An AI interaction and minimalist intelligent accompanying system and method suitable for the elderly

Abstract

This invention provides an age-friendly AI interaction and minimalist intelligent companionship system and method, belonging to the field of AI behavior understanding and safe interaction technology for elderly users. The system includes: an age-friendly interface engine module; an AI intent understanding module; a companion-style voice interaction module supporting dialect recognition, emotional soothing, and unlimited repeated responses; a risk prediction and blocking module, pre-setting 12 high-risk scenarios, implementing graded blocking at low, medium, and high levels with an anti-bypass mechanism; and a family ecosystem linkage module supporting one-click help, remote screen sharing, electronic fences, and emergency medical linkage. The method includes: initializing a minimalist interface, collecting operational behavior and identifying intent, implementing graded blocking or companionship guidance based on intent, linking family and community systems, and continuously self-learning and optimizing. This invention achieves a triple closed loop of "ease of use + safety + companionship," increasing the operation success rate to over 90%, effectively blocking accidental touches and fraud risks, and providing a warm and age-friendly interactive experience.

Description

Technical Field

[0001] This invention belongs to the field of AI behavior understanding and safe interaction technology for elderly users, and relates to an age-friendly AI interaction and minimalist intelligent companionship system and method. Background Technology

[0002] With the accelerating aging of the population, the demand for smart devices among the elderly is increasing. However, the age-friendly modification of existing smart devices still suffers from the following technical shortcomings: First, most existing age-friendly modifications only focus on superficial visual optimizations such as "enlarging fonts and increasing contrast," failing to deeply understand the operational intentions, cognitive habits, and behavioral characteristics of elderly users, and thus unable to fundamentally lower the barrier to entry. When using smart devices, elderly users often experience problems such as accidental touches, operational confusion, and process interruptions due to complex interface layouts, redundant functions, and frequent multi-level navigation, resulting in low success rates and a poor user experience.

[0003] Secondly, scams and malicious applications often use complex operating procedures or hidden, misleading buttons to induce elderly users to click repeatedly and perform high-risk operations. Existing devices lack risk prediction and early intervention mechanisms tailored to the behavioral characteristics of elderly users, easily leading to financial losses.

[0004] Third, existing smart devices have a limited range of interaction modes and lack continuous, warm, and supportive interaction. Elderly users often experience loneliness while using them and are prone to giving up when they encounter operational frustrations, lacking mechanisms for emotional support and patient guidance.

[0005] Fourth, the age-friendly design is not integrated with safety protection. The simplified interface, safety reminders, and risk prevention are disconnected, failing to provide elderly users with comprehensive protection throughout the entire process from start to finish.

[0006] Fifth, the standards for age-friendliness are not uniform across different smart devices, and their operating logic varies significantly. Elderly users face a high learning curve when using different devices, making it difficult to develop a consistent age-friendly experience.

[0007] Sixth, elderly users lack targeted age-friendly solutions for high-frequency scenarios such as health management, government affairs processing, travel navigation, emergency medical care, and remote assistance, and the interaction needs of special elderly groups such as those with hearing impairments and visual impairments are not fully met.

[0008] Therefore, there is an urgent need in this field for an age-friendly intelligent interactive system and method that can simultaneously solve the problems of ease of use, security, and companionship. Summary of the Invention

[0009] In view of this, the purpose of this invention is to provide an age-friendly AI interaction and minimalist intelligent companionship system and method to solve the technical problems in the prior art, such as age-friendly modifications only remaining at the visual surface, lacking intent understanding and risk prevention, having a single interaction mode and lacking companionship, and the separation between safety and ease of use.

