Monitoring system, information processing device, information processing method, and program

Abstract

This invention provides a monitoring system, information processing device, information processing method, and program that enable monitoring from childhood to old age using a single system, without requiring the use of different monitoring systems for each generation, while ensuring continuity of operation. [Solution] The monitoring system 1 comprises a server 20, a dedicated monitoring terminal 10, and a communication terminal 30. The server comprises a switching control unit 21 that switches the operating mode of the dedicated monitoring terminal from child monitoring mode to elderly monitoring mode based on attribute information including the user's age information, and a UI management unit 23 that acquires the operation pattern in child monitoring mode and determines the display settings based on the operation pattern in elderly monitoring mode. The dedicated monitoring terminal 10 comprises a GPS function and a display control unit 13 that controls the screen display based on the display settings received from the server 20. The communication terminal 30 comprises a monitoring display unit 305 that receives and displays monitoring information from the server 20.

Description

Technical Field

[0001] The present disclosure relates to a monitoring system, an information processing apparatus, an information processing method, and a program.

Background Art

[0002] Conventionally, a monitoring service for children and a monitoring service for the elderly have been provided as independent systems (Patent Document 1). However, inheritance of data and continuity of operability according to generational changes have not been considered.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the above prior art, there were problems such as the need for different systems for each generation.

[0005] Therefore, an object of the present disclosure is to provide a monitoring system that solves the above problems.

Means for Solving the Problems

[0006] To achieve the above objective, the generational succession type monitoring system of this disclosure comprises a server, a dedicated monitoring terminal, and a communication terminal, wherein the server comprises a switching control unit that switches the operating mode between a child monitoring mode and an elderly monitoring mode based on the age information of the person being monitored, and a UI management unit that acquires operation patterns for the dedicated monitoring terminal in the child monitoring mode and determines display settings based on the operation patterns in the elderly monitoring mode, the dedicated monitoring terminal has a GPS function and comprises a display control unit that controls the screen display based on the display settings received from the server, and the communication terminal comprises a monitoring display unit that receives location information from the server and displays the monitoring status. [Effects of the Invention]

[0007] According to this disclosure, generational succession enables long-term monitoring services with a single system, avoiding redundant device investments. Furthermore, by utilizing an individual's past operation patterns, the burden of learning how to operate the system in old age is significantly reduced, and a familiar operating feel can be maintained. The effects described above are examples of effects obtainable by the embodiments of this disclosure, and the effects obtainable by this disclosure are not limited to these. Furthermore, in some embodiments of this disclosure, some of the above effects may not be obtained, but even in such cases, the invention is not excluded from the technical scope of this disclosure. In addition, this disclosure may produce other effects different from those described above. [Brief explanation of the drawing]

[0008] [Figure 1] This is an overview diagram of the entire system. [Figure 2] This is a server hardware configuration diagram. [Figure 3] This is a hardware configuration diagram for a dedicated monitoring terminal. [Figure 4] This is a hardware configuration diagram of a communication terminal. [Figure 5] This is a server function block diagram. [Figure 6] This is a block diagram of the functions of a dedicated monitoring terminal. [Figure 7] This is a flowchart of the generation succession process. [Figure 8] This is a flowchart of the step-by-step notification process. [Figure 9] This is an example of the child mode UI screen. [Figure 10] This is an example of the UI screen for the elderly mode. [Figure 11] This is a data structure diagram. [Figure 12] This is a processing sequence diagram. [Figure 13] This is a screen transition diagram. [Modes for carrying out the invention]

[0009] As shown in Figure 1, the generational succession type monitoring system 1 of this disclosure has a configuration in which a dedicated monitoring terminal 10, a server 20, and a communication terminal 30 are connected via a network N.

[0010] The dedicated monitoring terminal 10 is a GPS-enabled device carried by the person being monitored (a child or an elderly person). The server 20 is an information processing device that centrally manages the entire monitoring service. The communication terminal 30 is a communication device such as a smartphone or tablet that family members use to check the monitoring status.

[0011] Network N consists of the Internet, LTE / 5G mobile communication network, and wired / wireless communication means, either individually or in combination. The communication protocol is optimized for the communication situation through the adaptive selection of HTTPS, MQTT, WebSocket, and CoAP. Data transmission frequency is dynamically adjusted to 5-minute intervals under normal circumstances and 10-second intervals in emergencies. These intervals are examples only.

[0012] As shown in FIG. 1, the generational inheritance monitoring system 1 of the present disclosure has a configuration in which a dedicated monitoring terminal 10, a server 20, and a communication terminal 30 are connected via a network N. In this specification, the term "server" is not limited to a single physical server that performs centralized processing, but includes a concept that encompasses a group of multiple servers, a group of microservices, cloud computing resources, edge computing resources, or a combination thereof that cooperate to provide the same monitoring service. The important point is that these processing resources uniformly implement generational inheritance functions such as operation mode switching corresponding to the age change of the monitored person, UI personalization based on operation patterns, and step-by-step notification control. The physical location and implementation form are not the essence of this technology. Therefore, differences in technical implementation methods such as distributed processing, edge processing, and P2P communication do not affect the technical scope of the present disclosure.

[0013] As shown in FIG. 1, the generational inheritance monitoring system 1 of the present disclosure has a unified system architecture in which a plurality of dedicated monitoring terminals 10 and communication terminals 30 cooperate with each other, centered around a single server 20, a unified database infrastructure, and a common authentication system. It performs consistent user ID management from childhood to old age, continuous data accumulation and analysis, and step-by-step function inheritance, which is technically different from simple product migration or independent service combinations. This system is distinguished from conventional age-separated independent systems in that it comprehensively supports the entire life cycle of the monitored person on a single technical platform.

[0014] Regarding the technical details of data inheritance, the data accumulated during childhood (0 - 18 years old) consists of location history (GPS coordinates, time, residence time), operation history (UI operation pattern, frequency, accuracy), behavior pattern (movement route, activity time, frequency of going out), and emergency response history (fall, getting lost, SOS history). After anonymizing the personal identification information (k-anonymization, k ≥ 5), these data are utilized for personalization of the elderly monitoring service.

[0015] As shown in FIG. 2, the server 20 is designed as a high-performance computing platform that executes core control processing for the entire system. It adopts an enterprise-grade server architecture in which a CPU 201, a memory 202, a storage 203, a network interface 204, a GPU 205, and a power supply device 206 are integrally arranged by a system bus. Each hardware component is configured with emphasis on redundancy and scalability to cope with high-load processing such as position information processing from a large number of monitoring dedicated terminals, operation pattern analysis by machine learning, and real-time notification processing.

[0016] The CPU 201 adopts a high-performance multi-core processor (equivalent to 3.0 GHz, 24 cores and 48 threads) to efficiently execute simultaneous request processing from a large number of users and complex machine learning operations. With this configuration, it is possible to execute in parallel the age determination process of the monitored person, operation pattern analysis, and anomaly detection algorithm, significantly improving the response performance of the entire system.

[0017] The memory 202 is equipped with DDR4-3200 ECC RAM of 128 GB or more to realize high-speed processing of large-scale user data and machine learning models, and guarantees stable operation 24 hours a day, 365 days a year with an error detection and correction function. Also, by optimizing the memory bandwidth, it maximizes the database access performance and cache efficiency.

[0018] The storage 203 adopts a tiered configuration that uses an SSD (NVMe PCIe 4.0, 2TB) for system programs that require fast access and user data that is frequently accessed, and an HDD (SATA 3.0, 20TB) for historical data and analysis results that require long-term storage. This two-layer storage configuration achieves both high-speed responsiveness required for real-time processing and long-term storage of large amounts of data while maintaining cost performance.

[0019] Network interface 204 employs an active-standby redundancy configuration with two 10GbE ports to ensure high availability. It also implements a mechanism to automatically switch over if one port fails, thereby preventing service interruptions.

[0020] The GPU205 features a graphics processor with 16GB of GDDR6 and computing performance equivalent to 2560 CUDA cores, designed for high-speed execution of machine learning algorithms necessary for learning operation patterns and generating personalized UIs. This achieves a 10-50 times improvement in processing speed compared to conventional CPU-only configurations.

[0021] To ensure stable operation of the entire system, power supply unit 206 has a redundant configuration of 1200W x 2 that meets the 80 Plus Platinum efficiency standard. By working in conjunction with a UPS (uninterruptible power supply), it guarantees continuous operation for 30 minutes even in the event of a power outage, ensuring safe system shutdown and data protection.

[0022] As shown in Figure 3, the dedicated monitoring terminal 10 includes a main processor 101, memory / storage 102, GPS receiver 103, communication module 104, sensor group 105, battery 106, display 107, voice input unit 108, and voice output unit 109. These are integrated by a system bus.

[0023] The main processor 101 employs a high-performance SoC (System on Chip) to execute multiple processes in parallel, such as real-time location information processing, sensor data analysis, speech recognition processing, and screen display control.

[0024] The memory and storage 102 configuration employs LPDDR5 8GB (low-power, high-speed memory) and UFS 3.1 128GB (high-speed flash storage) to enable smooth multitasking and storage of large amounts of operation history data.

[0025] The GPS receiver 103 employs, for example, a multi-band RTK compatible receiver to achieve positioning accuracy of 1 cm. The positioning system supports five systems: GPS, GLONASS, BeiDou, Galileo, and QZSS (Michibiki).

[0026] The communication module 104 enables multimode communication such as 5G NR Sub-6, LTE-M, NB-IoT, Wi-Fi 6E, and Bluetooth 5.2 LE.

[0027] The sensor group 105 integrates, for example, a 3-axis accelerometer (±16G, 14-bit resolution), a 3-axis gyroscope (±2000dps), a 3-axis magnetic sensor, a barometric pressure sensor, and a temperature and humidity sensor. The battery 106 uses a Li-ion polymer battery (3.7V, 5000mAh, 18.5Wh) and provides 7-10 days of continuous operation under normal use and up to 30 days in power-saving mode.

[0028] The Battery 106 employs a high-energy-density Li-ion polymer battery (3.7V, 5000mAh, 18.5Wh) to enable long-term continuous use, achieving 7-10 days of continuous operation in normal use mode and up to 30 days in power-saving mode.

[0029] The Display 107 employs a 2.4-inch TFT LCD (240 x 320 pixels, 65K color support) to achieve both outdoor visibility and power-saving operation. Its automatic brightness adjustment function, linked to an ambient light sensor, ensures clear display at a maximum brightness of 800 cd / m², from low-light indoor environments to direct sunlight outdoors. Furthermore, screen protection with tempered glass (equivalent to Gorilla Glass) and a waterproof coating ensures durability against scratches and water damage during everyday use.

