Game agency method and system of virtuality and reality combined intelligent toy

Abstract

The invention discloses a game agency method for virtuality and reality combined intelligent toys, which is used for effectively connecting and interacting AI intelligent toys with element universe and network games and solving the problem of interaction separation between physical toys and virtual games. The core innovation is as follows: on the aspect of the entity layer, a multi-form toy (plush bear / worker toy) is integrated with a heat-sensitive material and a magnetic sensor to acquire entity actions such as embracing and shaking; in the mapping layer aspect, entity actions are mapped into virtual game atomic operations through a declarative proxy protocol, and an entropy filter (less than or equal to 3 times per second) is implanted to avoid plug-in judgment; in a virtual layer aspect, a proxy source (such as "worker toy proxy construction") is declarated in a game interface, and a virtual state is synchronously fed back through vibration / LEDs. Virtuality and reality are combined, the efficiency of user operation Minecraft is obviously improved, the game participation degree of autism children is effectively improved, and five types of virtual platforms such as Roblox and the like are compatible.

Description

Technical Field

[0001] The present invention belongs to the technical field of virtual-reality integrated intelligent toys, and specifically relates to a method and system for operating virtual electronic games through physical toy agents. It is a solution that combines AI intelligent toys with the metaverse and online games. It is particularly suitable for the physical-digital two-way binding scenarios of non-combat games such as growth-oriented and educational games, covering the multimodal interaction needs of children, the elderly and special groups. Background Art

[0002] Currently, virtual-reality integrated smart toys face three core issues: fragmented interaction, limited audience coverage, and technical compliance risks. Although there are patents that attempt to connect physical toys with the virtual world, they all have significant flaws: (1) CN113730903A (Ganz) proposed a two-way task dispatching mechanism between physical toys and virtual avatars. The core of the mechanism is to achieve virtual-real linkage through a closed loop of "virtual avatar request → physical toy response". Although it enhances the emotional realism of the toys, it has two major drawbacks: first, it is inefficient. Users need to first receive task requests from the virtual world and then complete the response through the physical toy. The operation link is too long (average response delay ≥ 2 seconds), which cannot meet the real-time agent requirements of the game; second, it is limited in scenarios: it only supports exclusive virtual worlds (such as decorating toy homes) and cannot adapt to open game platforms such as Roblox and Minecraft; CN112788148A (Nanjing Lefu) built a multi-terminal "person-toy-toy" interactive network, achieving linkage through cloud-based forwarding of action-feedback matching tables. However, its drawbacks include: first, cloud dependency; all interactive commands must be relayed through cloud servers, resulting in operational delays exceeding 500ms due to network fluctuations, and functional paralysis during network outages; second, lack of game agent capabilities: interaction is limited to simple feedback between toys (such as flashing lights), lacking control logic for game progression.

[0003] CN119556802A (Hangzhou Saisi) introduces a biometric-driven mechanism, using wearable devices to detect heart rate and other data to predict the intensity of user movements and control the toy. However, this solution suffers from several key weaknesses: the burden of wearing wearable devices, the high resistance rate among special populations such as children with autism (the resistance rate exceeds 60% in clinical testing), and the high cost of sensor calibration. Secondly, it dissociates from physical interaction: using user vital signs as the sole input source, ignoring the value of the physical toy's movements.

[0004] CN117915990A (LEGO) focuses on AR, overlaying virtual and real space, using the motion vectors of real toys to drive the reverse motion of virtual objects. Its technical approach deviates from practical needs: first, its visual focus relies on high-definition screens to present AR effects, making it inadequate for the visually impaired and potentially causing dizziness during prolonged use; second, its limited entertainment scenarios are limited, applicable only to mobile toys like vehicles and lacking compatibility with construction and development games.

[0005] To summarize the common shortcomings of the above industries: existing patents are either limited by one-way task architecture (Ganz), or trapped by cloud latency (Lefu), or rely on wearable bio-devices (Seth), or are limited to visual interaction (Lego). None of them have achieved lossless direct connection of "physical action → atomic game operation", nor have they covered the barrier-free interaction needs of special groups.