[0010] To achieve the above objectives, the present invention provides the following technical solution: An age-friendly AI interaction and minimalist intelligent companion system, comprising: The age-friendly interface engine module adopts a minimalist layout of "one core function per screen," with menu levels ≤ 2. The size of regular function buttons is ≥ 60dp×60dp, the size of high-frequency function buttons is ≥ 80dp×80dp, the button spacing is ≥ 15dp, the font size is ≥ 24sp and supports adjustment from 24sp to 48sp, the color contrast is ≥ 4.5:1, and it has built-in high-frequency scene entry points for health management, government affairs processing, travel navigation, emergency assistance, and video calls. There are no ad pushes or forced update prompts, and the interface loading latency is < 300ms. The AI ​​intent understanding module is based on a lightweight LightGBM model with less than 2M parameters to recognize operation intent. It collects multi-dimensional operation behavior features, including click interval, click position deviation, and page dwell time. It can recognize 6 types of intents, including normal operation, accidental touch, hesitant operation, repeated operation, induced operation, and panic operation. The recognition accuracy is ≥95% and the false judgment rate is <3%. It also supports intelligent diagnosis and step-by-step guidance for operation failure. The companion-style voice interaction module supports full-process voice-first interaction, including voice wake-up, voice commands, and voice feedback, with a wake-up success rate of ≥92%. It provides at least 3 elderly-friendly voice tones, and the speech rate can be adjusted from 40 words / minute to 80 words / minute. It supports at least 6 dialects with an accuracy rate of ≥88%, and has functions such as repeated responses, emotion recognition based on voice tone, and soothing. The emotion recognition accuracy rate is ≥85%. The risk prediction and blocking module presets 12 high-risk operation scenarios and performs tiered blocking based on the recognition results of the AI ​​intent understanding module. Low-risk scenarios execute a lightweight confirmation pop-up, medium-risk scenarios initiate a 10-second delay confirmation, and high-risk scenarios directly block the operation. The prediction latency is <50ms, the blocking response latency is <30ms, and it has an anti-bypass mechanism. The family ecosystem linkage module supports quick calls for at least 3 frequently used contacts, one-click help triggered by long press, remote screen sharing and voice guidance. It can synchronize device usage status and fuzzy location to the family social system. The help information push delivery latency is <300ms with a 100% reach rate. It also supports voice control of smart devices throughout the house with a control accuracy of ≥90%.

[0011] Furthermore, the age-friendly interface engine module also supports at least 3 preset high-contrast color schemes and a night mode, with the night mode supporting automatic switching between sunrise and sunset; all buttons use dual "text + icon" labels; supports 200% interface scaling and voice prompts with synchronized text display; and provides both visual and tactile feedback after operation is triggered, with the tactile feedback vibration intensity ≥500Hz and duration 100ms.

[0012] Furthermore, the high-frequency scenario adaptation of the age-friendly interface engine module includes: in health management scenarios, it supports voice input of health data and automatically generates large-font statistical reports; in government affairs scenarios, the operation process is simplified to ≤3 steps; in travel navigation scenarios, only core navigation information is displayed; in video call scenarios, the contact avatar is ≥100dp×100dp and supports voice or automatic answering; in shopping payment scenarios, it hides misleading items and supports voice confirmation of payment amount; health data and operation status can be synchronized to family members' devices.

[0013] Furthermore, the AI ​​intent understanding module's intent recognition model learns online based on user operation data, fine-tuning the model every 300 sets of valid behavioral data collected, with a model update latency of <500ms, and supports one-click reset of the learning model; it can also identify the reasons for operation failures and provide corresponding voice solutions; the companion-style voice interaction module also supports user-defined active interaction frequencies, ranging from 1 to 6 hours, with interactive content including weather reminders, health tips, holiday greetings, and light topic chat, and supports one-click disabling of active interaction; the false wake-up rate is <1 time / hour; and there is a 1-second pause every 10 words during voice guidance.

[0014] Furthermore, the risk prediction and blocking module presets 12 types of high-risk operation scenarios, including: abnormal jump scenarios, sensitive operation scenarios, suspicious page scenarios, continuous inducement scenarios, large payment scenarios, and unfamiliar merchant payment scenarios; the blocking page is a system-level pop-up window that cannot be closed in the background; when ≥3 attempts to bypass the blocking are detected, the risk status is automatically synchronized to preset family members.

[0015] Furthermore, the home ecosystem linkage module also supports electronic fence settings, automatically reminding family members when they go beyond the fence range; it supports emergency medical assistance, automatically dialing 120 emergency number and synchronizing the user's health record to the emergency center and community medical service platform after being triggered; during remote assistance, family members can mark the operation location on their own terminal, and the marked content is displayed in an extra-large size on the elderly user's terminal with a synchronization latency of <50ms; the response latency of voice control of smart devices throughout the house is <1s.

[0016] Furthermore, the system supports three deployment modes: system-level service mode, integrated into the operating system, with background power consumption ≤200mW; application plug-in mode, with a plug-in size <8MB; and smart hardware adaptation mode, adapted to smart speakers, elderly care robots, smartwatches, or public service terminals; and compatible with Android 7.0 and above, iOS 10.0 and above, and HarmonyOS 2.0 and above operating systems.

