An AI toy close-range detection active interaction system, method and device

Abstract

The application discloses an AI toy close-range detection active interaction system, method and device, and belongs to the technical field of intelligent human-computer interaction. The system comprises a hierarchical distance sensing module, a master control module, a user identification module, a voice interaction module, an expression feedback module and a vibration feedback module; the device is a hardware carrier for realizing the system, and comprises a master control board assembly, an infrared distance sensor, a sound pickup assembly, a loudspeaker assembly and a quick-change safety shell structure; the method divides four human body distance intervals through infrared induction, actively wakes up the system in the trigger interval, and realizes personalized multi-modal interaction in combination with long-term memory of a user. The application solves the problems of passive interaction, high wake-up false trigger rate and insufficient safety of children using the existing intelligent toys, realizes accurate active wake-up through hierarchical distance detection, outputs personalized content in combination with user identity recognition, and simultaneously adopts child-proof magnetic attraction structure and tool-free quick-change design for the hardware, so that the convenience and safety of use are taken into account, and the interaction experience of young children is significantly improved.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology for intelligent toys, and in particular to an AI toy active interaction system, method and device with graded proximity sensing. Background Technology

[0002] With the rapid development of artificial intelligence and sensor technology, smart toys have become mainstream products for children's companionship and early education. Currently, mainstream smart toys on the market are mainly divided into three categories based on their interaction methods and hardware structure. All categories of products have significant technical shortcomings, as detailed below: 1. Purely passively triggered intelligent toys These products rely on physical buttons, fixed wake words, and touch controls for interaction, and the toys themselves lack active sensing capabilities. For children aged 3-6, this presents a high barrier to entry, as children cannot trigger interactions independently, resulting in a stiff user experience. Furthermore, these products operate entirely in a passive response state, with limited interaction scenarios and a lack of engaging interactive elements.

[0003] 2. Smart toys equipped with single-level distance sensors Some high-end toys are equipped with a single infrared distance sensor, which has only a fixed distance threshold. It wakes up as soon as a person is detected approaching within the range. This solution does not classify the sensing area, and environmental interference such as people walking around, pets passing by, and furniture obstructing or shaking can easily cause frequent false wake-ups. This not only increases device power consumption and shortens battery life, but also generates invalid noise, seriously affecting the user experience. Furthermore, it cannot provide differentiated responses based on the distance of the person, and the interaction logic is too simple.

[0004] 3. Conventional integrated intelligent toys Most existing toys use a single-piece shell that is sealed with glue and fixed with screws, making it difficult to disassemble, repair, and replace the appearance. The battery compartment generally uses a common buckle or screw cap structure without a child-proof opening design. Young children can easily pry open the battery compartment, posing safety hazards such as accidentally swallowing batteries or electric shock. At the same time, the products lack standardized expansion interfaces, making it impossible to connect external devices such as early education modules, lighting modules, and motion sensing modules, resulting in poor functional expandability.

[0005] In summary, existing smart toys generally suffer from problems such as passive interaction modes, low wake-up accuracy, insufficient security, and weak disassembly and expansion capabilities. The industry urgently needs a smart toy solution that can be graded in sensing, actively interact, is safe and reliable, and supports functional expansion. Summary of the Invention

[0006] 1. Purpose of the invention The purpose of this invention is to overcome several shortcomings of existing technologies and provide an AI toy proximity detection active interaction system, method, and device. On the one hand, it reduces the probability of false wake-up through four-level zoned distance detection, enabling active perception and personalized multimodal interaction; on the other hand, it optimizes the hardware structure by adopting a magnetic anti-opening battery compartment, a tool-free quick-change shell, and standardized expansion interfaces, taking into account child safety, ease of disassembly and assembly, and functional expandability, thus comprehensively improving the overall performance of the product.