[0030] The audio input section 108 achieves high-precision audio input with suppressed ambient noise through a beamforming configuration using two MEMS microphones.

[0031] The audio output unit 109 uses a full-range speaker (maximum output of 5W, maximum sound pressure level of 85dB) to achieve clear audio output even outdoors.

[0032] The communication terminal 30 is a smartphone or tablet device. As shown in Figure 4, for example, it has a processor 301 equivalent to or better than an ARM Cortex-A76, 4GB or more of RAM 302, 64GB or more of storage 303, and communication functions 304 of LTE Cat.6 or higher, Wi-Fi 5 or higher, and Bluetooth 4.2 or higher. The communication terminal 30 is equipped with a monitoring display unit 305. The monitoring display unit 305 is a functional unit that displays monitoring information (location information, anomaly detection information, health status information, etc.) received from the server 20 on the display. The monitoring display unit 31 includes a map display function that displays the current location of the person being monitored on a map, a notification display function that displays an alert screen when an anomaly is detected, and a history display function that displays the behavioral history of the person being monitored in chronological order.

[0033] This document explains the technical differences between this technology and existing general-purpose platforms (smartwatch family settings, family management apps, etc.). General-purpose products progressively restrict management functions for children aged 13 and older, and do not offer specialized functions for the elderly (fall detection accuracy, wandering behavior pattern learning, medical institution collaboration, etc.). Family management apps automatically terminate at age 18, and there is no technical continuity for monitoring the elderly afterward. In contrast, this disclosure realizes a specialized monitoring system that integrates (1) high-precision sensing using a dedicated monitoring terminal (GPS accuracy 3m, fall detection sensitivity 3G, battery life 7-30 days), (2) personalized learning through 15 years of long-term data accumulation, (3) standard linkage with medical and nursing care systems (FHIR R4 compliant), and (4) autonomy protection through phased notifications. In particular, the personalization of the user interface in old age by utilizing the operation history from childhood 5-15 years is technically impossible to achieve with general-purpose products.

[0034] The technical scope and functional equivalence of the dedicated monitoring terminal 10 will be explained. In this disclosure, "dedicated monitoring terminal" is not limited to dedicated devices designed and manufactured primarily for monitoring purposes, but is a concept that encompasses all electronic devices used to realize monitoring functions. Specific examples of such devices include, for example, dedicated devices such as GPS monitoring terminals, smartwatches for the elderly, and emergency call terminals; general-purpose devices such as smartphones (running dedicated applications), tablet terminals, and general-purpose smartwatches; IoT sensor devices such as motion sensors, door sensors, camera systems, and voice recognition devices; and in-vehicle and infrastructure-linked devices such as in-vehicle GPS, public transport IC cards, and community monitoring system linkage terminals. Furthermore, electronic devices that perform the following functions are included in the technical scope of "dedicated monitoring terminals," regardless of their implementation form. (a) Location information acquisition function (GPS, Wi-Fi positioning, cellular positioning, Bluetooth positioning, etc.) (b) Biometric and behavioral information acquisition function (accelerometer, heart rate, fall detection, voice input, etc.) (c) Communication function (bidirectional data exchange with the server) (d) Personalized UI display function (screen display or equivalent means) The goal is to integrate these functions and contribute to generational monitoring services; the physical characteristics of the device, such as its specificity / versatility, individuality / distributedness, shape, and size, do not affect the essence of the technology.

[0035] As shown in Figure 5, the server 20 includes a switching control unit 21 and a location information management unit 22 (location information Control unit, UI management unit 23, notification control unit 24, database 25, communication control unit 26 These are integrated via a system bus. The location information management unit 22 will, in the event of an emergency, Detailed location information is transmitted to the communication terminal 30, and under normal circumstances, approximate location information (within a 500m radius) is transmitted. Includes a location information control function that transmits area information.

[0036] The switching control unit 21 of the server 20 implements age determination and mode switching functions. In this specification, "switching control unit" means means for executing control processing to realize generation succession functions, and its implementation location is not limited to inside the server. Specifically, if the "age-based operating mode switching" function is realized through the processor in the dedicated monitoring terminal 10, the home Wi-Fi router, edge server, cloud server, blockchain node, or their coordinated processing, then it falls under the category of "switching control unit," regardless of the implementing entity. Similarly, the UI management unit, described later, encompasses all technical means (such as in-device AI processing, distributed learning, and collaborative filtering) to realize the functions of acquiring and analyzing operation patterns and determining display settings. The essence of the technology is the personal optimization function of display settings that utilizes an individual's past operation history, and this function cannot be avoided by the decentralization and distribution of processing.

[0037] The switching control unit 21 consists of a user attribute information management module 211, an age determination module 212, and a mode switching execution module 213.

[0038] The user attribute information management module 211 comprehensively manages the basic information (name, date of birth, address), family relationship information (relationship, contact information, permission settings), medical and health information (medical history, medication information, allergies), and personal settings (language, screen settings, notification settings) of the person being monitored. The server 20 further includes an information sharing control unit 28. The information sharing control unit 28 is a functional unit that provides monitoring information to multiple communication terminals 30 at different information sharing levels based on the family relationship information of each communication terminal 30. Specifically, the information sharing control unit 28 controls the access permission level for each family member (Level 1: basic information only, Level 2: basic information + location information, Level 3: basic information + location information + medical information) and appropriately adjusts the balance between privacy protection and emergency response.

[0039] The age determination module 212 comprehensively performs calculations of the number of days elapsed since the date of birth, developmental stage assessment (physical, cognitive, and social development levels), and independence assessment (ADL and IADL assessment). The age determination module 212 comprehensively evaluates the number of days elapsed since birth, as well as developmental stage and independence levels, to determine the appropriate timing for switching to monitoring mode. The switching timing is basically controlled in the following five stages, but is flexibly adjusted according to the individual's developmental stage, family requests, and social circumstances. The basic patterns are shown below. Basic modes: Child mode (0-15 years old), Transition mode (15-18 years old), Independent mode (18-65 years old), Pre-elderly mode (65-75 years old), Elderly mode (75 years and older) The key point is that operation learning data from the early stages of life (generally within the range of 0-20 years old) is used for monitoring services in the later stages of life (generally within the range of 60 years and older). Specific implementation details such as age categories, the presence or absence of intermediate stages, and data retention periods do not affect the essential technical philosophy of the technology.

[0040] The mode switching execution module 213 performs the following processes based on the judgment result. At the switching timing, it performs a gradual transition from child mode → transition mode → independent mode → elderly mode, and automatically enables / disables functions in each mode, changes data access permissions, and inherits UI settings.

[0041] The UI management unit 23 consists of an operation history collection module 231, a pattern analysis engine 232, and a personalized UI generation module 233. The operation history collection module 231 collects operation history from both the dedicated monitoring terminal 10 and the communication terminal 30. From the dedicated monitoring terminal 10, it records touch coordinates, pressure, time, swipe trajectory, speed, screen dwell time, and error events at a sampling rate of 10 times per second. From the communication terminal 30, it acquires display setting history such as text size settings, icon size settings, screen layout settings, and color settings for the monitoring application.

[0042] In this system, generational succession includes not only the pattern where the person being monitored changes from a child to an elderly person, but also the pattern where the person providing the monitoring changes to an elderly person who is being monitored. Specifically, when elderly person A monitors grandchild B, they use the monitoring application on the communication terminal 30, and the display settings (font size, icon size, contrast settings, etc.) during that time are recorded by the operation history collection module 231. Later, when elderly person A themselves becomes the person being monitored and uses the dedicated monitoring terminal 10, the display setting history accumulated on the communication terminal 30 is applied as the initial settings of the dedicated monitoring terminal 10. As a result, elderly person A can continue to use the familiar operating environment they were accustomed to when using the monitoring application on the dedicated monitoring terminal 10, reducing the burden of learning new operations.

[0043] The operation history collection module 231 functions as a data collection system that records and analyzes in detail the user's operation behavior on the dedicated monitoring terminal 10 and the communication terminal 30. From the dedicated monitoring terminal 10, touch coordinates (x,y), pressure value (0-1024 levels), and contact time (in milliseconds) are precisely recorded by high-frequency sampling 10 times per second. In addition, for swipe operations, trajectory data (start coordinates, end coordinates, intermediate path points), velocity vector, and acceleration vector are recorded to quantify the individual's operation characteristics. Furthermore, the dwell time on each screen, screen transition patterns, types of operation errors (mis-tap, out-of-range tap, recognition failure), and error recovery actions are recorded in detail. From the communication terminal 30, the display setting history of the monitoring application is acquired. The display setting history includes font size settings (selection history of standard, large, extra-large, etc.), icon size settings (selection history of standard, enlarged, etc.), screen contrast settings, background color settings, notification volume settings, and screen brightness settings. These setting histories are accumulated throughout the period during which the caregiver (for example, an elderly person caring for their grandchild) uses the monitoring application on the communication terminal 30, and are used for the initial setup of the dedicated monitoring terminal 10 when the elderly person becomes the person being monitored. The collected data is encrypted and stored after personal identification information is anonymized, and is used in conjunction with the pattern analysis engine 232 to generate a personalized UI.

[0044] The pattern analysis engine 232 quantifies operation patterns through ensemble learning of Random Forest (100 decision trees), Gradient Boosting (LightGBM), and LSTM (128 units x 3 layers).

[0045] The UI management unit 23 further includes a learning control unit 234. The learning control unit 234 is a functional unit that inputs operation patterns continuously acquired in the elderly monitoring mode into the pattern analysis engine 232 and analyzes them using machine learning to continuously optimize the display settings. The learning control unit 234 performs model retraining in daily batch processing (2-4 AM) and updates personalization parameters in real time every three hours. This ensures continuous UI optimization that adapts to changes in the physical and cognitive functions of the elderly over time.

[0046] The personalized UI generation module 233 determines display setting parameters based on pattern analysis results and display setting history obtained from the communication terminal 30. In particular, the font size and icon size set by the elderly person as the caregiver on the communication terminal 30 are preferentially applied as the initial settings for the dedicated monitoring terminal 10. The parameters include font size (14pt × 0.8-2.0x), line spacing (1.2 × 0.8-2.0x), text color and background color (RGB 0-255 each), button size (48dp × 0.8-2.0x), voice guidance frequency, and screen transition speed. If a display setting history exists on the communication terminal 30, the most frequent or last set value in that history is adopted as the initial value for font size and icon size to reproduce the display environment that the elderly person is familiar with.

[0047] As shown in Figure 6, the dedicated monitoring terminal 10 includes a location information processing unit 11, a sensor information processing unit 12, a display control unit 13, a location information transmission unit 14, and a communication control unit 15.