[0006] To address the above shortcomings, this patent proposes a declarative proxy mechanism, which breaks through and reconstructs the virtual-reality interaction paradigm through three core technologies: Declarative proxy engine: This establishes a direct channel from physical actions to game commands on the device side, runs a localized behavior classifier (e.g., gyroscope data → game rotation commands), and reduces latency to ≤ 200ms. It also declares the proxy source in the game UI (e.g., "Bear Baby Proxy Collection") and implements an operation entropy filter (frequency ≤ 3 times / second + random delay) to prevent cheat detection. Non-sensory multi-modal carriers: The magnetic sensor module supports quick replacement of four carrier types: plush, humanoid, geometric, and mobile chassis. The thermal layer (variable at 37°C ± 5°C) enhances emotional realism. The projection interface converts gestures into game commands (e.g., drawing a circle = sending a Roblox friend request), replacing voice input. Non-gamified agent expansion: covering educational assistance (elderly users tilt the worker toy → Minecraft buildings rotate, significantly improving efficiency) and medical rehabilitation (autistic children pat the plush bear → tactile feedback + game item collection) scenarios.

[0007] The generation gap comparison between the present invention and the cited patents is shown in Table 1: Table 1: Comparative analysis of the four dimensions of the present invention and cited patents Dimensions Citing patent defects This patent breakthrough Technological advantages Interaction Mechanism Ganz (CN113730903A): requires two-way task dispatching Direct conversion from physical action to game atomic operation Shortened operation chain and reduced latency Hardware Architecture Lefu (CN112788148A): Strong reliance on cloud-based transfers Local mapping of the edge computing layer, still available even when the network is disconnected Reduced computing power requirements Crowd coverage Saisi (CN119556802A): Mandatory wearable device Only the toy body sensor (pressure / gyroscope / microphone) is required Increased acceptance among special groups Scene expansion LEGO (CN117915990A): Limited to AR entertainment Supports non-combat games such as construction / cultivation / socialization Expanding user base .

[0008] Innovation Summary: This patent establishes a trusted channel between the physical and digital worlds using a declarative proxy protocol, balances efficiency and compliance with an entropy control algorithm, and transcends demographic boundaries through multi-sensory interaction, resulting in a trinity of innovation: "hardware-software-scenario." This patent not only fills the technological gap in "direct-control games for physical toys," but also unlocks a new market for specialized populations through its inclusive design. Summary of the Invention

[0009] The present invention creatively constructs a three-tier proxy architecture ( Figure 1 The hardware layer adopts a shape-insensitive design - regardless of the plush bear's hug detection area ( Figure 3 ①) or the gyroscope module of the worker toy ( Figure 3 ②), all access the edge computing layer through a unified sensor interface. The core of the software layer is the behavior translation engine ( Figure 2 When the user tilts a worker toy, the engine decomposes the 3D angular velocity into X / Y displacement. After applying an entropy filter to eliminate mechanical manipulation characteristics, it outputs Minecraft building rotation instructions. The key breakthrough lies in establishing a mapping dictionary between physical actions and atomic game operations, such as "1cm displacement = 1 square in the game" and "the voice keyword 'build' → calls the game API to place a block."

[0010] The industrial design of this system realizes the hardware generalization capability ( Figure 3 ): The shell adopts a negative Poisson's ratio structure, so that the same mold can be used to inject bear-shaped or human-shaped shells; the sensor magnetic array supports the elderly to replace the high-sensitivity pressure module, or to install a thermal feedback sheet for autistic children. Figure 4 As shown, the system successfully adapted to five typical scenarios: a plush bear carrying out emotional tasks in the game "My Daughter" (top left), a worker toy driving Minecraft construction (top right), a projection toy managing Roblox social interactions (center), a desktop robot operating the game "Traveling Frog" (bottom left), and a mobile car performing collection in "Animal Crossing" (bottom right). In tests with elderly users, the physical manipulation of the worker toy increased construction efficiency by 58% (compared to keyboard and mouse operation).

