the disaster), user's daily behavior pattern data (location information).

[1638] Processing: The server analyzes the disaster area and extracts users who are in that area from the database.

[1639] Output: A list of users in the disaster area.

[1640] Step 5:

[1641] Collect the latest data when a disaster occurs.

[1642] Input: Disaster notification, latest location information of the user's smartphone, activity data, and signal reception status.

[1643] Processing: Upon receiving disaster information, the smartphone immediately collects the latest sensor data and sends it to the server.

[1644] Output: The latest sensor data sent to the server.

[1645] Step 6:

[1646] The latest data is compared with daily behavior patterns to calculate the level of risk.

[1647] Input: Latest sensor data, daily behavior pattern data.

[1648] Processing: The server uses the generated AI model to compare the latest data with daily behavior patterns and calculate the user's risk level. The prompt sentence is input into the model and the analysis results are obtained.

[1649] Output: Calculated user risk level (levels such as normal, caution, danger, etc.).

[1650] Step 7:

[1651] If the risk level exceeds a certain threshold, an alert will be issued to check the safety of the person.

[1652] Input: Risk data, user's emergency contact information.

[1653] Processing: If the server detects a risk level that exceeds the threshold, it will automatically send an email or SMS to the user's emergency contacts to confirm their safety.

[1654] Output: Safety alert sent to emergency contacts.

[1655] Step 8:

[1656] Automatically suggests evacuation routes to users at high risk.

[1657] Input: User's current location, risk level data, and surrounding environment information.

[1658] Processing: The server uses AI to analyze surrounding geographical information and disaster information to generate an appropriate evacuation route, which is then sent to the user's smartphone.

[1659] Output: Evacuation route displayed on the user's smartphone.

[1660] 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.

[1661] This invention is an automatic safety confirmation system for use in disaster situations, and by combining it with an emotion engine that recognizes the user's emotions, it is possible to perform more accurate risk assessments. This system uses AI to analyze smartphone usage data, and aims to automatically confirm and notify the user's safety when a disaster occurs.

[1662] Overall system configuration

[1663] The system consists of the following main elements:

[1664] 1. A means of collecting and storing users' daily behavior patterns

[1665] 2. Means of receiving disaster information in real time

[1666] 3. A method for identifying users in target areas based on disaster information

[1667] 4. A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) in the event of a disaster

[1668] 5. A method for calculating risk by comparing the latest collected data with daily behavior patterns

[1669] 6. A method for issuing safety confirmation alerts to pre-registered contacts when the risk level exceeds a certain threshold

[1670] 7. Emotion engine that recognizes user emotions

[1671] Specific embodiments of each means will be described in detail below.

[1672] Collecting and storing users' daily behavior patterns

[1673] As users use their smartphones on a daily basis, the device collects location information, app and web usage, activity information (e.g., walking, driving, stationary status), and health data (number of steps, heart rate, etc.). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[1674] The server receives the data sent from the device and stores it in a database, thereby accumulating the user's daily behavioral patterns for later AI analysis.

[1675] Learning daily behavior patterns

[1676] The server periodically retrieves daily behavior data stored in the database and inputs it into the AI ​​model. The AI ​​model analyzes this data and learns the user's usual behavior patterns (for example, usual commuting route and activity times). Based on the learning results, a behavioral pattern profile for each user is generated and continuously updated.

[1677] Receiving disaster information and identifying affected areas

[1678] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the area where the disaster occurred is identified and users in that area are extracted from the database.

[1679] Data collection in the event of a disaster

[1680] When the device receives disaster information, it immediately collects the current location, the last update time, information about nearby users (via short-range communication), and the radio wave reception status. This data is then promptly sent to the server.

[1681] Comparison of the latest data and calculation of risk

[1682] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change at all even during normal activity times, it will be detected as an abnormal pattern.

[1683] Use of emotion engine

[1684] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions, voice, and actions, and sends it to the server.

[1685] The server uses the emotion data sent by the user to assess the risk level. The emotion data indicates the user's stress level and tension, and the accuracy of the risk level is improved based on this. For example, if a user is in an abnormally stressed state when a disaster occurs, this information is reflected in the risk level calculation.

[1686] Issuance of safety confirmation alerts

[1687] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[1688] Specific examples

[1689] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives and stores this data, and the AI ​​learns user B's daily behavioral patterns.

[1690] One day, an earthquake occurs, and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area, and extracts from the database whether User B is in that area.