[0048] The position information processing unit 11 consists of a GNSS receiving module 141, a position estimation filter 142, and a transmission control module 143. The GNSS receiving module 141 performs multiband reception (L1 / L2 / L5) with a 5Hz update. The position estimation filter 142 implements inertial navigation combined positioning using an Extended Kalman Filter, achieving positioning accuracy of 3m CEP in open sky environments, 5-10m CEP in urban areas, and 20-50m CEP indoors.

[0049] The sensor information processing unit 12 performs motion state estimation (stationary, walking, running, vehicle movement), abnormal behavior detection (deviation from normal range of movement, prolonged dwelling, sudden cessation of movement), and fall detection processing. For fall detection, a primary determination is made if the combined acceleration exceeds 3G for 0.1 seconds or more, followed by a secondary determination through motion pattern analysis over a 60-second period, and finally a final determination is made based on a voice confirmation response.

[0050] The server 20 further includes a fall detection unit 27. The fall detection unit 27 is a functional unit that determines the fall status of the person being monitored based on detection data from a 3-axis acceleration sensor included in the sensor group 105 of the dedicated monitoring terminal 10. The fall detection unit 27 implements a three-stage determination algorithm that performs acceleration threshold determination to detect when the combined acceleration exceeds 3G as the first stage, motion pattern analysis for 60 seconds after detecting a fall candidate as the second stage, and final determination by voice confirmation response as the third stage. The fall detection unit 27 works in cooperation with the notification control unit 24 to initiate a stepwise notification process when a fall is confirmed.

[0051] The display control unit 13 consists of a UI rendering engine 131, an adaptive display module 132, and a touch input processing unit 133. The UI rendering engine 131 implements a 60FPS rendering engine compatible with OpenGL ES 3.2. The touch input processing unit 133 implements multi-touch support (up to 10 points), pressure sensitivity detection (256 levels), and gesture recognition (tap, swipe, pinch, long tap).

[0052] The location information transmission unit 14 is responsible for communication control functions that ensure the secure and efficient transmission of GPS location data processed by the location information processing unit 11 to the server 20. It dynamically adjusts the transmission frequency, typically every 5 minutes and every 10 seconds in emergencies. It also adjusts the data compression ratio according to the communication conditions, ensuring reliable location information transmission even with limited communication bandwidth.

[0053] The communication control unit 15 centrally manages all communication processing between the dedicated monitoring terminal 10, the server 20, and the communication terminal 30, and automatically selects the most suitable communication protocol from HTTPS, MQTT, WebSocket, and CoAP depending on the communication situation. Furthermore, it maintains stable communication quality by performing automatic retransmission control in the event of a communication error, ensuring security through encrypted communication, and controlling the switching between multiple communication methods (5G, LTE, Wi-Fi).

[0054] The dedicated monitoring terminal 10 is further equipped with an identity notification unit 16. The identity notification unit 16 is a functional unit that outputs pre-registered identity information (name, emergency contact information) via voice output unit 109 in response to user operations (such as pressing and holding the emergency button for 3 seconds or entering a voice command). The identity notification unit 16 works in cooperation with the communication control unit 15 to acquire identity information from the server 20 and generates voice data by performing voice synthesis processing. This makes it possible for a third party to confirm the identity information even if the person being monitored is unconscious or in a similar state.

[0055] The adaptive display module 132 of the display control unit 13 dynamically adjusts the screen display in real time by reflecting the personalization parameters (in JSON format, encrypted) received from the server 20. The personalization parameters include the learning source data period, the learning algorithm identifier, and the fitness score, and this information can be checked externally from the settings menu on the terminal screen or from the log file. Furthermore, whether or not learning has been applied can be determined from the difference in response time when the UI settings are changed (personalized learning enabled: average 2.1 seconds, preset settings: average 0.3 seconds).

[0056] As shown in Figure 7, the generation succession process is carried out in the following steps.

[0057] At the start of processing, the system checks the status of all registered users. In S101, an age information check is performed once a month, and the switching conditions are determined (automatic age determination, manual determination by the guardian, and comprehensive AI determination). Automatic age determination is performed by calculating the number of days elapsed since the date of birth. Manual determination by the guardian is performed based on a switching request from the communication terminal 30. Comprehensive AI determination is performed by integrating developmental stage assessment, independence assessment, and behavioral pattern analysis. If the switching conditions are met, the process proceeds to S102; otherwise, the process ends.

[0058] In S102, the preparatory process involves saving the child mode settings, organizing and analyzing the operation history, and calculating personalized parameters. The operation history data is organized into detailed data for the most recent year (approximately 500MB) and summary data for the entire past period (approximately 50MB). As personalized parameters, the ranking of frequently used functions, average operation completion time, error occurrence tendency, and screen dwell time distribution are calculated. After the preparatory process is complete, the process proceeds to S103.

[0059] In S103, Graduation Mode is activated by displaying a congratulatory message, generating and sharing memory data, and notifying the designated successor. The memory data consists of a school route map (history of frequently traveled routes), an adventure map (history of newly visited places), and a growth timeline (history of expanding range of activity at each age). Sharing with family is performed based on privacy settings. After processing is complete, proceed to S104.

[0060] In S104, the system transitions to a dormant state by disabling GPS functionality, minimizing communication frequency (only checking for survival once a week), and performing inheritance preparation processing (registering new target information). After the transition to the dormant state is complete, the system proceeds to S105.

[0061] In S105, the system determines when to begin monitoring the elderly. The start conditions are determined by automatic start based on pre-set ages (65, 70, or 75 years old) or by a manual start request from a family member. If the start conditions are met, the system proceeds to S106; otherwise, it remains dormant (terminates).

[0062] In S106, the initial setup for elderly mode includes applying accumulated personalization parameters, increasing font size by 1.5-2.0 times, improving contrast by 2-3 times, and enabling elderly-specific functions (fall detection, wandering detection, medication reminder, emergency call, etc.). After the initial setup is complete, the process ends. Upon completion of the process, a completion notification will be sent to the user and their family, the processing results will be recorded in the system log, and the next scheduled periodic check will be set (end).

[0063] Furthermore, the generational succession process is not limited to the transition from childhood to old age of the person being monitored. For example, if elderly person A was monitoring grandchild B using the monitoring application on the communication terminal 30, the display settings history (text size settings, icon size settings, etc.) on elderly person A's communication terminal 30 is stored in the database 25 on the server 20. Later, when grandchild B grows up and is no longer a person being monitored, and elderly person A becomes the person being monitored and starts using the dedicated monitoring terminal 10, the stored display settings history is applied as the initial settings for the dedicated monitoring terminal 10. This process allows elderly person A to operate the new terminal with familiar text and icon sizes, significantly reducing the burden of learning how to operate it. The switching control unit 21 searches the display settings history of the communication terminal 30 associated with elderly person A's user ID and instructs the UI management unit 23 to apply the history.

[0064] This system provides a dual monitoring control function. Dual monitoring control is a control system that enables a state in which the first user (for example, an elderly person A aged 70) monitors the second user (for example, grandchild B), while the first user himself is simultaneously monitored by a third user (for example, C, who is the child of elderly person A and the parent of grandchild B).

[0065] Specifically, the switching control unit 21 of server 20 registers elderly person A as a user with both "monitoring provider" and "monitored person" attributes. While displaying grandchild B's monitoring information on elderly person A's communication terminal 30, it also transmits location information from elderly person A's own dedicated monitoring terminal 10 to C's communication terminal 30. This allows elderly person A to maintain their social role by participating in monitoring grandchild B while also ensuring their own safety. Dual monitoring control has the effect of building an intergenerational mutual monitoring network and improving the sense of security for the entire family.

[0066] Figure 8 is a flowchart of the step-by-step notification process executed by server 20. The step-by-step notification process in this embodiment includes continuous monitoring by the AI ​​anomaly detection system and a three-stage escalation function according to the severity of the detected anomaly.

[0067] First, the notification control unit 24 performs continuous anomaly detection processing using its AI anomaly detection function (S201). In S201, it analyzes various data such as sensor data, location information, and operation patterns to determine the occurrence of an anomaly. If no anomaly is detected (NO), the notification control unit 24 continues to perform the anomaly detection processing in S201. This ensures that the system maintains a 24 / 7 continuous monitoring state.

[0068] If an anomaly is detected in S201 (YES), the notification control unit 24 performs level detection processing (S202). In S202, the severity, urgency, and scope of impact of the detected anomaly are comprehensively analyzed and classified into three levels: Level 1 (mild), Level 2 (moderate), and Level 3 (severe). Level determination is performed based on the following criteria: (a) Level 1: Minor anomalies (e.g., slight deviation from the set range, temporary communication delay) (b) Level 2: Moderate abnormality (e.g., deviation from critical safety range, partial failure of function) (c) Level 3: Anomaly requiring immediate attention (e.g., life-threatening situation, complete system shutdown)

[0069] If Level 1 is determined in S202, the notification control unit 24 performs a direct response confirmation to the person (S203). In S203, a response is requested from the person being monitored's communication terminal 30 via voice call, message display, audible alarm, etc. Response confirmation is carried out using a combination of multiple means such as push notification, SMS, and voice guidance to ensure reliable identity verification.

[0070] If a response from the individual is confirmed in S203 (YES), the notification control unit 24 executes response processing (S204). In S204, the response content is recorded, the situation is confirmed, additional support is suggested as needed, and the process is terminated after confirming a return to a normal state. During response processing, the response time, response method, response content, etc. are saved in detail as a log and used to improve the accuracy of monitoring in the future. On the other hand, if no response is received from the person in S203 (NO), the notification control unit 24 performs a status escalation process (S205). In S205, the status is elevated from level 1 to level 2, and the severity of the anomaly is re-evaluated. The status escalation process includes collecting additional sensor information, checking the surrounding situation, and preparing emergency contact information.

[0071] If the situation is determined to be Level 2 in S202, or if it transitions to Level 2 through the promotion process in S205, the notification control unit 24 performs a response confirmation to family members, etc. (S206). In S206, the system notifies pre-registered emergency contacts (family, relatives, neighbors, etc.) of the situation via telephone, email, a dedicated application, etc., and requests a response.

[0072] If a response from a family member is confirmed in S206 (YES), the notification control unit 24 executes family response processing (S207). In S207, the system assists the family in confirming the situation, coordinating on-site responses as needed, and contacting relevant organizations. It also records the family's response status, estimated arrival time, and measures taken, and continuously follows up until appropriate support is completed.

[0073] On the other hand, if no response is received from family members or others in S206 (NO), the notification control unit 24 performs a status escalation process (S208). In S208, the status is elevated from Level 2 to Level 3, and preparations are made to treat it as an emergency. The status escalation process includes prior notification to emergency service agencies, mass notification to relevant parties, and detailed analysis of the situation on site.