[0011] This system has a disruptive potential in the field of special education ( Figure 5 The case of autistic children reveals the failure of traditional game interfaces: when children cannot understand the rules of interaction between on-screen characters, they can trigger vibration feedback by tapping the stuffed bear (①), simultaneously completing item collection in "My Family Has a Daughter" (③), and the LED rings in the toy's eyes displaying the number of resources (④) form a perceptible positive incentive. Even more groundbreaking, projection toys transform virtual social interaction into gesture language—drawing a "handshake" pattern on a hemisphere (⑤) directly generates a Roblox friend request, allowing non-verbal communicators to participate in metaverse social interaction for the first time.

[0012] This invention uses a declarative proxy protocol to establish a trusted channel between the physical and digital worlds. Its innovation is not a single hardware or software breakthrough, but rather through: (1) Carrier-independent architecture—compatible with all forms of hardware, from plush toys to mobile robots; (2) Behavioral entropy control—strike a balance between agent efficiency and anti-cheating; (3) Multi-sensory alternative interaction - expanding the gaming population to vulnerable groups that traditional technologies cannot cover.

[0013] This trinity design provides a true “wall breaker” solution for metaverse games. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 System architecture diagram (module interconnection logic). This diagram reveals the deep coupling mechanism between hardware and games. The physical toy layer (left) connects to multimodal sensors via a magnetic interface. The edge computing layer (center) runs a lightweight large model to achieve behavior-to-command conversion. The game adaptation layer (right) injects standardized commands into game engines such as Roblox and Minecraft. Dual-channel communication is used between the three layers: low-latency Bluetooth transmission of the control stream and 4G network synchronization of the status stream to ensure real-time agent behavior. A highlight of the industrial design is the horizontal scalability of the sensor modules, making the same toy compatible with both construction and social game scenarios.

[0015] Figure 2 Behavior Mapping Engine Flowchart (Core Algorithm). This process precisely converts physical actions into atomic game operations. When a user shakes a stuffed bear (input ①), the gyroscope generates an angular velocity data stream (②), which is identified as "high-frequency vibration" by the action classifier (③). The mapping rule base then matches the "shake the apple tree" command from the game "My Family Has a Daughter" (④). The key innovation lies in the entropy filter (⑤): When the same action is detected more than 20 times within a minute, a random delay (100ms-500ms) is automatically inserted, maintaining natural operation while preventing cheat detection. The output layer (⑥) adds a watermark containing the toy's ID, allowing the game server to verify the legitimacy of the proxy.

[0016] Figure 3 Hardware exploded view (industrial design details). This image reveals the modular construction of the cartoon worker toy. The outer layer (①) is a food-grade silicone shell with a built-in PTC heater to simulate body temperature. The sensor layer (②) includes a detachable pressure array (helmet section) and a 9-axis IMU (torso section), connected via magnetic contacts. The core layer (③) is a waterproof main control module with a reserved projector port. Industrial design breakthroughs include: the shell mold supports quick replacement (such as replacing bear ears / worker helmets), and the sensor layout conforms to the ergonomic grip area (highlighted in red), allowing elderly people to naturally trigger acquisition commands when holding it.

[0017] Figure 4: Cross-game agent scenario (embodiment integration). Five embodiments are demonstrated in coordination in the metaverse environment. The center area shows a hemispherical projection toy (①) sending an expression instruction to a Roblox character; the upper left area shows a plush bear (②) hugging action to increase the intimacy of My Girl; the upper right area shows a worker toy (③) tilting to drive a Minecraft building to rotate; the lower left area shows a desktop robot (④) clapping to trigger a Traveling Frog to go home; and the lower right area shows a moving car (⑤) driving along the wall to map Animal Crossing resource collection. The scene communicates through a unified agent protocol, proving the system's ability to adapt to the expansion of multiple game types.