[0007] 2. Technical Solution To achieve the above objectives, the present invention provides the following technical solution: 2.1 Device Design: The device adopts a hardware architecture of hierarchical infrared distance sensing + multimodal feedback, integrating a main control module, sensor module, and voice / visual / tactile execution module. It is equipped with a child-safe quick-change shell, magnetic protective battery compartment, and blind-plug expansion interface to achieve active wake-up, safe use, and functional expansion.

[0008] 2.2 Methodology: The sensing area is divided into four distance intervals to achieve layered detection and graded response; user identification is completed by combining the local long-term memory module and personalized interactive content is matched; the cloud APP supports custom wake-up thresholds to adapt to different usage scenarios and user groups.

[0009] 3. Beneficial effects (1) Excellent interactive experience: It relies on graded distance detection to achieve active perception and wake-up, without the need for children to operate manually, which is suitable for the usage habits of young people and the interaction process is more natural and smooth. (2) High wake-up accuracy: Four-level partition logic + customizable wake-up threshold effectively filters environmental interference, greatly reduces the frequency of false wake-ups, and reduces the standby power consumption of the device; (3) Personalized interaction: Equipped with a local long-term memory module, it can identify different user identities and match exclusive interactive content to enhance interaction stickiness; (4) High safety level: The shell with rounded corners and the magnetic anti-opening battery compartment fully comply with the national children's toy safety standards and eliminate safety hazards; (5) Strong versatility and expandability: The tool-free quick-change shell can quickly change the appearance, and the standardized quick-connect interface supports the connection of various external functional modules, which can be adapted to multiple scenarios such as early education, companionship, and interactive games; (6) Easy maintenance: The elastic buckle structure does not require screws or glue for fixing, making disassembly and maintenance simple and reducing after-sales costs.

[0010] 4. Comparative Analysis of Existing Technologies and the Invention To clearly demonstrate the inventiveness of this invention, a comparison and explanation of existing mainstream solutions with this invention is provided from the perspective of core technologies: As shown in the table above, this invention simultaneously solves the deficiencies of existing products in four dimensions: interaction logic, sensing and detection, hardware security, and structural design, forming a complete technical solution with outstanding substantive features and significant technological progress. Attached Figure Description

[0011] Figure 1 is a schematic diagram of the AI ​​toy proximity detection active interaction system of the present invention; Figure 2 is an exploded structural diagram of the AI ​​toy close-range detection active interaction device of the present invention; The markings in Figure 2 are as follows: 201 Top cover sealing box, 202 Main control board assembly, 203 Antenna module, 204 Infrared distance sensor module, 205 Dual microphone pickup assembly, 206 Speaker sound guide cavity assembly, 207 Vibration motor suspension shock absorption assembly, 208 Capacitive touch sensor panel, 209 Removable battery compartment assembly, 2091 Child-proof magnetic safety cover, 210 Symmetrical elastic buckle mounting bracket, 211 Adjustable universal fixing buckle, 212 Standardized quick-connect expansion connector. Detailed Implementation

[0012] Those skilled in the art should understand that the following embodiments are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0013] Example 1: Basic embodiment of the device (refer to Figure 2) The hardware structure of the AI ​​toy device of this invention is as follows: The sealed housing 201 has an array of sound outlets and houses a main control board assembly 202. The main control board integrates an MCU, AI chip, WIFI / Bluetooth module, and antenna module 203, enabling data processing, AI semantic interaction, and wireless communication. An infrared distance sensor module 204 is embedded in the sensing window of the device housing and electrically connected to the main control board. Its detection range is 0-200cm, and it can stably identify the distance to a human body.

[0014] The dual-microphone pickup assembly 205 employs a sealed pickup cavity and a windproof noise-reducing cotton structure. The dual-microphone array layout achieves environmental noise reduction and sound source localization, improving speech recognition accuracy in noisy environments. The speaker sound guide cavity assembly 206 incorporates a full-range speaker, optimizing the sound field through the cavity to ensure high-quality external sound. The vibration motor suspension damping assembly 207 uses a silicone damping sleeve to encase the motor body, providing tactile feedback while preventing vibration transmission and resulting noise.