[0074] If S202 determines that the situation is at level 3, or if the situation transitions to level 3 through the promotion process in S208, the notification control unit 24 executes emergency response (S209). In S209, the system automatically performs actions such as calling 119 (emergency services), contacting the police and fire departments, sending emergency notifications to neighbors and the management association, and contacting care managers and medical institutions.

[0075] Following the emergency response in S209, the notification control unit 24 executes the response completion process (S210). In S210, it checks the response status of each relevant organization, confirms arrival at the scene, confirms the implementation of emergency measures, and confirms the destination if hospital transport is required. It also sends a final report of the situation to all relevant parties and adjusts the future monitoring system.

[0076] This document details the technical significance of the operation pattern learning engine 100B. This disclosure is a UI personalization that technically utilizes an individual's past usage experience (procedural memory), and this technical concept is independent of the type and acquisition method of the learning data. In this disclosure, "operation patterns" include any information relating to an individual's past device usage behavior. This includes physical operation history (touch, swipe, voice input, etc.), physiological response history (eye tracking, electroencephalogram, heart rate variability, etc.), behavioral selection history (menu selection, frequency of function use, etc.), and environmental adaptation history (illumination and volume adjustment patterns, etc.). Methods for achieving personal optimization include a wide range of technological means, such as statistical learning, machine learning, genetic algorithms, neural networks, Bayesian estimation, collaborative filtering, biometric information analysis, and utilization of genetic information. The essence of the technology is "maintaining continuous ease of use in response to the chronological changes of the same individual (from childhood to old age)," and the diversity of technological means to achieve this objective does not limit the scope of rights. Therefore, circumvention through changes in learning algorithms, data types, personalization methods, etc., is included within the technical scope of this disclosure.

[0077] The specific structure of the operation pattern data is described below. Touch events are recorded using touch coordinates (x,y), pressure value (0-1024 levels), contact time (ms), and contact area (mm^2). Swipe trajectories are recorded using start coordinates, end coordinates, intermediate path points, velocity vector, and acceleration vector. Screen transition history is recorded using source screen ID, destination screen ID, transition method (button, swipe, voice), and transition time.

[0078] As a detailed record of error events, the system will record the error type (mistake, out-of-range tap, recognition failure, timeout), the screen where the error occurred, the error recovery action (re-tap, select a different operation, refer to help), and the time taken to recover. From this information, we will identify individual user difficulty patterns and elements that make learning difficult, and use this information to improve the UI for older adults.

[0079] The feature extraction process extracts temporal features (average operation time, operation interval, intraday variation pattern), spatial features (tap accuracy distribution, frequently used screen areas, avoided screen areas), frequency features (function usage frequency, error frequency, help reference frequency), and learning features (operation improvement rate, new function acquisition speed, forgetting pattern).

[0080] The detailed configuration of the machine learning model employs Random Forest (100 decision trees, maximum depth 15, Gini impurity criterion), Gradient Boosting (learning rate 0.1, 100 stages, regularization parameter 0.01), and LSTM (input dimension 64, 128 units x 3 hidden layers, dropout rate 0.2).

[0081] For training data management, a minimum of 30 days' worth of data (approximately 1000 operation events) is required, and convergence is judged based on reaching 85% accuracy within 90 days. The training update frequency involves daily batch processing (2-4 AM) for model retraining, and real-time personalization parameter updates are performed every 3 hours. Evaluation metrics include the percentage reduction in operation time, the percentage reduction in error rate, and the user satisfaction score (on a 5-point scale).

[0082] As shown in Figure 11, the data structure of this system consists of multiple tables to realize a generational inheritance type monitoring service. The user attribute management table manages user ID, age information, operation mode, and registration date and time, and records the transition history from child mode to elderly mode for the same user. The operation pattern recording table manages user ID, operation type, operation frequency, recording period, and mode, and stores basic data to utilize operation characteristics from the child mode era for UI optimization in elderly mode. The monitoring information management table manages information ID, user ID, location information, detection status, and notification level, and is used to determine step-by-step notification processing. The UI display settings table manages user ID, font size, and color settings, and realizes personalized display that reflects the setting history from the child mode era. Through the cooperation of these tables, it becomes possible to inherit operation patterns across generations.

[0083] The fall detection system implements a three-stage high-precision algorithm to prevent serious injuries caused by falls among the elderly.

[0084] As shown in Figure 12, the generation succession process sequence is executed between three parties: server 20, dedicated monitoring terminal 10, and communication terminal 30. First, the dedicated monitoring terminal 10 sends an age information update notification (T1) to server 20. Server 20 performs internal processing of generation succession determination (T2), operation pattern acquisition (T3), and UI setting determination (T4). Next, server 20 sends a mode switching instruction (T5) and display setting application (T6) to the dedicated monitoring terminal 10. The dedicated monitoring terminal 10 performs screen display control (T7) and sends a switching completion notification (T8) back to server 20. Server 20 sends monitoring information to communication terminal 30 (T9), and communication terminal 30 performs monitoring display update (T10). After that, the dedicated monitoring terminal 10 notifies server 20 to start new pattern recording (T11), and server 20 starts continuous learning processing (T12). This sequence enables a smooth generational transition from child mode to elderly mode.

[0085] In the first stage of acceleration threshold determination, data from a 3-axis accelerometer is acquired with a sampling frequency of 100Hz, a measurement range of ±16G, and a resolution of 14 bits. The combined acceleration √(ax^2+ay^2+az^2) is calculated to detect an acceleration exceeding 3G. A 0.5-second moving average filter is applied to remove noise, and a fall candidate event is detected based on a duration of 0.1 seconds or more.

[0086] In the second stage of motion pattern analysis, acceleration data for 60 seconds after detecting a potential fall is analyzed in detail. This utilizes the fact that in a normal fall, the movement stops after impact and the standard deviation of acceleration is less than 0.2G. The presence or absence of an getting-up movement (large body movement within 1 minute) is also detected, and if no getting-up movement is detected, a fall is confirmed. The posture estimation algorithm (estimated from gyro sensor data to determine the body axis direction) is used to determine the direction of the fall and the final posture.

[0087] In the third stage of confirmation processing, after a fall is confirmed, the voice guidance "A fall has been detected. Are you okay?" is repeated three times at maximum volume (85dB). During the 30-second waiting period for a response, the response detection process (screen touch, voice input "I'm okay", or shaking the device) is performed. Voice recognition recognizes keywords such as "yes," "I'm okay," and "help." If the response is positive, the process ends as a false positive; if the response is negative or there is no response, an emergency is detected and an automatic notification is sent to family or a medical institution.

[0088] To prevent false positives, the system considers the continuity of actions (excluding continuous actions lasting 3 seconds or more), periodicity (excluding regular actions), and ambient noise (excluding sounds of sitting down on a bed or chair) to distinguish between normal actions and everyday movements (squatting, sitting, lying down). Furthermore, threshold adjustments are performed to adapt to individual movement characteristics through learning each user's behavioral patterns.

[0089] As shown in Figures 9 and 10, the UI adaptive control system performs optimization by utilizing the individual's past operating characteristics during generational transition from child mode 90A to elderly mode 90B.

[0090] Figure 9 shows an example of the screen layout for Child Mode 90A. The design guidelines for Child Mode 90A emphasize visual appeal (bright primary color scheme, RGB values ​​in the range of R=200-255, G=150-255, B=100-255), ease of operation (font size 14pt approximately 4.9mm, icon size 48dp approximately 12mm), and safety (prevention of accidental operation, restriction of access to inappropriate content). As shown in Figure 9, the top of the screen consists of a current location display area 91 (map icon + address display), the center consists of a group of main operation buttons 92 (large buttons for "Go Home," "Help," and "Contact Family" arranged vertically), and the bottom consists of a status display area 93 (battery level, communication status, and time arranged horizontally). The button design uses rounded rectangles (radius 8px) to create a friendly feel, and when touched, an animation effect (0.2 seconds of shrinking and restoring) is displayed as if the button is being pressed. The font used is a universal design textbook font, achieving both readability and approachability. The background color is based on a light blue (#E8F4F8), a color scheme designed to give children a sense of security.

[0091] Figure 10 shows an example screen configuration for the Elderly Mode 90B. The design guidelines for Elderly Mode 90B emphasize visibility (calm color tone with RGB values ​​of R=50-150, G=50-150, B=50-150, character size 18-24pt, approximately 6.3-8.5mm), operability (1.5 times the line spacing, button size 64dp or larger), and accessibility (contrast ratio of 4.5:1 or higher, WCAG 2.1 AA compliant, and enhanced voice guidance). As shown in Figure 10, the screen consists of a large time display 94 (24pt font, clearly showing year, month, day, day of the week, hour, and minute) at the top, a current location display 95 (address + distance from nearest station) below that, a vertical arrangement of main function buttons 96 in the center ("Emergency Contact," "Contact Family," "Notify Current Location," each 64dp x 48dp in size), and a detailed status display 97 at the bottom (battery level displayed as a percentage and icon, communication status displayed as an antenna mark, steps and health status displayed numerically). The button design is a rectangle with rounded corners (radius 4px) prioritizing ease of operation, and each button has both an icon and text indicating its function. A universal design font is used to maximize visibility. The background color is based on eye-friendly ivory (#F8F8F0), and the color scheme is designed to reduce eye strain even when looking at it for a long time. In addition, a spacing of 12dp or more is ensured between each button to prevent accidental touches.

[0092] As is clear from the comparison of Figures 9 and 10, in the personalized learning process, k-means clustering (k=5, categorization of operation patterns) and principal component analysis (dimensionality reduction, components with a contribution rate of 90% or higher are selected) are performed as unsupervised learning. In the supervised learning process, classification learning (support vector machine, kernel: RBF, regularization parameter C=1.0) is performed using operation success / failure as the training signal, and regression learning (random forest regression, 50 decision trees) is performed with operation time as the target variable.

[0093] In the conversion from the child mode screen 90A in Figure 9 to the elderly mode screen 90B in Figure 10, the operational characteristics extracted as features include handedness determination (analysis of bias in tap positions), grip estimation (pressure distribution pattern), fingertip size estimation (statistical analysis of touch area), gaze pattern estimation (distribution of screen dwell time), and attention concentration characteristics (times when operational errors frequently occur).

[0094] The detailed calculation formulas for the UI generation rules in the elderly mode screen 90B in Figure 10 are as follows: Character size coefficient = 1.0 + (100 - operation accuracy score) × 0.01, Button size coefficient = 1.0 + (100 - operation speed score) × 0.008, Line spacing coefficient = 1.0 + (100 - visibility score) × 0.006, Voice guidance frequency = (100 - learning adaptation score) × 0.02. The parameter adjustment ranges are: Character size -20% to +100%, Button size -10% to +80%, Line spacing -20% to +100%, Voice guidance frequency 0 to 2.0 times / operation.

