Task execution method, apparatus, device, and product

Abstract

The present disclosure relates to a task execution method, apparatus, device and product. The method comprises obtaining pet status information, the pet status information comprising pet behavior information and pet emotion information. The method comprises determining, by a first machine learning model, pet demand information based on the pet status information. The method comprises determining, by a second machine learning model, an action strategy based on the pet demand information, pet attributes and a pet interaction strategy library. The method comprises controlling an intelligent robot to perform an action for the pet based on the action strategy.

Description

Technical Field

[0001] This disclosure relates to the field of computers, and more specifically to methods, apparatus, devices and products for performing tasks. Background Technology

[0002] Intelligent robots (hereinafter referred to as robots) are most widely used in manufacturing. They can perform various complex, repetitive, or dangerous tasks, such as welding, painting, assembly, and handling, thereby improving production efficiency, reducing costs, and ensuring consistent product quality. Furthermore, with continuous technological advancements, the collaboration between robots and humans is constantly evolving, and the application of robots is expanding beyond the industrial sector to more areas, such as home education and pet companionship. Summary of the Invention

[0003] In a first aspect of the embodiments of this disclosure, a task execution method is provided, applicable to an intelligent robot. The method includes acquiring pet state information, including pet behavior information and pet emotional information. The method includes determining pet demand information using a first machine learning model based on the pet state information. The method includes determining an action strategy using a second machine learning model based on the pet demand information, pet attributes, and a pet interaction strategy library. The method includes controlling the intelligent robot to perform actions for the pet based on the action strategy.

[0004] In a second aspect of the embodiments of this disclosure, a task execution apparatus suitable for an intelligent robot is provided. The apparatus includes a pet state information acquisition module configured to acquire pet state information, including pet behavior information and pet emotional information. The apparatus includes a pet demand information determination module configured to determine pet demand information based on the pet state information using a first machine learning model. The apparatus includes an action strategy determination module configured to determine an action strategy based on the pet demand information, pet attributes, and a pet interaction strategy library using a second machine learning model. The apparatus includes a control module configured to control the intelligent robot to perform actions for the pet based on the action strategy.

[0005] In a third aspect of the embodiments of this disclosure, an electronic device is provided. The electronic device includes one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the method of the first aspect.

[0006] In a fourth aspect of the embodiments of this disclosure, a computer program product is provided. The computer program product is tangibly stored on a non-transitory computer-readable medium and includes machine-executable instructions that, when executed, cause a machine to implement the method of the first aspect.

[0007] The summary section is provided to present the chosen concepts in a simplified form, which will be further described in the detailed description below. The summary section is not intended to identify key or principal features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter. Attached Figure Description

[0008] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0009] Figure 1 A schematic diagram of an example environment in which several embodiments of the present disclosure may be implemented is shown;

[0010] Figure 2 A flowchart of a task execution method according to some embodiments of the present disclosure is shown;

[0011] Figure 3 A schematic diagram illustrating an example flow of task execution according to some embodiments of the present disclosure is shown;

[0012] Figure 4 A schematic diagram illustrating example steps of environment perception and initialization according to some embodiments of the present disclosure is shown;

[0013] Figure 5 A schematic diagram illustrating example steps for pet condition monitoring according to some embodiments of the present disclosure is shown;

[0014] Figure 6 A schematic diagram illustrating example steps of interactive decision-making and action according to some embodiments of the present disclosure is shown;

[0015] Figure 7 A schematic diagram illustrating example steps of data logging and feedback according to some embodiments of the present disclosure is shown;

[0016] Figure 8 A schematic diagram illustrating example steps of continuous learning and optimization according to some embodiments of the present disclosure is shown;

[0017] Figure 9 A block diagram of a task execution apparatus according to some embodiments of the present disclosure is shown; and

[0018] Figure 10 A block diagram of a device capable of implementing several embodiments of the present disclosure is shown. Detailed Implementation

[0019] It is understood that all user-related data involved in this technical solution should be obtained and used only after authorization from the user. This means that if it is necessary to use a user's personal information in this technical solution, the user's explicit consent and authorization are required before obtaining this data; otherwise, no related data collection and use will be carried out. It should also be understood that when implementing this technical solution, relevant laws and regulations should be strictly followed in the process of data collection, use, and storage, and necessary technical measures should be taken to protect user data security and ensure the secure use of data.