[1691] User B's device receives the earthquake information and immediately collects and sends the current location information, last update time, information about nearby users, signal reception status, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[1692] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family or workplace). This is expected to enable User B's safety to be confirmed quickly and accurately, and to enable necessary rescue operations to be carried out promptly.

[1693] The processing flow will be explained below.

[1694] Step 1:

[1695] The device collects data on the user's daily activities (location information, activity information, app usage, health data, and emotional data). Emotional data is collected by an emotion engine that recognizes the user's facial expressions, voice, and behavior.

[1696] Step 2:

[1697] The data collected by the terminal is sent to the server at regular intervals (for example, every 15 minutes).

[1698] Step 3:

[1699] The server receives the data sent from the terminal and stores it in a database.

[1700] Step 4:

[1701] The server inputs the stored data into the AI ​​model, which learns the user's daily behavioral patterns and emotional data. Based on the learning results, a behavioral pattern profile for each user is generated and updated.

[1702] Step 5:

[1703] The server receives disaster information such as earthquakes, weather warnings, fires, and large-scale accidents in real time from various disaster information services.

[1704] Step 6:

[1705] The server identifies the disaster-affected area based on the disaster information received, and extracts users who are in that area from the database.

[1706] Step 7:

[1707] When the device receives disaster information, it immediately collects the current location information, the last update time, information about nearby users, radio wave reception conditions, and emotional data.

[1708] Step 8:

[1709] The latest data collected by the device is sent to the server.

[1710] Step 9:

[1711] The server compares the latest data received from the device with the daily behavioral pattern profile and emotional data, and the AI ​​calculates the risk level, taking into account the user's emotional data (e.g., stress level and tension).

[1712] Step 10:

[1713] If the server detects that the risk level exceeds a certain threshold, it automatically sends a safety confirmation alert to the user's registered contacts.

[1714] Step 11:

[1715] The server will send an alert along with detailed information such as the user's last location, last update time, information about nearby users, radio wave reception status, and emotional data.

[1716] Example 2

[1717] 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."

[1718] In the event of a disaster, it is extremely important to quickly and accurately confirm the safety of users and take appropriate action. However, conventional systems do not take into account the user's daily behavioral patterns or emotional state in risk assessment, resulting in low accuracy in safety confirmation. The present invention aims to solve these problems.

[1719] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in a target area based on the received disaster information, means for collecting the latest data of users (location information, last update time, information on surrounding users, and communication environment) when a disaster occurs, means for comparing the collected latest data with daily behavior patterns to calculate a risk level, means for recognizing user emotion data and incorporating it into risk assessment, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables highly accurate risk assessment that takes into account the user's behavior patterns and emotional state.

[1720] "Means for collecting and storing users' daily behavioral patterns" refers to means for collecting users' daily activities, location information, app usage, and activity information, and storing this in a database.

[1721] "Means for receiving disaster information in real time" refers to means for obtaining disaster information such as earthquakes, weather, fires, and large-scale accidents in real time and transmitting it to a server.

[1722] The "means for identifying users in a target area based on received disaster information" is a means for identifying a disaster-affected area from received disaster information and extracting users who are in that area from a database.

[1723] "Means for collecting the latest user data (location information, last update time, surrounding user information, communication environment) when a disaster occurs" refers to means for collecting the user's current location information, last update time, user information via surrounding communications, and the status of radio waves and the communication environment when a disaster occurs.

[1724] The "means for calculating the risk by comparing the latest collected data with daily behavior patterns" is a means for comparing the latest collected data with the user's saved daily behavior pattern profile, evaluating the difference, and calculating the risk.

[1725] "Means for recognizing the user's emotional data and incorporating it into risk assessment" refers to means for collecting and analyzing data on the user's emotional state (e.g., facial expressions and voice) and reflecting it in risk assessment.

[1726] "Means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold" refers to a means for automatically sending a safety confirmation alert to the user's registered emergency contacts (family, workplace, etc.) when the calculated risk level exceeds a pre-set threshold.

[1727] This invention is an automatic safety confirmation system for use in the event of a disaster, which performs more accurate risk assessment based on the user's daily behavioral patterns and emotional data. The system uses AI to analyze smartphone usage data and automatically confirms and notifies the user's safety in the event of a disaster.

[1728] System Configuration

[1729] The system consists of the following main elements:

[1730] 1. A means of collecting and storing users' daily behavior patterns

[1731] The device collects data on the user's daily behavior using the smartphone's GPS, health sensors, app usage logs, camera, microphone, etc. This data includes location information, activity data (walking, driving, and standing still), heart rate, and app usage.