[0095] The architecture of this system adopts a Kubernetes microservice configuration based on a cloud-native design, enabling independent scaling for each service, rolling updates, and fault isolation.

[0096] The main service configuration consists of API Gateway (request routing, rate limiting, authentication), authentication services (OAuth 2.0, SAML 2.0, multi-factor authentication), location information processing services (GPS data normalization, geofencing determination, anomaly detection), machine learning services (operation pattern analysis, personalization parameter generation), notification services (push notifications, SMS, email, voice calls), and database services (user data, operation history, analysis results).

[0097] Load balancing is implemented using L7 load balancing via Application Load Balancer (ALB), with dynamic selection of routing methods including round-robin, weighted routing (server performance ratio), minimum connection priority, and response time priority. Health checks are performed every 30 seconds using the HTTP GET / health endpoint; disconnection occurs after 3 consecutive failures, and reactivation occurs after 2 consecutive successes.

[0098] Auto-scaling is configured to scale out if any of the following conditions persist for 5 minutes: CPU usage exceeding 70%, memory usage exceeding 80%, response time exceeding 500ms, or number of concurrent connections exceeding 1000. The scale-in condition is when all metrics remain below 50% of the threshold for 15 minutes. The maximum scale configuration consists of 100 API Gateway Pods, 50 machine learning service Pods, and 10 read-only database replicas.

[0099] The database will employ a polyglot configuration, utilizing PostgreSQL 14 (structured data such as user information and configuration information), MongoDB 5.0 (time-series data such as location history and operation history), and Redis 6.2 (session information and cached data). PostgreSQL will be configured as a master-slave setup with one master and three read replicas, while MongoDB will be configured as a replica set of three instances.

[0100] Our data retention policy stipulates that detailed data (operation history, location history) will be stored for 3 years, aggregated data (daily and monthly statistics) for 10 years, and statistical data (anonymized analysis results) will be permanently stored. Personally identifiable information will be automatically deleted one year after discontinuation of use. Backups will consist of daily full backups and hourly incremental backups, guaranteeing an RTO (Recovery Time Objective) of 4 hours and an RPO (Recovery Point Objective) of 1 hour.

[0101] By complying with FHIR R4, the system supports international standards for medical information sharing. Implementation resources include Patient (patient information), Observation (observation information: location, vital signs, behavioral patterns), Location (location information), Device (device information), and Diagnostic Report (diagnostic report: fall detection results, etc.). RESTful API provision enables standard integration with electronic medical record systems, community comprehensive support center systems, and long-term care record systems.

[0102] The disaster emergency response flow provides comprehensive support functions in the event of a natural disaster. Meanwhile, the emergency identity verification function implements a three-stage information disclosure control system that balances privacy protection with emergency response.

[0103] The identity notification function can be activated in three ways: (1) manually by the user (press and hold the emergency button for 3 seconds), (2) automatically by the user (no response for 60 seconds after fall detection, or detection of unconscious state), and (3) by a third party (reading a QR code (registered trademark), NFC communication, or Bluetooth connection). Each trigger has a different level of access permission.

[0104] In the first phase, only emergency response instructions will be displayed, and personal information will be kept completely confidential. The displayed content will include messages such as, "There may be an emergency situation for the owner of this device," "Please call an ambulance (dial 119)," and "If the person is unconscious, please show this device screen to the paramedics." Voice guidance will also be provided simultaneously to ensure accessibility for visually impaired individuals.

[0105] In the second phase, limited personal information will be disclosed. Disclosure conditions will be keyword recognition via voice input for terms such as "paramedic," "police officer," and "firefighter," or identity verification through communication with an NFC ID card embedded in the paramedic or police officer's employee ID card. The disclosed information will be limited to name, age, blood type, address, and two emergency contacts (names and phone numbers only).

[0106] In the third stage, detailed medical information will be disclosed. The conditions for disclosure are NFC authentication of medical professional qualification certificates such as a medical license or nursing license, or authenticated access from a terminal at a medical institution. The disclosed information will include detailed medical information (medical history, names, usage, and dosage of currently taken medications, allergy information, contact information for primary care physician, long-term care insurance card number, and guarantor information).

[0107] As a security measure, all access logs will be encrypted and recorded. The log contents will include the date and time of access, GPS coordinates, access method (manual / QR code / NFC / voice), access terminal information (smartphone model, OS, MAC address), and, if possible, a photo of the operator taken with the front camera. To prevent misuse, continuous access restrictions (up to 3 times in 5 minutes), abnormal time zone access detection (2-5 AM), and suspicious location access detection using GPS will be implemented.

[0108] As shown in Figure 13, screen transitions consist of a transition during generational succession from the child mode screen to the elderly mode screen, and transitions to the settings screen in each mode. In the child mode screen, the SOS button, settings button, and family button are placed, and the font size is set to small. The generational succession process during age update transitions to the elderly mode screen. In the elderly mode screen, based on the operation patterns from the child mode era, frequently used functions are enlarged and placed higher. For example, if the settings button was used frequently, it will be enlarged and placed higher; if the family contact function was used moderately, it will be placed in the center and highlighted; and if the history function was used infrequently, it will be smaller and placed lower. In the settings screen, items such as font size adjustment, color settings, and notification level are displayed, and frequently used items are displayed at the top, reflecting past setting history. The care manager collaboration system enables standard information sharing based on the long-term care insurance system and contributes to the advancement of scientifically-based care.

[0109] The collaboration protocol adopts the API design compliant with the Ministry of Health, Labour and Welfare's "Standard Specifications for Long-Term Care Information Collaboration Systems" and HL7 FHIR R4. Communication security will be implemented using VPN (Virtual Private Network) encrypted communication, multi-factor authentication (ID / password + SMS authentication + digital certificate), and qualification verification using the digital certificate of the care support specialist certificate.

[0110] The linked data items include information on activities of daily living (ADL), such as mobility (quantitative measurement of walking distance, speed, and frequency), outing patterns (automatic recording of outing time, destination, and companions), and lifestyle rhythms (estimation of wake-up, bedtime, and meal times). Information on instrumental activities of daily living (IADL) is also automatically analyzed, such as shopping behavior (frequency, location, and duration), visits to medical institutions (frequency, location, and estimated medical department), and participation in social activities (frequency, location, and estimated activity content).

[0111] The care plan creation support function automatically performs quantitative evaluations of ADL / IADL abilities (automatic calculation of scores based on the Barthel Index and Lawton IADL Scale) from accumulated objective behavioral data, estimates cognitive function levels (calculated from the complexity of outing patterns, ability to move to new places, and time management ability), and evaluates social participation (calculated from the frequency of outings, opportunities for contact with people, and diversity of activity locations), thereby supporting the care manager's professional judgment with objective data.

[0112] The voice interface function of the dedicated monitoring terminal 10 implements a specialized design that takes into account the hearing and speech characteristics of the elderly.

[0113] The dedicated monitoring terminal 10 is further equipped with a voice guidance unit 110. The voice guidance unit 110 is a functional unit that, in elderly monitoring mode, outputs voice guidance such as operation guidance linked to screen operation, periodic announcements of time and weather, and medication reminders from the voice output unit 109. The voice guidance unit 110 includes a text-to-speech (TTS) engine that converts text to speech and provides personalized settings such as speech speed adjustment (0.8-1.5x speed), pitch adjustment (±20%), and dialect support (8 dialects).

[0114] The audio input section 108 is equipped with two MEMS microphones (frequency response 100Hz-8kHz, S / N ratio of 64dB or higher) and uses beamforming technology (delay sum algorithm) to achieve directional control and ambient noise suppression (20dB or higher). The audio processing engine implements an embedded application optimized ASR (Automatic Speech Recognition) engine, guaranteeing recognition accuracy of 90% or higher.

[0115] The speech recognition process consists of a step-by-step process involving preprocessing (noise reduction: spectral subtraction, volume normalization: RMS normalization, frequency filtering: 200Hz-4kHz bandpass), feature extraction (MFCC: Mel-Frequency Cepstral Coefficients 12-dimensional + Δ + ΔΔ), acoustic model (DNN-HMM: Deep Neural Network - Hidden Markov Model, 1024 hidden states), and language model (tri-gram, 1000-word vocabulary).

[0116] The recognition vocabulary will be specialized in emergency terms ("Help," "It hurts," "I'm in pain," "It's okay," "Yes," "No"), everyday terms ("Thank you," "Excuse me," "I don't understand," "One more time"), and operational terms ("Stop," "Bigger," "Softer," "Brighter," "Darker"), and will employ a model that has learned the speech characteristics of the elderly (indistinct endings, volume fluctuations, and breathing).

[0117] The Text-to-Speech (TTS) function employs a design that takes into account the hearing characteristics of the elderly (decreased hearing in the high-frequency range and reduced speech intelligibility). The audio output unit 109 is equipped with a full-range speaker (frequency response 200Hz-8kHz, maximum output 5W, maximum sound pressure level 85dB) to achieve clear audio output even in outdoor environments. The speech synthesis engine implements high-quality synthesized speech based on WaveNet technology and performs sound quality adjustments that prioritize naturalness, clarity, and familiarity.

[0118] Personalized voice settings include speech speed adjustment (0.8-1.5x speed, default 1.0x speed), pitch adjustment (±20%, adjusted according to individual listening comfort), automatic volume adjustment (environmental noise level +10dB), and dialect support (8 dialects including standard Japanese, Kansai dialect, Tohoku dialect, Kyushu dialect, and Okinawa dialect). It also implements a "family voice mode" that allows users to pre-record family voices to provide reassurance in emergencies. Voice data is encrypted and stored in the device's Secure Element and is not transmitted externally.

[0119] The security design of this system implements multi-layered defense in accordance with the Personal Information Protection Act, the Guidelines for the Secure Management of Medical Information Systems (Ministry of Health, Labour and Welfare), the IoT Security Guidelines (Ministry of Internal Affairs and Communications and Ministry of Economy, Trade and Industry), and the NIST Cybersecurity Framework.

[0120] Communication encryption will be based on TLS 1.3 (Transport Layer Security), with AES-256-GCM (Advanced Encryption Standard) as the encryption method, ECDH (Elliptic Curve Diffie-Hellman) P-384 for key exchange, and HMAC-SHA-256 for message authentication. Certificates will use RSA-4096 or ECDSA P-384 public key cryptography, and their validity period will be within one year.