[0020] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0021] For example, upon receiving a user's proactive request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0022] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0023] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0024] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0025] In the description of embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects unless explicitly stated. Other explicit and implicit definitions may also be included below.

[0026] As mentioned above, robots have demonstrated a high level of capability in collaborating with humans, and their main applications largely focus on effective interaction and cooperation with humans. However, as more and more people in modern society choose to keep pets as companions, a significant demand gap is gradually emerging: current robot applications lack robots capable of interacting with pets, especially robots that can interact with pets based on their condition.

[0027] In view of this, this disclosure proposes a task execution method applicable to intelligent robots, capable of replacing humans in communicating and interacting with pets. In the solution provided in this disclosure, a first machine learning model determines the pet's needs based on the obtained pet's state information, which includes the pet's behavioral and emotional information. Then, a second machine learning model determines the actions to be performed by the robot on the pet based on the pet's needs information, pet attributes, and a pet interaction strategy library.

[0028] In this way, pet companion robots can care for pets when humans are unable to be with them, effectively alleviating pets' anxiety caused by loneliness, reducing undesirable behaviors, and improving their quality of life. At the same time, they also make it easier and more convenient for humans to care for their pets, enhancing the human experience of using the robot.

[0029] Figure 1 A schematic diagram of an example environment 100 in which various embodiments of this disclosure may be implemented is shown. For example... Figure 1 As shown, the example environment includes a robot 110 and a pet 120. In embodiments of this disclosure, the robot 110 is a programmable machine capable of automatically performing tasks, comprising mechanical structures, a control system, sensors, and actuators.

[0030] In some embodiments, the robot 110 is equipped with hardware devices, including LiDAR, a high-definition camera, a WiFi device, a Bluetooth beacon receiver / transmitter, a sound acquisition device, and other hardware devices. In some embodiments, the robot 110 is also equipped with physiological monitoring devices, such as an infrared body temperature sensor, to obtain the pet's body temperature.

[0031] In some embodiments, the robot 110 is configured with a machine learning model for calculation and analysis, used to calculate and analyze information such as the pet's behavior and status in real time. In some embodiments, the machine learning model also has a built-in or callable interactive strategy library to achieve control of the robot based on action strategies. In some embodiments, for data stability, the robot 110 can also upload the collected and analyzed data to the cloud, which can be a computing system, a single server, a distributed server, etc.

[0032] Pet 120 refers to an animal kept or kept as a companion animal by humans. These animals live with humans and provide them with companionship, joy, and emotional support. Pet 120 can be a dog, cat, bird, fish, hamster, rabbit, etc., and the embodiments disclosed herein do not limit this to any particular species.

[0033] In embodiments of this disclosure, robot 110 can interact and communicate with pet 120 based on acquired pet status information. In some embodiments, robot 110 can patrol the pet's environment (e.g., a home environment) to acquire pet 120 status information. In some embodiments, the patrol time of robot 110 can be at certain time intervals. Pet status information may include behavioral information and emotional information.

[0034] For example, robot 110 can continuously capture images of its surrounding environment using its equipped high-definition camera and analyze these images in real time using computer vision algorithms (e.g., convolutional neural network algorithms, CNN). Taking a home environment as an example, robot 110 can inspect each room to determine whether pet 120 is standing, sitting, lying down, or moving. Using computer vision algorithms, it can identify the pet's movement sequences, thereby continuously analyzing the pet's behavioral information.