[1732] The collected data is sent to the server at regular intervals (for example, every 15 minutes) and stored on the server.

[1733] 2. A means of learning daily behavior patterns

[1734] The server periodically retrieves daily behavior data stored in the database and analyzes it using a generative AI model, primarily using deep learning frameworks such as TensorFlow and PyTorch.

[1735] The AI ​​model learns the user's usual behavioral patterns (for example, daily commuting route and activity times) and generates and updates a behavioral pattern profile based on this.

[1736] 3. Means of receiving disaster information and identifying affected areas

[1737] The server works in conjunction with various disaster information services (e.g., the Japan Meteorological Agency's API) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents.

[1738] When disaster information is received, the area where the disaster occurred is identified based on that information, and users who are in that area are extracted from the database.

[1739] 4. How to collect the latest user data when a disaster occurs

[1740] When the terminal receives disaster information, it immediately collects the current location information, the last update time, information on nearby users (via short-range communication), and the communication environment.

[1741] This data is quickly transmitted to the server.

[1742] 5. Comparing the latest data with daily behavior patterns and calculating the risk level

[1743] The server compares the latest data sent from the device with the stored daily behavior pattern profile.

[1744] The generative AI evaluates the differences in this data and calculates the risk level, detecting anomalies based on deviations from normal behavioral patterns.

[1745] 6. Methods for Recognizing User Emotional Data and Incorporating It into Risk Assessment

[1746] The device collects emotional data from the user's facial expressions, voice, and actions, and sends it to a server using a camera and microphone.

[1747] The server performs a risk assessment based on the emotional data analyzed by the emotion engine, taking into account the user's stress level and tension.

[1748] 7. How to issue safety confirmation alerts

[1749] If the calculated risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family or workplace).

[1750] The alert includes the user's last location, last update time, information about nearby users, communication environment, emotional data, etc.

[1751] Specific examples

[1752] For example, user B uses a smartphone on a daily basis, and location information, activity data, and emotional data are sent from the device to the server. The server receives this data and stores it in a database, and the generative AI model learns user B's daily behavioral patterns.

[1753] One day, an earthquake occurs and the server receives earthquake information from the Japan Meteorological Agency. It confirms that the earthquake occurred in a specific area and extracts from the database whether User B is in that area.

[1754] User B's device receives the earthquake information and immediately sends its current location, last update time, information about nearby users, communication environment, and emotional data to the server. The server analyzes this data and calculates the risk level based on the difference between User B's behavioral patterns and emotional data.

[1755] Because the danger level exceeds the threshold, the server automatically sends a safety confirmation alert and detailed information to User B's registered contacts (e.g., family, workplace). This process is expected to quickly and accurately confirm User B's safety, and enable necessary rescue operations to be carried out promptly.

[1756] Prompt Sentence Examples

[1757] "Please explain in detail your automated safety confirmation system in the event of a disaster. Please tell us the specific processing steps, from data collection to risk assessment and safety confirmation alerts. Please also include examples of application in real-life scenarios."

[1758] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1759] Processing Steps

[1760] Step 1: Data collection

[1761] The device collects information about the user's daily activities, including location information using GPS, activity information obtained from smartphone sensors (e.g., walking, driving, standing still), app usage, and heart rate.

[1762] Input: User's smartphone usage, sensor information

[1763] Output: Collected daily behavior data (e.g., location information, heart rate)

[1764] Specific operation: The device activates the sensor every 15 minutes and packages the acquired data in JSON format.

[1765] Step 2: Send data

[1766] The device sends the collected daily behavior data to the server at regular intervals (e.g., every 15 minutes).

[1767] Input: Collected daily behavior data

[1768] Output: Data sent to the server

[1769] Specific operation: The device sends JSON formatted data to the server's API endpoint using HTTPS.

[1770] Step 3: Save Data

[1771] The server receives the daily activity data sent from the terminal and stores it in a database.

[1772] Input: Daily activity data sent from the device

[1773] Output: Behavioral data stored in a database

[1774] Specific behavior: The server parses the received JSON data and stores it in the database in the appropriate format.

[1775] Step 4: Learning behavioral patterns

[1776] The server periodically retrieves the behavioral data stored in the database, inputs it into the generative AI model, and analyzes it.