[0017] This invention also provides an age-friendly AI interaction and minimalist intelligent companionship method, which includes the following steps: S1. After the system is deployed and started, initialize the age-friendly interface engine module and load the minimalist age-friendly interface. This interface has extra-large buttons, high contrast and no ads, and has built-in high-frequency scene entry points. At the same time, complete the initialization of the AI ​​intent understanding module and the companion voice interaction module. S2. When elderly users operate the device, the age-friendly interface engine provides a one-step access to the interactive environment and adapts the interaction mode according to the user type, providing real-time feedback on the operation status; the AI ​​intent understanding module collects multi-dimensional operation behavior data and identifies the operation intent. S3. If the operation is recognized as normal, the voice interaction module will provide voice confirmation feedback; if the operation is recognized as hesitant or repetitive, the repeated explanation and step-by-step guidance will be triggered; if an abnormal emotion is detected, the emotional reassurance script will be activated; if a high-risk scenario is detected, the risk prediction module will execute tiered blocking. S4. When a user triggers the "Call Family", "One-Click Help" or "Emergency Medical" functions, the family ecosystem linkage module sends information to preset family members, emergency centers and community platforms, synchronizes device usage status and location, and supports remote assistance. S5. The system continuously collects user operation data, regularly optimizes the AI ​​intent recognition model, and communicates with anti-fraud systems, smart home systems, and community medical systems through standardized interfaces to form a full-scenario age-friendly protection system.

[0018] Furthermore, in step S3, the risk prediction module presets 12 types of high-risk scenarios, where low risk triggers a lightweight confirmation pop-up, medium risk initiates a 10-second delay confirmation and displays a risk warning, and high risk directly blocks the operation, with a blocking execution success rate of ≥99.9% and an anti-bypass mechanism trigger accuracy of 100%.

[0019] Furthermore, in step S4, during remote assistance, family members can mark the operation location on their own terminal screen, and the marked content is synchronized to the elderly user's terminal in real time and displayed in an extra-large size with a synchronization latency of <50ms; in case of emergency medical assistance, 120 is automatically dialed and the user's health record is synchronized.

[0020] The beneficial effects of this invention are as follows: This invention upgrades from "surface adaptation" to "deep intent understanding and full-scene coverage," resulting in more precise adaptation. Breaking away from traditional surface-level visual optimization methods like "enlarging fonts," this invention uses AI to identify six types of operational intents, including accidental touches, hesitation, and being misled, providing personalized operation guidance. It also covers ten high-frequency scenarios such as health, government affairs, travel, and payment, employing a minimalist design of "one function per screen, one-step access," with no more than two menu levels. The success rate is increased to over 90%, and it is compatible with special groups such as those with hearing and visual impairments, achieving a leap from "usable" to "easy to use and enjoyable to use."

[0021] This invention constructs a triple technological closed loop of "ease of use + security + companionship". Ease of use is ensured through a minimalist interface, security is guaranteed through risk prediction and blocking, and loneliness and frustration are alleviated through companion-like voice interaction, forming a unique technological barrier that distinguishes it from single-function products on the market.

[0022] This invention provides a warm and interactive design that fully caters to the psychological needs of elderly users. It supports emotion recognition and reassurance, unlimited repeated responses, dialect recognition, customized speech speed and tone, and proactive, lightweight interaction, avoiding the mechanical and cold tool-like interaction and truly meeting the psychological and emotional needs of elderly users.

[0023] This invention provides end-to-end security protection, eliminating potential risks. It presets 12 high-risk scenarios, combines intent recognition to proactively block accidental touches, misleading, and fraudulent behaviors, employs a tiered intervention strategy and anti-bypass mechanism, and links with anti-fraud systems. It upgrades from "simple ease of use" to "security + ease of use," effectively addressing the financial security risks of elderly users.

[0024] This invention employs a modular decoupling and multi-form deployment design, resulting in strong commercial adaptability. The system's functional modules are completely decoupled, supporting three forms: system-level services, application plugins, and hardware adaptation. It can be adapted to all types of devices, including mobile phones, televisions, speakers, robots, and public terminals. It can be independently licensed to hardware manufacturers or deployed collaboratively, offering diverse commercial monetization paths.