[0015] The detachable battery compartment assembly 209 is located at the bottom of the device and houses a polymer lithium battery pack. It comes with a child-proof magnetic safety cover 2091, which can only be opened with a special magnetic tool, preventing young children from disassembling it by hand. The housing uses a symmetrical elastic snap-fit ​​mounting bracket 210 combined with an adjustable universal fixing buckle 211. Pressing unlocks and inserting locks it in place, requiring no tools and allowing for quick replacement of different housing designs. The standardized quick-connect expansion connector 212 is a blind-plug structure, allowing for connection to external devices such as lighting modules, motion sensing modules, and early education card recognition modules.

[0016] The robot's structure and layout are as follows: an infrared sensor window is placed on the robot's head, an expression display screen is embedded in the face, and a speaker outlet and a pickup micro-hole are integrated and arranged on the robot's chest and abdomen. The hardware layout is adapted to the aforementioned hierarchical infrared zone interactive control logic.

[0017] Automatic return to standby mode when idle: The system status management unit continuously monitors the infrared sensor signal. When there is no human body sensor input within a preset time, it automatically controls the whole machine to return from the working mode to the initial low power standby state.

[0018] All exposed edges and corners of the device are rounded, in accordance with GB 6675 Safety Standard for Children's Toys.

[0019] Example 2: Basic Example of Interaction Method (Refer to Figure 1) The interaction flow under the default operating conditions of this invention is as follows: S1, Standby Zone Monitoring Steps: The robot is normally in a low-power standby state. The infrared sensing module continuously collects the measured distance between the human body and the robot, dividing the sensing zone into four levels according to distance: >200cm undetected standby zone, 150cm~200cm long-distance untriggered zone, 80cm~150cm medium-distance pre-identification zone, and 0~80cm short-distance trigger zone. S1 includes the following sub-steps for hierarchical zone power consumption control: >200cm range, the entire robot remains in sleep mode and is powered off; 150~200cm long distance, only infrared sampling is retained, and other functional modules are in sleep mode; 80~150cm medium distance, user data is pre-buffered, and the robot enters standby mode; only the 0~80cm trigger zone sends a wake-up command and powers on all functional modules.

[0020] S2, Human Approach Sampling Step: When a human moves closer to the robot, the infrared distance sensor collects distance data in real time and transmits the distance data to the distance judgment module; S3. Distance threshold determination steps: The distance determination module retrieves the preset trigger threshold from the threshold configuration storage module. The default wake-up threshold is 80cm. The measured distance is compared with the preset threshold. A valid wake-up trigger signal is generated only when the measured distance is ≤ the preset threshold. S4. User identification steps: After receiving the pre-trigger signal, the user identification module retrieves the user identity file and historical interaction data stored in the long-term memory module to complete the matching and identification of the visiting user. S5. Full-machine mode wake-up steps: The wake-up control module receives a valid trigger signal and, under the control of the system status management unit, controls the robot to switch from low-power standby mode to full-power wake-up working mode. S6, Personalized Voice Broadcast Steps: The voice interaction module retrieves the corresponding customized greeting data from the knowledge base module based on the recognized user identity, and outputs the exclusive voice greeting through the speaker output hole; S6 includes a knowledge base tag retrieval sub-step, in which the knowledge base module stores greeting and question-and-answer data according to user identity tags, and the voice interaction module retrieves the exclusive data in a targeted manner after matching the user identity tags to generate broadcast content; S7, Expression + Vibration Synchronization Feedback Steps: The expression control module retrieves dynamic materials from the expression resource library and drives the expression display screen to synchronously play dynamic expressions that match the voice; at the same time, the vibration feedback module retrieves vibration parameters from the feedback strategy management module and synchronously outputs body vibration feedback that follows the rhythm of the voice; S7 includes an expression timing synchronization control sub-step, where the expression control module receives the playback timing signal from the voice interaction module and retrieves the corresponding dynamic expressions according to the start and pause nodes of the voice, realizing synchronous output of expression animation and voice playback timing; S7 also includes a vibration adaptive adjustment sub-step, where the feedback strategy management module pre-stores multiple levels of vibration intensity and intermittent vibration timing parameters, and the vibration feedback module dynamically switches vibration parameters according to the voice playback speed and sentence breaks; S8. Continuous Dialogue Start and Stop Steps: After the initial greeting interaction is completed, the system automatically switches to continuous dialogue mode. The microphone built into the micro-hole picks up the user's voice in real time. The voice interaction module continuously parses the voice and links with the knowledge base to complete multiple rounds of continuous question and answer interaction, achieving uninterrupted human-computer interaction in a loop.