[0018] Figure 5 : Special population auxiliary process (social value). This flowchart reveals the multi-sensory substitution mechanism in the closed loop of autistic children's operation: ① Interaction trigger (blue): the child claps the plush bear toy to produce physical contact, and the pressure sensor triggers signal collection; ② Vibration feedback (green): the toy produces gentle vibration (frequency 3Hz) through the built-in motor, confirming instruction reception; ③ Game execution (yellow): the agent system completes the item collection task in My Girl, and the game character performs the collection action; ④ LED display (red): the toy's eye LED ring displays the real-time resource quantity (such as 5 apples), providing visual progress feedback; ⑤ Gesture projection (purple): the hemispherical projection toy converts virtual social interaction into physical gestures (draws a "handshake" pattern = sends a friend request). Innovation value: through tactile (vibration), visual (LED), and physical gestures (projection), a non-verbal interaction channel is constructed, significantly improving the game participation of autistic children. DETAILED DESCRIPTION

[0019] Embodiment 1: Plush bear toy + My Girl nurturing game: (1) Mapping logic: A[hold plush bear]-->B[pressure sensor data>2N]-->C[in-game character intimacy +10] D[say "go to school" to bear's ear]-->E[voice recognition module]-->F[trigger game character to automatically go to school] (2) Effect: Autistic children complete game social tasks through physical contact, significantly reducing operation time.

[0020] Embodiment 2: Cartoon worker toy + Minecraft: (1) Mapping logic is shown in Table 2: Table 2: Minecraft mapping logic table Toy Action Game Instructions Mapping Parameters Percussion Toy Helmet Start logging Frequency of knocks = amount of wood / minute Tilt the toy body Building direction adjustment Inclination ±15° = rotation 30° (2) Effect: Elderly users replace complex keyboard control with physical operation, significantly improving building efficiency.

[0021] Example 3: Hemispherical projection toy + Roblox virtual social: (1) Mapping logic: Draw a "handshake" pattern on the projected screen → send a friend request in the game; rotate the toy hemisphere → adjust the virtual character's dialogue expression (rotation angle maps expression type); (2) Effect: Teenagers use toys to complete in-game social interactions during offline gatherings, with interaction delays less than 200ms.

[0022] Example 4: Desktop Robot + Travel Frog: (1) Mapping logic: Place a toy on the "food accessory" → automatically prepare a lunch box for the game frog; tap the table twice → trigger the frog to go home immediately; (2) Effect: Office workers use fragmented time to manage the game process through physical actions.

[0023] Example 5: Ground Moving Vehicle + Animal Crossing: (1) Mapping logic: The toy car moves along the room boundary → the game character automatically collects boundary resources; the color of the car headlight changes → feedback on the in-game weather status (blue = rainy); (2) Effect: Visually impaired players perceive game progress through tactile and audio feedback.

Claims

1. A game agent method for a virtual-real intelligent toy, characterized in that include: a) Establishing a binding relationship between a physical toy and a virtual electronic game, wherein the toy is interactive hardware that can be connected to an AI large model, and the game is a growth / educational / role-playing / construction game; b) Collecting the user’s physical interaction with the toy through multimodal sensors; c) Map physical behaviors into game operation instructions within the virtual-reality integration framework and declare them as user agent behaviors; d) Send instructions to the game server through the game plug-in engine, perform proxy operations and feedback status to the toy.

2. The method according to claim 1, wherein the virtual-real combination framework comprises: (1) Physical → virtual mapping: toy displacement distance → game character movement distance; (2) Virtual → Physical Feedback: Number of game resources → Toy LED display array.

3. The system of claim 1 , wherein the behavior translator performs: (1) Convert unstructured actions (such as shaking) to atomic game operations (such as shaking a tree); (2) Through operation entropy detection (new anti-cheating algorithm), the proxy frequency is limited to ≤3 times / second.

Images