[0025] This invention achieves ecological collaboration and full-scenario coverage, providing more comprehensive protection. Through standardized interfaces, it links with family social systems, anti-fraud systems, whole-house smart systems, and community medical systems, enabling quick calls to family members, one-click assistance, remote support, voice control of home appliances, and emergency medical coordination. This forms a complete closed loop of "device interaction - family protection - smart living - medical security," exponentially amplifying the ecological value.

[0026] This invention features low latency and high reliability, resulting in a smooth user experience. Interaction response latency is less than 100ms, interface loading latency is less than 300ms, voice recognition accuracy is no less than 92%, dialect recognition accuracy is no less than 88%, risk prediction latency is less than 50ms, there is no noticeable lag, and operating power consumption is low, not affecting terminal battery life, fully meeting the core needs of elderly users for device stability.

[0027] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Detailed Implementation

[0028] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0029] Example 1: System Overall Architecture and Deployment This embodiment provides an age-friendly AI interaction and minimalist intelligent companionship system. The system adopts a five-layer modular architecture, including an age-friendly interface engine module, an AI intent understanding module, a companion-style voice interaction module, a risk prediction and blocking module, and a family ecosystem linkage module. The modules are decoupled and collaboratively linked through standardized interfaces.

[0030] The system can be deployed on a smartphone for seniors running Android 11, equipped with a Snapdragon 662 processor and 4GB of RAM. After installation, the system runs as a system-level service, automatically starting upon boot, with background power consumption of approximately 150mW, far below the design limit of 200mW.

[0031] After system startup, the age-friendly interface engine module is automatically initialized, loading a minimalist age-friendly interface. The phone's home screen presents a "one screen, one core function" layout, displaying only five large icons: "Health," "Government Affairs," "Travel," "Calls," and "Help." Each icon is 90dp x 90dp in size, with an icon spacing of 18dp. The default font is a 28sp sans-serif typeface, with a white background and black text color scheme set to a 7:1 contrast ratio.

[0032] Example 2: Operation Recognition Process of AI Intent Understanding Module An elderly user attempted to make a video call using the aforementioned smartphone. On the first attempt, the user failed to accurately hit the "call" icon, instead tapping a blank area next to it, with a click position deviation of approximately 25 dp. The user then immediately tapped again, with an interval of approximately 180 ms between the two taps.

[0033] The AI ​​intent understanding module collects user behavior data at a frequency of 10Hz, including click location coordinates, click timestamps, and page dwell time. The lightweight LightGBM model determines the operation as an "accidental touch" based on the input features, with a recognition confidence level of 97.3%.

[0034] The companion-style voice interaction module then announced: "It seems you didn't tap the button. Would you like to try again? I've enlarged the button for you." At the same time, the age-friendly interface engine module temporarily enlarged the "call" icon to 110dp×110dp to reduce the probability of accidental tapping again.

[0035] The user's second click accurately struck the enlarged "Call" icon, with a click position deviation of 5dp and a click interval of 1.2 seconds. The AI ​​intent understanding module determined it to be a "normal operation," and the voice interaction module announced, "Okay, calling your daughter for you." The entire process, from the accidental touch to the completion of the guidance and successful dialing, took approximately 2.8 seconds, requiring no additional settings or intervention from the user.

[0036] Example 3: Blocking instance of risk prediction and blocking module While browsing the web, the elderly user's phone screen suddenly displayed an advertisement that read, "Congratulations on winning a free medical check-up opportunity, valid only today." The page included a prominent "Claim Now" button. After being lured into clicking the button, the user was redirected to four different pages before finally arriving at a page requiring them to enter their ID number and bank card number.

[0037] The risk prediction and blocking module monitors page redirection behavior in real time. It detected "more than 3 consecutive redirects to unfamiliar pages in a short period of time" and "redirects from unofficial channels to pages containing sensitive information." At the same time, the AI ​​intent understanding module identified the intent to "be induced to perform an action." The overall assessment was deemed a high-risk level.

[0038] The module immediately executes high-risk blocking actions: a system-level blocking page pops up before the user fills in information, displaying the message "This page may be a scam, it is recommended to stop immediately," accompanied by a voice announcement: "This page is very dangerous, do not fill in any information! Should we call your family to check it out?" The blocking page provides two large buttons, "Call Family" and "OK," and the user cannot close the blocking page using the back or Home button.

[0039] After the user clicked "Call Family," the family ecosystem linkage module automatically sent a distress message to the pre-set son's mobile phone, including the device's approximate location and risk warning. The entire process, from the user clicking the trigger button to the blocking taking effect, took only about 45ms, and the user did not suffer any property loss.