[1777] Input: Saved daily activity data

[1778] Output: Learned behavioral pattern profile

[1779] Specific operation: The server uses TensorFlow and PyTorch to run a behavior prediction model and learn user behavior patterns.

[1780] Step 5: Receiving disaster information

[1781] The server receives information in real time from the disaster information service.

[1782] Input: Data from disaster information service

[1783] Output: Disaster information received by the server

[1784] Specific operation: The server periodically checks the API endpoint to obtain the latest disaster information.

[1785] Step 6: Identify users in your target area

[1786] The server identifies the disaster-affected area from the received disaster information and extracts users within that area.

[1787] Input: Disaster information, user location information in the database

[1788] Output: A list of users within the identified target area

[1789] Specific operation: The server uses GIS to map the disaster area and extracts users in that area from the database.

[1790] Step 7: Collect the latest data

[1791] When the terminal receives disaster information, it collects the current location information, the last update time, information on nearby users, and the communication environment.

[1792] Input: Disaster information, device sensor information

[1793] Output: Latest collected data (location information, communication environment, etc.)

[1794] What it does: The device performs GPS and Bluetooth scans to collect and send the latest data.

[1795] Step 8: Data comparison and risk calculation

[1796] The server compares the latest data sent from the device with the saved daily behavior patterns and calculates the level of risk.

[1797] Input: Latest data, daily behavior pattern profile

[1798] Output: Calculated risk

[1799] Specific operation: The server uses a generative AI model to detect abnormalities in behavioral patterns and evaluates the level of risk based on the results.

[1800] Step 9: Collect and evaluate emotion data

[1801] The terminal collects the user's emotional data (facial expressions, voice, etc.) and sends it to the server.

[1802] Input: User's facial expression and voice data

[1803] Output: Parsed emotion data

[1804] Specific operation: The device uses the camera and microphone to capture emotion data and sends it to the server.

[1805] Step 10: Finalize your risk assessment

[1806] The server recalculates the risk assessment based on the emotion data.

[1807] Input: Emotion data, initial risk assessment

[1808] Output: Adjusted risk assessment

[1809] Specific operation: The server takes in the emotional data, reassess the danger level, and generates the final result.

[1810] Step 11: Issue an alert

[1811] If the calculated risk level exceeds a threshold, the server sends a safety confirmation alert to the registered destination of the target user.

[1812] Input: Adjusted risk assessment, user registration information

[1813] Output: Safety confirmation alert issued

[1814] Specific operation: The server generates an alert message containing the necessary information (e.g., last location, nearby user information, etc.) and sends it to the registered recipient via email or SMS.

[1815] In this way, the system takes into account the user's daily behavior and emotional state, enabling quick and highly accurate confirmation of safety during a disaster.

[1816] (Application example 2)

[1817] 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."

[1818] Although systems for confirming users' safety during disasters already exist, they typically rely on users to provide information voluntarily, which often results in inaccurate information being unavailable. Furthermore, these systems assess risk based solely on location and activity data, without taking into account the user's emotional state, limiting the accuracy of risk assessments. Therefore, there is a need for a system that can more quickly and accurately confirm users' safety and provide appropriate notifications.

[1819] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting and saving daily behavior patterns of users, means for receiving disaster information in real time, means for identifying users in the target area based on the received disaster information, means for collecting user emotion data and using it for risk assessment, means for calculating the risk level by comparing the latest collected data with daily behavior patterns, and means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold. This enables more accurate risk assessment and faster safety confirmation by analyzing the user emotion data and reflecting the stress level and tension level in the risk assessment.

[1820] "User's daily behavior patterns" refers to the totality of data such as the user's daily actions, movements, and app usage.

[1821] "Disaster information" refers to information about emergencies such as earthquakes, fires, weather warnings, and large-scale accidents.

[1822] "Target area" refers to the geographical area identified as being affected by a disaster based on disaster information.

[1823] The "latest data" is real-time data including the user's location information, the last update time, information about nearby users, radio wave reception conditions, and emotion data at the time of the disaster.

[1824] The "risk level" is a numerical value or classification of the degree of risk in the situation in which the user is placed.

[1825] "Emotion data" is data that indicates the emotional state of the user as recognized from facial expressions, tone of voice, and the like.

[1826] "Risk assessment" is the process of calculating a user's risk level based on collected data.

[1827] A "safety confirmation alert" is a message or notification sent to a contact to confirm whether the user is safe.