[0121] Terminal authentication implements a combination of device-specific certificates (X.509 v3) and Trusted Platform Module (TPM 2.0). Each terminal has a unique device certificate embedded during manufacturing, and mutual authentication (bidirectional verification of device certificates and server certificates) is performed during the initial connection to the server. Authentication information is encrypted and stored in the terminal's Secure Element (Common Criteria EAL5+ authentication) to ensure tamper resistance against physical attacks.

[0122] For data protection, all stored data is encrypted using AES-256-CBC, and encryption keys are managed by a dedicated Key Management System (KMS) using a Hardware Security Module (HSM). Personally identifiable information (PII) is anonymized (k-anonymization, k≧5) and differential privacy (ε=1.0) is applied to minimize the risk of personal identification during statistical and machine learning processing. Database access is controlled by Role-Based Access Control (RBAC), and access privileges are granted based on the Principle of Least Privilege.

[0123] As part of incident response, a SIEM (Security Information and Event Management) system will be implemented to establish a 24 / 7 / 365 Security Operation Center (SOC) monitoring system. The system will monitor for brute-force attacks (more than 10 authentication failures per minute), SQL injection (abnormal SQL statement patterns), XSS (Cross-Site Scripting, script execution attempts), DDoS attacks (abnormal traffic exceeding 1000 requests per second), and internal unauthorized access (access outside of working hours or from abnormal locations). If thresholds are exceeded, the system will automatically block connections and notify relevant parties.

[0124] This section details the emergency response functions in the event of a natural disaster (earthquake, tsunami, typhoon, heavy rain, volcanic eruption). Disaster detection is achieved through API integration with the Japan Meteorological Agency's Earthquake Early Warning System, J-Alert (National Instant Warning System), and local government disaster prevention information systems. After receiving disaster information, the system automatically switches to "disaster mode," changing from normal monitoring functions to survival confirmation and evacuation support functions.

[0125] As a change in operation during disaster mode, the frequency of location information transmission will be changed from the normal 5-minute interval to a 30-second interval to enable real-time safety confirmation. If the battery level falls below 50%, power saving settings (screen brightness 10%, reduced GPS positioning accuracy, communication frequency 1-minute interval) will be automatically applied to ensure operation continues for as long as possible. In the event of a communication infrastructure failure, an automatic switching function to satellite communication (Iridium, Inmarsat) or LPWAN (LoRa, Sigfox) will be implemented.

[0126] As an evacuation support function, it provides voice and screen guidance for the optimal route from your current location to the nearest designated evacuation shelter or designated emergency evacuation site. The route calculation prioritizes safety during disasters (avoiding liquefaction-prone areas, landslide-prone areas, and tsunami inundation prediction areas) rather than the shortest route under normal circumstances. Evacuation shelter information is linked in real time with open data from local governments (shelter location, capacity, opening status, and available facilities), and the information is updated every 5 minutes.

[0127] In the event of a disaster posing an immediate threat to life, such as a tsunami or landslide, emergency voice warnings such as "Evacuate to higher ground immediately" and "Stay away from mountains" will be continuously output at maximum volume (85dB). The screen will display large arrows indicating the direction of evacuation and distance information, conveying emergency information through both visual and auditory means.

[0128] To ensure the safety of family members, the system will temporarily lift the usual location sharing restrictions (only general place names within a 500m radius) during disasters, sharing the exact location, movement history, and last confirmed time with all family members. Furthermore, to expedite identity verification at evacuation centers, it will implement a QR code (registered trademark) display function (including name, age, emergency contact information, and medical information) and an automatic registration function to the evacuee list (linked with local government systems). After the disaster subsides (24 hours after the weather warning is lifted), the system will automatically revert to normal mode and provide trauma care support information (contact information for specialized organizations, psychological first aid information).

[0129] The disaster emergency response flow provides comprehensive support from disaster detection to evacuation completion. The flow consists of six stages: disaster information reception, disaster mode switching, safety confirmation, evacuation route guidance, family notification, and evacuation center coordination. In disaster information reception, warnings from the Japan Meteorological Agency, J-Alert, and local government disaster prevention systems are automatically received. In disaster mode switching, the location information transmission frequency is changed from every 5 minutes to every 30 seconds, and power saving settings are applied. In safety confirmation, the user's status is confirmed through voice guidance and vibration. In evacuation route guidance, a safe route from the current location to the nearest evacuation center is guided by voice and on screen. In family notification, accurate location information and safety status are shared with all family members. In evacuation center coordination, identity verification is performed using a QR code (registered trademark) and automatic registration to the evacuee list is performed.

[0130] This section details our multilingual support measures, with a view to global expansion. In the first phase, we will support five languages: Japanese, English, Chinese (simplified and traditional), and Korean. In the second phase, we will add four more languages: Thai, Vietnamese, Spanish, and Portuguese, bringing the total to nine languages.

[0131] The internationalization (i18n) design implements externalization of string resources (managed in property files), regional support for date, time, number, and currency display (locale settings: ja_JP, en_US, zh_CN, etc.), support for character rendering direction (left-to-right: LTR, right-to-left: RTL), and full support for multibyte characters (unified UTF-8).

[0132] Localization (l10n) involves adapting to each country's legal systems (personal data protection laws, medical laws, elderly welfare laws), cultural considerations (meaning of colors, taboo colors, avoidance of physical contact and personal space, religious considerations), and communication standards (frequency band, output restrictions, authentication requirements). For example, in the Chinese market, it is necessary to comply with government approval for location-based services, domestic data storage requirements, and export control regulations for encryption technology.

[0133] To support multiple languages ​​in the voice interface, dedicated speech recognition and speech synthesis engines will be implemented for each language. The speech recognition vocabulary will be optimized for emergency medical terminology in each country (essential emergency vocabulary such as "Help," "pain," and "chest pain"), everyday conversational expressions, and proper nouns (place names, personal names, facility names). For speech synthesis, sound quality adjustments will be made considering standard pronunciation in each country, regional dialects, and speaker characteristics (gender, age, speaking speed) that are familiar to the elderly.

[0134] To comply with international medical and care standards, the system will implement integration with ICD-11 (International Classification of Diseases, 11th Revision), ICF (International Classification of Functioning, Disability and Health), and SNOMED CT (Systematic Clinical Medicine Terminology). It will also establish protocols for integration with national emergency call systems and information sharing with local medical institutions, care facilities, and administrative agencies. To address time zone differences, it will implement unified management using UTC standard time and a function to switch between displaying information in different regional time zones.

[0135] The multilingual UI switching system provides the functionality to dynamically change the screen display according to the user's language settings. The system consists of a language detection unit, a resource management unit, a layout control unit, and a character rendering engine. The language detection unit identifies the user's language by integrating the terminal's language settings, GPS-based region determination, and language recognition via voice input. The resource management unit efficiently manages string resources, font data, and image resources for nine languages ​​and performs dynamic loading as needed. The layout control unit optimizes the screen layout according to the number of characters, rendering direction (LTR / RTL), and cultural considerations for each language. The character rendering engine, with UTK-8 support, achieves accurate display of multilingual characters and automatically adjusts the font size and line spacing to suit each language. When switching languages, personal settings accumulated through operation pattern learning are also inherited as language-specific settings, maintaining consistent ease of use.

[0136] (Example 1: Cloud / on-premises mixed configuration) As the first variation of this system, we will describe a hybrid cloud configuration according to the data sensitivity level. A segregated architecture is adopted in which highly confidential personal information (secrecy level 3: name, address, medical history, family information) is stored in an on-premises private cloud, while anonymized data for statistical processing and machine learning (secrecy level 1: statistical data, trained models) is processed in a public cloud (cloud service).

[0137] The data classification policy categorizes data into three levels: secrecy level 3 (personally identifiable information, requires special management), secrecy level 2 (behavioral history / medical information, requires management), and secrecy level 1 (statistical data / trained models, general management). Each level has different storage locations (on-premise, private cloud, public cloud), encryption methods (AES-256, AES-192, AES-128), and access permissions (executive level, administrator level, general level). Data transfer between on-premise and cloud locations is conducted via dedicated lines (closed network, guaranteed bandwidth) or highly encrypted VPNs (AES-256 + RSA-4096, Perfect Forward Secrecy compatible) to prevent interception and tampering by third parties.

[0138] (Modification 2: Configuration utilizing edge computing) As a second variation of this system, we will describe a distributed processing configuration using edge computing technology that leverages 5G / Beyond 5G infrastructure. By using edge servers (MEC: Multi-access Edge Computing) located in each region, real-time processing such as location information processing, anomaly detection, speech recognition, and image analysis is performed near the user (at the edge), significantly reducing response delay (average 200ms → less than 50ms).

[0139] The optimal placement of edge servers takes into account population density, geographical conditions, communication infrastructure, and disaster risk. In urban areas, one server should be placed within a 5km radius (latency within 5ms); in suburban areas, one server within a 20km radius (latency within 20ms); and in mountainous areas, one server within a 50km radius (latency within 50ms, with satellite communication backup). Dynamic load balancing and failover are achieved between edge servers using SD-WAN (Software-Defined Wide Area Network), eliminating single points of failure (SPOF).

[0140] (Variation 3: 5G communication support and high-speed data transfer) As a third variation of this system, we will describe ultra-high-speed, high-capacity, and ultra-low-latency communication compatible with 5G (fifth-generation mobile communication system) and Beyond 5G. It supports the Sub-6GHz band (3.7GHz, 4.5GHz) and millimeter-wave band (28GHz, 39GHz) of 5G NR (New Radio), achieving a theoretical maximum communication speed of 20Gbps (downlink), an effective speed of 1-5Gbps, and a latency of 1ms or less.

[0141] Advanced functions utilizing 5G include 4K video monitoring (checking the situation in case of a fall, remote diagnosis by a doctor), virtual meetings with family using VR / AR technology (360-degree camera and VR headset integration), advanced conversational functions by an AI concierge (integration with a large-scale language model), and immersive family communication through hologram communication. In addition, by using 5G base station location information (CID: Cell ID, high-precision positioning support) in conjunction with positioning, the accuracy of location estimation outside of GPS coverage (indoors, underground, tunnels, etc.) will be significantly improved from the conventional 100m to less than 10m.

[0142] (Modification 4: Vital Sign Monitoring Function) As a fourth variation of this system, we will describe the integration of vital sign monitoring functionality. The dedicated monitoring terminal is equipped with a PPG heart rate sensor (photoplethysmography, green LED + photoelectric sensor), an SpO2 blood oxygen saturation sensor (infrared LED + red LED), a skin temperature sensor (thermistor, accuracy ±0.1℃), and a sweat sensor (electrical conductivity measurement), enabling early detection of physiological abnormalities.