[0021] Automatic return to standby mode when idle: The system status management unit continuously monitors the infrared sensor signal. When there is no human body sensor input within a preset time, it automatically controls the whole machine to return from the working mode to the initial low power standby state.

[0022] Example 3: Preferred Example of Parameter Variant (Adaptive Parameter Configuration) This solution supports custom parameter adaptation, allowing for differentiated configurations for different user groups and scenarios, and represents a preferred embodiment of the present invention. 1. Distance threshold adaptive adjustment For children aged 3-4: The trigger zone threshold is set to 60cm via a mobile app, and the four-level intervals are adjusted as follows: Undetected zone >180cm, far distance zone 120-180cm, medium distance zone 60-120cm, and trigger zone 0-60cm, shortening the trigger distance to suit children's habit of playing at close range. For school-aged children aged 7 and above: The trigger zone threshold is set to 100cm to expand the sensing range and adapt to children's long-distance interaction scenarios.

[0023] 2. Hardware architecture variations The magnetic safety cover can be replaced with a multi-point magnetic structure to further enhance its anti-opening capability; the dual microphones can be arranged in a staggered top-bottom layout to further enhance the directional noise reduction effect; the quick-connect expansion interface can be equipped with a waterproof rubber ring to adapt to humid environments such as bathrooms and outdoors.

[0024] 3. Timeout parameter variants For home indoor scenarios: the no-interaction timeout is set to 60 seconds; for noisy public scenarios such as shopping malls and amusement parks: the no-interaction timeout is shortened to 30 seconds to reduce ineffective standby power consumption.

[0025] Example 4: Scenario Application 1: Daily Companionship with Young Children This embodiment is designed for home companionship scenarios for children aged 3 to 6, and the device parameters are configured according to the toddler mode (trigger threshold 60cm).

[0026] When a child enters within 0-60cm of the toy, the device automatically wakes up, identifies the user through voiceprint recognition, plays a personalized greeting, and simultaneously displays a cartoon smiley face on the expression display screen and gently vibrates the device. After a child touches the capacitive touch sensor on the device, it automatically switches to nursery rhyme or story playback mode. The device automatically goes into sleep mode if the child leaves the toy for more than 60 seconds.

[0027] In this scenario, the graded detection function effectively filters out false wake-ups caused by family members moving around at a distance. The magnetic battery compartment and the fully rounded corner shell ensure the safety of young children. The tool-free shell can be replaced with cartoon shells regularly to enhance children's sense of novelty.

[0028] Example 5: Scenario Application 2: Early Childhood Education Scenario This embodiment is equipped with an early education recognition module (externally connected via a quick-connect expansion connector) and is suitable for early education scenarios for children aged 4 to 8.

[0029] The device has a default trigger threshold of 80cm. When a child approaches and wakes up the device, they can ask questions via voice, such as Chinese characters, pinyin, English words, or fun science questions. The AI ​​chip, combined with a cloud-based early education knowledge base, provides real-time responses. When a child places an early education card in the recognition area, an external module collects the card information and links with the main unit to provide text, images, and voice explanations.