[1828] The present invention is an automatic safety confirmation system for use in the event of a disaster, which collects data on users' daily behavior patterns and emotions and performs highly accurate risk assessments in the event of a disaster. Hereinafter, embodiments of the present invention will be described in detail.

[1829] Overall system configuration

[1830] This system is mainly composed of three elements: a server, a terminal, and a user.

[1831] Collecting and storing daily behavior patterns

[1832] As users use their smartphones on a daily basis, the device collects location information, activity data (e.g., walking, driving, and standing still), app usage, and emotional data (e.g., facial expressions and tone of voice). The collected data is sent to a server at regular intervals (e.g., every 15 minutes).

[1833] Learning daily behavior patterns

[1834] The server periodically retrieves daily behavior data stored in the database and inputs it into an AI model (e.g., TensorFlow, PyTorch). The AI ​​model analyzes this data and learns the user's usual behavioral patterns (e.g., commuting route, activity times). Based on the learning results, a behavioral pattern profile is generated and continuously updated.

[1835] Receiving disaster information and identifying affected areas

[1836] The server works in conjunction with various disaster information services (e.g., real-time data from the Japan Meteorological Agency) to receive real-time disaster information such as earthquakes, weather warnings, fires, and large-scale accidents. When disaster information is received, the server identifies the area where the disaster occurred and extracts users who are in that area from the database.

[1837] Data collection in the event of a disaster

[1838] When a device receives disaster information, it immediately collects its current location, the last update time, information about nearby users (e.g., device detection via short-range communication), and the status of radio wave reception. This data is then promptly sent to the server. The device also collects emotion data and sends it to the server.

[1839] Comparison of the latest data and calculation of risk

[1840] The server compares the latest data sent from the device with the stored daily behavioral pattern profile. AI evaluates the differences between these data and calculates the level of risk. For example, if the device's location information does not change even during normal activity times, it will be detected as an abnormal pattern.

[1841] Use of emotion engine

[1842] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine collects emotion data from the user's facial expressions and tone of voice and sends it to a server. The server analyzes the emotion data and reflects stress levels and tension in the risk assessment, improving the accuracy of the risk assessment.

[1843] Issuance of safety confirmation alerts

[1844] If the risk level exceeds a certain threshold, the server automatically sends a safety confirmation alert to the user's registered contacts (e.g., family, workplace), along with detailed information such as the user's last location, last update time, information about nearby users, signal reception status, and emotional data.

[1845] Specific examples

[1846] For example, when a user uses a smartphone on a daily basis, location information, app usage status, and emotional data are sent from the device to a server. The server receives and stores this information, and AI learns the user's daily behavioral patterns. One day, an earthquake occurs. The server receives disaster information, confirms that the disaster occurred in a specific area, and extracts from the database whether the user is in that area. The device receives the earthquake information and immediately collects current location information, last update time, information about surrounding users, signal reception status, and emotional data, and sends them to the server. The server analyzes this data and calculates the level of risk based on the difference between the behavioral pattern and the emotional data. If the level of risk exceeds a threshold, the server automatically sends a safety confirmation alert and detailed information to the user's registered contacts.

[1847] Example prompts for generative AI models

[1848] Enter the user's face photo and voice data:

[1849] Facial photo data:[path_to_image]\

[1850] Audio data:[path_to_audio]\

[1851] \

[1852] Parse the following data attributes:

[1853] 1. Emotion recognition (e.g., joy, anxiety, anger, sadness)

[1854] 2. Tone of voice (e.g., nervous, stressed, calm)

[1855] \

[1856] As output, return the user's emotional state and its associated numerical data (e.g., stress level 0-100).

[1857] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1858] Step 1:

[1859] Data collection

[1860] The device collects the user's daily behavioral patterns. Specifically, it obtains location information, activity data, app usage...

Claims

1. A means for collecting and storing daily behavior patterns of a user; A means for receiving disaster information in real time; A means for identifying users in a target area based on the received disaster information; A means of collecting the latest user data (location information, last update time, information on nearby users, and radio wave reception status) when a disaster occurs; A method for calculating the risk level by comparing the latest collected data with daily behavior patterns, and A means for issuing a safety confirmation alert to pre-registered notification destinations when the risk level exceeds a certain threshold; A system including:

2. A means of learning users' daily behavior patterns using AI, The system of claim 1 , further comprising means for generating a behavioral pattern profile based on the learning results.

3. The system according to claim 1 , further comprising means for receiving earthquake information, weather warning information, fire information, and large-scale accident information in cooperation with a disaster information providing service.

Images