[0040] Example 4: Remote Assistance and Family Collaboration An elderly user was unable to find the "voice input destination" option when using the navigation function. After several failed attempts, the "repeated operation" intent recognition was triggered. The system automatically synchronized the operation failure status to the family ecosystem linkage module, which then pushed a notification to the pre-set family members' mobile phones: "Mom is having trouble using navigation and has failed 3 times. Would you like remote assistance?" After clicking "Remote Assistance," family members can view the live shared screen of the elderly user's device on their own phones and circle the "Voice Input" button on their own screen. This marker appears as a large, bright, flashing circle on the elderly user's phone screen in real time, with a synchronization latency of approximately 45ms. The elderly user successfully locates the voice input button based on the marker and completes navigation settings by voice inputting their destination. Throughout the process, family members can only annotate the screen and provide voice guidance; they cannot remotely control the elderly user's phone, thus fully protecting user privacy.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An age-friendly AI interaction and minimalist smart companion system, characterized in that, The system includes: The age-friendly interface engine module adopts a minimalist layout of "one core function per screen," with a menu hierarchy of ≤2 levels, regular function button size ≥60dp×60dp, high-frequency function button size ≥80dp×80dp, button spacing ≥15dp, font size ≥24sp and supports adjustment from 24sp to 48sp, color contrast ≥4.5:1, and built-in entry points for high-frequency scenarios such as health management, government affairs processing, travel navigation, emergency assistance, and video calls. There are no ad pushes or forced update prompts, and the interface loading latency is <300ms. The AI ​​intent understanding module is based on a lightweight LightGBM model with less than 2M parameters to recognize operation intent. It collects multi-dimensional operation behavior features, including click interval, click position deviation, and page dwell time. It can recognize 6 types of intents, including normal operation, accidental touch, hesitant operation, repeated operation, induced operation, and panic operation. The recognition accuracy is ≥95% and the false judgment rate is <3%. It also supports intelligent diagnosis and step-by-step guidance for operation failure. The companion-style voice interaction module supports full-process voice-first interaction, including voice wake-up, voice commands, and voice feedback, with a wake-up success rate of ≥92%. It provides at least 3 elderly-friendly voice tones, and the speech rate can be adjusted from 40 words / minute to 80 words / minute. It supports at least 6 dialects with an accuracy rate of ≥88%, and has functions such as repeated responses, emotion recognition based on voice tone, and soothing. The emotion recognition accuracy rate is ≥85%. The risk prediction and blocking module presets 12 high-risk operation scenarios and performs tiered blocking based on the recognition results of the AI ​​intent understanding module. Low-risk scenarios execute a lightweight confirmation pop-up, medium-risk scenarios initiate a 10-second delay confirmation, and high-risk scenarios directly block the operation. The prediction latency is <50ms, the blocking response latency is <30ms, and it has an anti-bypass mechanism. The family ecosystem linkage module supports quick calls for at least 3 frequently used contacts, one-click help triggered by long press, remote screen sharing and voice guidance. It can synchronize device usage status and fuzzy location to the family social system. The help information push delivery latency is <300ms with a 100% reach rate. It also supports voice control of smart devices throughout the house with a control accuracy of ≥90%.

2. The age-friendly AI interaction and minimalist intelligent companion system according to claim 1, characterized in that, The age-friendly interface engine module also supports at least 3 preset high-contrast color schemes and a night mode. The night mode supports automatic switching between sunrise and sunset. All buttons use a dual label of "text + icon". It supports 200% interface scaling and synchronized text display with voice prompts; after the operation is triggered, it provides dual visual and tactile feedback, with the tactile feedback vibration intensity ≥500Hz and duration 100ms.

3. The age-friendly AI interaction and minimalist intelligent companion system according to claim 2, characterized in that, The age-friendly interface engine module's high-frequency scenario adaptations include: in health management scenarios, it supports voice input of health data and automatically generates large-font statistical reports; in government affairs scenarios, the operation process is simplified to ≤3 steps; in travel navigation scenarios, only core navigation information is displayed; in video call scenarios, the contact's avatar is ≥100dp×100dp and supports voice or automatic answering; in shopping payment scenarios, it hides misleading items and supports voice confirmation of payment amount; health data and operation status can be synchronized to family members' devices.