[0030] Thanks to its dual-microphone noise reduction structure, it can accurately recognize children's voice commands even in multi-person environments such as living rooms and studies; graded distance detection prevents family members from accidentally interrupting the early education process and improves learning focus.

[0031] Example 6: Scenario Application 3: Guiding Children's Social Skills This embodiment is a guidance and training program for introverted children and children with weak social skills, and belongs to the extended application scenarios.

[0032] The device is set up with a multi-person recognition mode. When the first child approaches the trigger area and wakes up the device, the toy will issue an invitation voice. When the second child enters the mid-distance area at the same time, the device will start interactive content such as two-person interactive games and fun dialogues to guide the two children to actively communicate.

[0033] This solution utilizes a hierarchical and zoned logic to differentiate between two states: "single person approaching" and "multiple people approaching," and outputs different interactive content accordingly. This allows children to develop their social skills through fun and interactive activities, expanding the application boundaries of smart toys.

[0034] The above embodiments have fully disclosed the present invention from the perspectives of hardware structure, operation method, parameter variations, and different application scenarios. Those skilled in the art, based on the content of this specification, the accompanying drawings, and the claims, can replicate and modify this technical solution without creative effort. The scope of protection of this invention is not limited to the specific embodiments described above; all equivalent substitutions and simple modifications made based on the core concept of this invention fall within the scope of protection of this invention.

Claims

1. An AI-powered toy proximity detection active interaction device, characterized in that, include: The top cover is a sealed box (201) with an array of sound outlet holes; The main control board assembly (202) is located inside the sealed box of the upper cover and integrates an MCU, an AI chip and a wireless communication module, serving as the control core of the device; The antenna module (203) is electrically connected to the main control board assembly to realize wireless communication function; The infrared distance sensor module (204) is installed at the front end of the device housing and is electrically connected to the main control board assembly. It is used for non-contact detection of human body distance. The dual-microphone pickup assembly (205) adopts a sealed pickup cavity + windproof and noise-reducing cotton structure and is electrically connected to the main control board assembly for collecting user voice signals; The speaker sound guide cavity assembly (206) has a built-in full-range speaker and is electrically connected to the main control board assembly for outputting voice interaction content; The vibration motor suspension damping assembly (207) uses a silicone damping sleeve and a damping mounting base, and is electrically connected to the main control board assembly for outputting vibration feedback. A capacitive touch sensor (208) is located in the middle of the device and is electrically connected to the main control board assembly to supplement touch interaction; A removable battery compartment assembly (209) is located at the bottom of the device, with a built-in lithium battery pack and a child-proof magnetic safety cover (2091). The symmetrical elastic snap-on mounting bracket (210) and the adjustable universal fixing buckle (211) constitute a tool-free quick-change housing mounting structure; The standardized quick-connect expansion connector (212) adopts a blind mating structure and is used for functional expansion connections of the device; All exposed corners of the device are rounded, which meets the safety standards for children's toys.