[0143] Heart rate variability (HRV) analysis is used to assess the state of the autonomic nervous system (sympathetic / parasympathetic balance), estimate stress levels (HRV frequency analysis), and evaluate sleep quality (REM / non-REM sleep determination). The abnormal value detection algorithm automatically detects abnormal heart rates (tachycardia >100 bpm, bradycardia <50 bpm, arrhythmia detection), decreased blood oxygen saturation (SpO2 <95%, early detection of respiratory diseases such as COPD), and abnormal body temperature (fever >37.5°C, hypothermia <35.0°C) by comparing them with individual reference values ​​that take into account age, sex, and medical history. This enables early medical intervention in conjunction with a step-by-step notification system.

[0144] (Variation 5: Medication management and reminder function) As a fifth variation of this system, we will describe a medication management support function aimed at improving medication adherence among the elderly. Through collaboration with pharmacists and physicians, the system automatically retrieves prescription drug information (drug name, ingredient name, dosage and administration, time of administration, duration of administration, side effect information, and interaction information) from the electronic prescription system and performs individualized medication schedule management.

[0145] The medication reminder function provides three-stage notifications: 30 minutes before the set time (preparation notification), at the set time (medication notification), and 30 minutes after the set time (confirmation notification). Notification methods include a combination of voice guidance (reading out the drug name and dosage), screen display (drug photo and administration instructions), and vibration alerts (gradual intensity adjustment). Medication confirmation objectively records medication adherence through voice recognition (keywords such as "I took my medicine" and "Medication completed"), camera recognition (confirmation of type and quantity by photographing the medication, AI image analysis), and IoT pill case integration (pill case with open / close sensor and weight sensor).

[0146] As a step-by-step response to missed medication, the system escalates in stages: 1st time (+1 hour): voice guidance is re-executed; 2nd time (+3 hours): family is notified with the message "Medication time has passed"; 3rd time (+6 hours): medical institution / pharmacy is notified with the message "Medication not confirmed." This prevents serious symptom worsening due to missed medication. In addition, to prevent duplicate medication, a warning message "Already taken. Risk of duplicate medication" is displayed when additional medication is to be taken for a drug already taken.

[0147] (Variation 6: Dementia progression assessment function) As a sixth variation of this system, we will describe a function that enables early detection and quantitative evaluation of changes in cognitive function from subtle changes in daily behavior patterns. Through long-term analysis of location information, movement patterns, and operation history, the progression from the prodromal stage of dementia (MCI: Mild Cognitive Impairment) to mild, moderate, and severe dementia can be objectively evaluated.

[0148] The evaluation indicators are quantified across four domains: orientation (time: disruption of scheduled behavior, place: getting lost or lost, people: confusion in recognizing family members), attention function (concentration: increased frequency of operational errors, persistence: frequency of work interruptions), executive function (planning: extended time spent preparing to go out, executive ability: failure to complete intended actions), and memory function (short-term memory: forgetting recent actions, long-term memory: getting lost in familiar places).

[0149] The machine learning-based progression assessment uses behavioral data from the past 24 months as training data to calculate a monthly cognitive function score (0-100 points, 100 being normal). If the score decline rate is 5 points or more per month, it is classified as "caution required," and if it is 10 points or more, it is classified as "further examination required," and notifications are sent to the family and the attending physician. The evaluation results employ a scoring system supervised by dementia specialists, ensuring objective evaluation based on medical evidence. Regular reports to family and medical professionals (monthly summaries and annual detailed analyses) support the initiation of medical consultations and care services at the appropriate time.

[0150] (Variation 7: Pet monitoring support) As a seventh variation of this system, we will describe a mutual monitoring system for people and their pets (dogs and cats) by extending the monitoring function for elderly people's pets. A small GPS terminal for pets (weighing less than 30g, waterproof IPX8, shock-resistant design, 7-day battery life) will monitor the pet's location, activity level, and health status, achieving a two-way monitoring effect between the elderly and their pets.

[0151] Pet behavior analysis records changes in activity level (steps, distance traveled, exercise intensity, measured by a 3-axis accelerometer), rest patterns (sleep duration, location, detected by a motion sensor), elimination behavior (time spent in specific locations, detected by an acoustic sensor), and eating behavior (time spent near food bowls, estimated intake using a weight sensor).

[0152] Pet abnormality detection includes monitoring for a sudden decrease in activity level (less than 50% compared to the previous week, possibility of illness or injury), movement to unfamiliar places (deviation from the usual range of activity, possibility of getting lost or escaping), prolonged dwelling in the same location (more than 2 hours of inactivity, possibility of injury or poor health), and prolonged periods of being more than a certain distance away from the elderly (more than 500m, more than 24 hours, abnormality in the mutual monitoring relationship). Pet abnormalities can also be used as a precursor indicator of changes in the health status of the elderly (progression of dementia, depressive symptoms, decline in physical function), improving the overall monitoring effect for both people and pets.

[0153] <Summary> [General tasks] One of the purposes of this disclosure is to realize a technology that can flexibly respond to changes in the life stage of the person being monitored and provide continuous and efficient monitoring services over the long term.

[0154] [Issues related to Appendix 1] One of the purposes of this disclosure is to enable monitoring from childhood to old age using a single system, without requiring separate systems for each generation, while ensuring continuity of operation. [Note 1] It is equipped with a server, a dedicated monitoring terminal, and a communication terminal. The aforementioned server, A switching control unit that switches the operating mode of the dedicated monitoring terminal from child monitoring mode to elderly monitoring mode based on attribute information including the user's age information, A UI management unit acquires the operation pattern in the child monitoring mode and determines the display settings (font size, color settings) based on the operation pattern in the elderly monitoring mode. Equipped with, The aforementioned monitoring terminal is GPS function, A display control unit that performs screen display control based on the display settings received from the server, Equipped with, The aforementioned communication terminal is The system includes a monitoring display unit that receives and displays monitoring information from the server. A monitoring system. [Effects of Appendix 1] The configuration described in Appendix 1 enables long-term monitoring services through generational inheritance using a single system, avoiding redundant device investments. Furthermore, by utilizing an individual's past operation patterns, the burden of learning how to operate the system in old age is significantly reduced, allowing users to maintain a familiar feel for the system. This configuration solves a common problem of flexibly responding to changes in the life stages of those being monitored and enabling continuous and efficient monitoring services over the long term.

[0155] [Issues related to Appendix 2] One of the purposes of this disclosure is to ensure the continuity of the operating environment when the person providing the monitoring becomes an elderly person and is subsequently the person being monitored. [Note 2] The monitoring system described in Appendix 1, The UI management unit acquires a display settings history, including the text size setting and icon size setting, on the communication terminal in the child monitoring mode, and determines the text size and icon size of the dedicated monitoring terminal based on the display settings history in the elderly monitoring mode. A monitoring system. [Effects of Appendix 2] As described in Appendix 2, the display settings used by the elderly person as the monitor provider on the communication terminal can be transferred to the dedicated monitoring terminal when they become the person being monitored, significantly reducing the burden of learning new operations.

[0156] [Issues related to Appendix 3] One of the purposes of this disclosure is to realize a notification system that, when an abnormality is detected in a person being monitored, respects the autonomy of the person and gradually expands support as needed. [Note 3] The monitoring system described in Appendix 1, The server further includes a notification control unit that, upon detecting an abnormal signal from the dedicated monitoring terminal, executes a primary notification to the dedicated monitoring terminal, and if a response to the primary notification is not received within a predetermined time, executes a secondary notification to the communication terminal in stages. A monitoring system. [Effects of Appendix 3] The configuration described in Appendix 3 ensures safety while respecting the autonomy of the person being monitored to the fullest extent, and only notifying family members when necessary, thereby eliminating a sense of excessive surveillance.

[0157] [Issues related to Appendix 4] One of the purposes of this disclosure is to provide a means for a third party to quickly verify the identity information of a person being monitored in an emergency situation such as becoming unconscious. [Note 4] The monitoring system described in Appendix 1, The aforementioned monitoring terminal further includes an identity notification unit that outputs identity information (name, emergency contact information) via voice in response to user operation. A monitoring system. [Effects of Appendix 4] The configuration described in Appendix 4 allows a third party to confirm the identity of the person being monitored by voice, even if the person is unconscious, enabling a rapid emergency response.

[0158] [Issues related to Appendix 5] One of the purposes of this disclosure is to enable high-precision detection of falls among the elderly, reduce false positives, and respond quickly. [Note 5] The monitoring system described in Appendix 1, The server further includes a fall detection unit that determines the state of a fall based on detection data from the 3-axis acceleration sensor of the dedicated monitoring terminal. A monitoring system. [Effects of Appendix 5] The configuration described in Appendix 5 enables objective fall detection based on acceleration data, allowing for early intervention to prevent serious injuries caused by falls in the elderly.

[0159] [Issues related to Appendix 6] One of the purposes of this disclosure is to automatically facilitate referral to professional caregivers when family members are unable to provide adequate care. [Note 6] The monitoring system described in Appendix 3, The notification control unit, if no response is received to the secondary notification, executes a tertiary notification to the care manager's terminal. A monitoring system. [Effects of Appendix 6] As described in Appendix 6, even if family members are unable to provide support, automatic referral to care managers with specialized caregiving knowledge is enabled, ensuring appropriate assistance.

[0160] [Issues related to Appendix 7] One of the purposes of this disclosure is to realize location information control that provides detailed location information in emergencies while protecting the privacy of the person being monitored. [Note 7] The monitoring system described in Appendix 1, The server further includes a location information control unit that transmits detailed location information to the communication terminal in emergencies and transmits approximate location information (area information within a 500m radius) under normal circumstances. A monitoring system. [Effects of Appendix 7] The configuration described in Appendix 7 allows for both protection of privacy during normal times and rapid location identification in emergencies, ensuring safety while maintaining the dignity of the person being monitored.

[0161] [Issues related to Appendix 8] One of the purposes of this disclosure is to establish appropriate levels of information sharing for multiple family members, tailored to each individual's relationship with them. [Note 8] The monitoring system described in Appendix 1, The server further comprises an information sharing control unit that provides monitoring information to multiple communication terminals at different information sharing levels based on the family relationship information of each communication terminal. A monitoring system. [Effects of Appendix 8] The structure described in Appendix 8 allows for the provision of information tailored to the relationships and responsibilities of each family member, enabling an appropriate balance between privacy protection and emergency response.

[0162] [Issues related to Appendix 9] One of the purposes of this disclosure is to enable monitoring using a small device that can be carried at all times by the person being monitored. [Note 9] The monitoring system described in Appendix 1, The aforementioned monitoring terminal is a wearable device and is equipped with an emergency notification button and a display unit. A monitoring system. [Effects of Appendix 9] The configuration described in Appendix 9 enables continuous monitoring in a wearable form that can be worn daily by the person being monitored, and allows for immediate notification in emergencies.