4. The age-friendly AI interaction and minimalist intelligent companion system according to claim 3, characterized in that, The AI ​​intent understanding module's intent recognition model learns online based on user operation data, fine-tuning the model every 300 sets of valid behavioral data collected, with a model update latency of <500ms, and supports one-click reset of the learning model; it can also identify the reasons for operation failures and provide corresponding voice solutions; the companion-style voice interaction module also supports user-defined active interaction frequency, ranging from 1 to 6 hours, with interactive content including weather reminders, health tips, holiday greetings, and light topic chat, and supports one-click disabling of active interaction; the false wake-up rate is <1 time / hour; and there is a 1-second pause every 10 words during voice guidance.

5. The age-friendly AI interaction and minimalist intelligent companion system according to claim 4, characterized in that, The risk prediction and blocking module presets 12 high-risk operation scenarios, including: abnormal jump scenarios, sensitive operation scenarios, suspicious page scenarios, continuous inducement scenarios, large payment scenarios, and unfamiliar merchant payment scenarios; the blocking page is a system-level pop-up window that cannot be closed in the background; when ≥3 attempts to bypass the blocking are detected, the risk status is automatically synchronized to preset family members.

6. The age-friendly AI interaction and minimalist intelligent companion system according to claim 5, characterized in that, The family ecosystem linkage module also supports electronic fence settings, automatically reminding family members when they go outside the fence range; It supports emergency medical assistance, automatically dialing 120 emergency number and synchronizing the user's health record to the emergency center and community medical service platform after being triggered; during remote assistance, family members can mark the operation location on their own terminal, and the marked content is displayed in an extra-large size on the elderly user's terminal, with a synchronization latency of <50ms; the response latency of voice control of smart devices throughout the house is <1s.

7. The age-friendly AI interaction and minimalist intelligent companion system according to claim 6, characterized in that, The system supports three deployment modes: system-level service mode, integrated into the operating system, with background power consumption ≤200mW; application plug-in mode, with plug-in size <8MB. It adapts to various smart hardware forms, including smart speakers, elderly care robots, smartwatches, and public service terminals; and is compatible with Android 7.0 and above, iOS 10.0 and above, and HarmonyOS 2.0 and above operating systems.

8. A method for age-friendly AI interaction and minimalist intelligent companionship, characterized in that, The method includes the following steps: S1. After the system is deployed and started, initialize the age-friendly interface engine module and load the minimalist age-friendly interface. This interface has extra-large buttons, high contrast and no ads, and has built-in high-frequency scene entry points. At the same time, complete the initialization of the AI ​​intent understanding module and the companion voice interaction module. S2. When elderly users operate the device, the age-friendly interface engine provides a one-step access to the interactive environment and adapts the interaction mode according to the user type, providing real-time feedback on the operation status; the AI ​​intent understanding module collects multi-dimensional operation behavior data and identifies the operation intent. S3. If the operation is recognized as normal, the voice interaction module will provide voice confirmation feedback; if the operation is recognized as hesitant or repetitive, the repeated explanation and step-by-step guidance will be triggered; if an abnormal emotion is detected, the emotional reassurance script will be activated; if a high-risk scenario is detected, the risk prediction module will execute tiered blocking. S4. When a user triggers the "Call Family", "One-Click Help" or "Emergency Medical" function, the family ecosystem linkage module sends information to preset family members, emergency centers and community platforms, synchronizes device usage status and location, and supports remote assistance. S5. The system continuously collects user operation data, regularly optimizes the AI ​​intent recognition model, and communicates with anti-fraud systems, smart home systems, and community medical systems through standardized interfaces to form a full-scenario age-friendly protection system.

9. The age-friendly AI interaction and minimalist intelligent companionship method according to claim 8, characterized in that, In step S3, the risk prediction module presets 12 types of high-risk scenarios, where low risk triggers a lightweight confirmation pop-up, medium risk initiates a 10-second delay confirmation and displays a risk warning, and high risk directly blocks the operation, with a blocking execution success rate of ≥99.9% and an anti-bypass mechanism trigger accuracy of 100%.

10. The age-friendly AI interaction and minimalist intelligent companionship method according to claim 8, characterized in that, In step S4, during remote assistance, family members can mark the operation location on their own terminal screen. The marked content is synchronized to the elderly user's terminal in real time and displayed in an extra-large size with a synchronization latency of <50ms. In case of emergency medical assistance, 120 is automatically dialed and the user's health record is synchronized.