2. An AI-powered toy proximity detection and active interaction method, applied to the device described in claim 1, characterized in that, Includes the following steps: S1. Standby Zone Monitoring Steps: The robot is normally in a low-power standby state. The infrared sensing module continuously collects the measured distance between the human body and the robot, and divides the sensing zone into four levels according to the distance: >200cm undetected standby zone, 150cm~200cm long distance untriggered zone, 80cm~150cm medium distance pre-identification zone, and 0~80cm close distance trigger zone. S2, Human Approach Sampling Step: When a human moves closer to the robot, the infrared distance sensor collects distance data in real time and transmits the distance data to the distance judgment module; S3. Distance threshold determination steps: The distance determination module retrieves the preset trigger threshold in the threshold configuration storage module. The default wake-up threshold is 80cm. The measured distance is compared with the preset threshold. A valid wake-up trigger signal is generated only when the measured distance is ≤ the preset threshold. S4. User identification steps: After receiving the pre-trigger signal, the user identification module retrieves the user identity file and historical interaction data stored in the long-term memory module to complete the matching and identification of the visiting user. S5. Full-machine mode wake-up steps: The wake-up control module receives a valid trigger signal and, under the control of the system status management unit, controls the robot to switch from low-power standby mode to full-power wake-up working mode. S6. Personalized voice broadcast steps: The voice interaction module retrieves the corresponding customized greeting data from the knowledge base module based on the recognized user identity, and outputs the exclusive voice greeting through the speaker output hole; S7, Expression + Vibration Synchronization Feedback Steps: The expression control module retrieves dynamic materials from the expression resource library and drives the expression display screen to synchronously play dynamic expressions that match the voice; at the same time, the vibration feedback module retrieves the vibration parameters from the feedback strategy management module and synchronously outputs body vibration feedback that follows the rhythm of the voice. S8. Continuous Dialogue Start and Stop Steps: After the initial greeting interaction is completed, the system automatically switches to continuous dialogue mode. The microphone built into the micro-hole picks up the user's voice in real time. The voice interaction module continuously parses the voice and links with the knowledge base to complete multiple rounds of continuous question and answer interaction.

3. The apparatus according to claim 1, characterized in that, It also includes a body structure layout method: an infrared sensing window is placed on the robot's head, an expression display screen is embedded in the face, and a speaker outlet and a pickup micro-hole are integrated and arranged in the robot's chest and abdomen. The hardware layout structure is adapted to the aforementioned hierarchical infrared partition interactive control logic. The dual-microphone pickup component (205) adopts a dual-microphone array design to achieve noise reduction and sound source localization, thereby improving the accuracy of speech recognition.

4. The method according to claim 2, characterized in that, It also includes a remote custom threshold configuration step: wirelessly connect to the threshold configuration storage module via a mobile APP to modify the wake-up distance threshold online and flexibly change the distance range of the near-field trigger zone.

5. The method according to claim 2, characterized in that, It also includes user profile iteration and update steps: after a single continuous dialogue interaction ends, the long-term memory module automatically stores the content of this round of dialogue and user interaction habit data, continuously expanding the user identification database for rapid identification of subsequent visiting users.

6. The method according to claim 2, characterized in that, Step S1 includes a hierarchical and zoned power consumption control sub-step: in the >200cm range, the entire machine remains in sleep mode and is powered off; in the 150~200cm long distance range, only infrared sampling is retained, and the other functional modules are in sleep mode; in the 80~150cm medium distance range, user data is pre-buffered and the machine enters standby mode; only in the 0~80cm trigger zone is a wake-up command sent downwards and all functional modules are powered on.

7. The method according to claim 2, characterized in that, Step S6 includes a knowledge base tag retrieval sub-step. The knowledge base module stores greeting and question-and-answer corpora according to user identity tags. The voice interaction module matches user identity tags and then retrieves exclusive corpora to generate broadcast content.

8. The method according to claim 2, characterized in that, Step S7 includes a vibration adaptive adjustment sub-step. The feedback strategy management module pre-stores multiple vibration intensity levels and intermittent vibration timing parameters, and the vibration feedback module dynamically switches vibration parameters based on the speech rate and sentence breaks in the voice broadcast.

9. The method according to claim 2, characterized in that, The continuous dialogue control method in step S8 is as follows: after entering the continuous dialogue mode, the microphone continuously picks up sound in real time, the voice interaction module parses the user's voice commands in real time, and immediately calls the knowledge base content to respond, thus realizing uninterrupted human-computer interaction in a loop.

10. The method according to any one of claims 4-9, characterized in that, It also includes an automatic return to standby mode when idle: the system status management unit continuously monitors the infrared sensor signal, and when there is no human body sensor input within a preset time, it automatically controls the whole machine to fall back from the working mode to the initial low power standby state.

Images