[0163] [Issues related to Appendix 10] One of the purposes of this disclosure is to improve usability by supporting elderly users through voice guidance. [Note 10] The monitoring system described in Appendix 1, The aforementioned monitoring terminal further includes a voice guidance unit that outputs operational instructions via voice in the elderly monitoring mode. A monitoring system. [Effects of Appendix 10] The configuration described in Appendix 10 compensates for the decline in visual function among the elderly and supports intuitive operation through voice guidance.

[0164] [Issues related to Appendix 11] One of the purposes of this disclosure is to ensure the security of communications regarding monitoring information. [Note 11] The monitoring system described in Appendix 1, Communication between the server, the dedicated monitoring terminal, and the communication terminal is protected by an encrypted communication protocol. A monitoring system. [Effects of Appendix 11] The configuration described in Appendix 11 prevents the interception and falsification of monitoring information and ensures the protection of personal information.

[0165] [Issues related to Appendix 12] One of the purposes of this disclosure is to build a mutual monitoring network in which elderly people can maintain their social role as monitors while also being monitored themselves. [Note 12] The monitoring system described in Appendix 1, The server further performs dual monitoring control, in which the first user monitors the second user while the first user himself is also monitored by the third user. A monitoring system. [Effects of Appendix 12] The configuration described in Appendix 12 enables the realization of a mutual monitoring network in which elderly people can maintain their social roles while ensuring their own safety by participating in watching over their grandchildren and other relatives.

[0166] [Issues related to Appendix 13] One of the purposes of this disclosure is to enable the long-term accumulation and management of operation patterns and display settings. [Note 13] The monitoring system described in Appendix 1, The server further comprises a database management unit that stores the operation patterns and the display setting history in chronological order. A monitoring system. [Effects of Appendix 13] The configuration described in Appendix 13 enables the accumulation and analysis of operation patterns over long periods, improving the accuracy of UI optimization during generational succession.

[0167] [Issues related to Appendix 14] One of the purposes of this disclosure is to enable automated coordination with public institutions in emergency situations where no response is received through all phased notifications. [Note 14] The monitoring system described in Appendix 6, If no response is received to the third notification, the notification control unit will execute a fourth notification to the local government system (community comprehensive support center, fire station). A monitoring system. [Effects of Appendix 14] As described in Appendix 14, even in emergencies where family members and care managers are unable to respond, rapid rescue can be achieved through automatic notification to public institutions.

[0168] [Issues related to Appendix 15] One of the purposes of this disclosure is to continuously achieve UI optimization that adapts to changes in the physical and cognitive functions of elderly people over time. [Note 15] The monitoring system described in Appendix 1, The UI management unit further includes a learning control unit that analyzes the operation patterns in the elderly monitoring mode using machine learning and continuously optimizes the display settings. A monitoring system. [Effects of Appendix 15] As described in Appendix 15, the UI is automatically optimized to adapt to changes in the physical and cognitive functions of elderly users, maintaining a comfortable operating environment over the long term.

[0169] [Issues related to Appendix 16] One of the purposes of this disclosure is to provide an information processing device that enables a generational succession type monitoring service. [Note 16] It comprises a processor and a memory unit, The aforementioned storage unit is The program is stored, The aforementioned processor, By executing the aforementioned program, Based on attribute information including the user's age, the process switches the operating mode of the dedicated monitoring terminal from child monitoring mode to elderly monitoring mode. The process involves acquiring the operation pattern in the child monitoring mode and determining the display settings in the elderly monitoring mode based on the operation pattern, Execute Information processing device. [Effects of Appendix 16] The configuration described in Appendix 16 enables the server device to realize the core functions of the generational succession type monitoring service.

[0170] [Issues related to Appendix 17] One of the purposes of this disclosure is to provide an information processing method that enables a generational succession type monitoring service. [Note 17] The processor, Based on attribute information including the user's age, the operating mode of the dedicated monitoring terminal is switched from child monitoring mode to elderly monitoring mode. The operation pattern in the aforementioned child monitoring mode is acquired, In the aforementioned elderly monitoring mode, the display settings are determined based on the operation pattern. The display settings are transmitted to the dedicated monitoring terminal. It receives requests for monitoring information from communication terminals and provides the monitoring information. Information processing methods. [Effects of Appendix 17] The configuration described in Appendix 17 allows for the protection of the generational succession type monitoring service as a method invention.

[0171] [Issues related to Appendix 18] One of the purposes of this disclosure is to provide a program that enables generational succession type monitoring functionality in a dedicated monitoring terminal. [Note 18] In the processor, The process of receiving operating mode information of the dedicated monitoring terminal from the server. A process to switch between child monitoring display and elderly monitoring display based on the aforementioned operating mode information. The process of recording the user's operation pattern in the child monitoring display and sending it to the server, In the aforementioned elderly monitoring display, a process of receiving and displaying location information from the dedicated monitoring terminal, To execute program. [Effects of Appendix 18] The configuration described in Appendix 18 enables the implementation of a generational succession type monitoring function as a program on the dedicated monitoring terminal side.

[0172] Furthermore, the elements of each of the above embodiments and modifications can be combined as appropriate. Moreover, this disclosure is not limited to the above embodiments, and various modifications are possible within the scope of the technical concept of this disclosure. [Explanation of Symbols]

[0173] 1…Generational succession type monitoring system, 10…Dedicated monitoring terminal, 20…Server, 30…Communication terminal, N…Network, 21…Switching control unit, 22…Location information management unit, 23…UI management unit, 24…Notification control unit, 25…Database, 26…Communication control unit, 11…Location information processing unit, 12…Sensor information processing unit, 13…Display control unit, 14…Location information transmission unit, 15…Communication control unit, 101…Main processor, 102…Memory / storage, 103…GPS receiver, 104…Communication module, 105…Sensor group, 106…Battery, 107…Display, 108…Voice input unit, 1 09…Audio output unit, 201…CPU, 202…Memory, 203…Storage, 204…Network interface, 205…GPU, 206…Power supply unit, 211…User attribute information management module, 212…Age determination module, 213…Mode switching execution module, 231…Operation history collection module, 232…Pattern analysis engine, 233…Personalized UI generation module, 131…UI rendering engine, 132…Adaptive display module, 133…Touch input processing unit, 141…GNSS receiving module, 142…Position estimation filter, 143…Transmission control module

Claims

1. It is equipped with a server, a dedicated monitoring terminal, and a communication terminal. The aforementioned server, A switching control unit that switches the operating mode of the dedicated monitoring terminal from child monitoring mode to elderly monitoring mode based on attribute information including the user's age information, A UI management unit that acquires the operation pattern in the child monitoring mode and determines the display settings based on the operation pattern in the elderly monitoring mode, The aforementioned monitoring terminal is GPS function, The system comprises a display control unit that performs screen display control based on the display settings received from the server, The aforementioned communication terminal is The system includes a monitoring display unit that receives and displays monitoring information from the server. A monitoring system.

2. A monitoring system according to claim 1, The UI management unit acquires a display setting history, including the text size setting and icon size setting, on the communication terminal in the child monitoring mode, and determines the text size and icon size of the dedicated monitoring terminal based on the display setting history in the elderly monitoring mode. A monitoring system.

3. A monitoring system according to claim 1, The server further includes a notification control unit that, upon detecting an abnormal signal from the dedicated monitoring terminal, executes a primary notification to the dedicated monitoring terminal, and if a response to the primary notification is not received within a predetermined time, executes a secondary notification to the communication terminal in stages. A monitoring system.

4. A monitoring system according to claim 1, The aforementioned monitoring terminal further includes an identity notification unit that outputs identity information (name, emergency contact information) via voice in response to user operation. A monitoring system.

5. A monitoring system according to claim 1, The server further includes a fall detection unit that determines the state of a fall based on detection data from the three-axis acceleration sensor of the dedicated monitoring terminal. A monitoring system.

6. A monitoring system according to claim 3, The notification control unit, if no response is received to the secondary notification, executes a tertiary notification to the care manager's terminal. A monitoring system.

7. A monitoring system according to claim 1, The server further includes a location information control unit that transmits detailed location information to the communication terminal in emergencies and transmits approximate location information (area information within a 500m radius) under normal circumstances. A monitoring system.

8. A monitoring system according to claim 1, The server further comprises an information sharing control unit that provides monitoring information to multiple communication terminals at different information sharing levels based on the family relationship information of each communication terminal. A monitoring system.

9. A monitoring system according to claim 1, The aforementioned monitoring terminal is a wearable device and is equipped with an emergency notification button and a display unit. A monitoring system.

10. A monitoring system according to claim 1, The aforementioned monitoring terminal further includes a voice guidance unit that outputs operational instructions via voice in the elderly monitoring mode. A monitoring system.

11. A monitoring system according to claim 1, Communication between the server, the dedicated monitoring terminal, and the communication terminal is protected by an encrypted communication protocol. A monitoring system.

12. A monitoring system according to claim 1, The server further performs a dual monitoring control in which the first user monitors the second user, while the first user himself is also monitored by the third user. A monitoring system.

13. A monitoring system according to claim 1, The server further comprises a database management unit that stores the operation patterns and the display setting history in chronological order. A monitoring system.

14. A monitoring system according to claim 6, If no response is received to the third notification, the notification control unit will execute a fourth notification to the local government system (community comprehensive support center, fire station). A monitoring system.

15. A monitoring system according to claim 1, The UI management unit further includes a learning control unit that analyzes the operation patterns in the elderly monitoring mode using machine learning and continuously optimizes the display settings. A monitoring system.

16. It comprises a processor and a memory unit, The aforementioned storage unit is The program is stored, The aforementioned processor, By executing the aforementioned program, Based on attribute information including the user's age, the process switches the operating mode of the dedicated monitoring terminal from child monitoring mode to elderly monitoring mode. The process involves acquiring the operation pattern in the child monitoring mode and determining the display settings in the elderly monitoring mode based on the operation pattern, Execute Information processing device.

17. The processor, Based on attribute information including the user's age, the operating mode of the dedicated monitoring terminal is switched from child monitoring mode to elderly monitoring mode. The operation pattern in the aforementioned child monitoring mode is acquired, In the aforementioned elderly monitoring mode, the display settings are determined based on the operation pattern. The display settings are transmitted to the dedicated monitoring terminal. It receives requests for monitoring information from communication terminals and provides the monitoring information. Information processing methods.

18. In the processor, The process of receiving operating mode information of the dedicated monitoring terminal from the server. A process to switch between child monitoring display and elderly monitoring display based on the aforementioned operating mode information. The process of recording the user's operation pattern in the child monitoring display and sending it to the server, In the aforementioned elderly monitoring display, a process of receiving and displaying location information from the dedicated monitoring terminal, To execute program.

Images