[0035] In some embodiments, the robot 110 can also utilize a sound acquisition device to listen to sounds in the home environment, capturing various sounds such as barking, footsteps, and breathing from the pet via a microphone array in the sound acquisition device. After acquiring these sounds, audio processing technology can be used to analyze the frequency, intensity, and rhythm of the sounds to determine the pet's emotional state and behavioral intentions. For example, if the robot 110 captures an urgent bark from the pet, it can combine this with image information captured by a high-definition camera to determine whether the pet is trapped somewhere or needs urgent attention, thereby determining the pet's emotional state, i.e., identifying the pet's emotional information.

[0036] In addition, the pet's status information may also include pet health information. Pet health information includes historical health information and currently acquired health information. In some embodiments, the robot 110 can determine the pet's health information through a sequence of acquired images. For example, the condition of the pet's fur can be used to determine if the pet is unwell. In some embodiments, the robot 110 can use an infrared temperature sensor to measure the pet's body temperature to identify whether the pet is unwell.

[0037] In embodiments of this disclosure, a first machine learning model can determine the pet's needs based on the pet's state information. The first machine learning model can be a large language model (LLM) or a multimodal large language model. For example, if the aforementioned pet state information indicates that pet 120 is lingering near its food bowl, sniffing it, and showing signs of eagerness, then pet 120 needs to eat. As another example, if the aforementioned pet state information indicates that pet 120 is frequently licking its mouth or moving towards its water bowl, then pet 120 needs to drink water.

[0038] For example, if the pet status information mentioned above indicates that pet 120 (such as a cat or dog) is actively approaching robot 110, wagging its tail, or rubbing its head against robot 110, this indicates that the pet needs attention and play. As another example, if the pet status information mentioned above indicates that pet 120 (such as a cat or dog) is pacing back and forth or biting its own tail, this indicates that the pet needs comfort and relaxation.

[0039] In embodiments of this disclosure, a second machine learning model can determine action strategies based on the pet's needs information, the pet's attributes, and a pet interaction strategy library. The second machine learning model can be a Large Action Model (LAM). In some embodiments, the pet's attributes can be its breed, personality, or other attributes. In some embodiments, the pet interaction strategy library predefines interaction strategies with pet attributes as the primary key. For example, if the aforementioned needs information indicates that pet 120 needs companionship and play, and if pet 120 is a dog, combined with the pet interaction strategy library, the action strategies can be determined to be throwing a ball, tug-of-war, or petting its head; if pet 120 is a cat, then the action strategies can be determined to be waving a cat toy, providing catnip toys, etc.

[0040] In embodiments of this disclosure, the robot can be controlled to perform actions directed at a pet, based on an action strategy. For example... Figure 1 As shown, based on the pet's needs, attributes, and pet interaction strategy library, it can be determined that the interaction strategy for the pet 120 is to pat its head. Then, the robot 110 can be controlled to walk to the side of the pet 120 and pat the head of the pet 120 with its equipped mechanical arm.

[0041] If the action strategy is to add water or food to pet 120, and the food or water bowl is in the kitchen while robot 110 is in the living room, then robot 110 can use lidar or high-definition cameras to calculate a collision avoidance path to reach the food or water bowl smoothly and add water or food to pet 120.

[0042] In this way, pet companion robots can care for pets when humans are unable to be with them, effectively alleviating pets' anxiety caused by loneliness, reducing undesirable behaviors, and improving their quality of life. At the same time, it makes pet care easier and more convenient for humans, enhancing the user experience of using the robot.

[0043] It should be understood that the architecture and functionality in example environment 100 are described for illustrative purposes only and do not imply any limitation on the scope of this disclosure. Embodiments of this disclosure can also be applied to other environments with different structures and / or functionalities. For example, in some embodiments, environment 100 may also include multiple pets and robots; in some embodiments, the robots can communicate directly with each other; one robot can serve multiple pets; one robot can exclusively serve one pet; and multiple robots can serve one pet.

[0044] The following will combine Figures 2 to 10 The method according to embodiments of this disclosure is described in detail. For ease of understanding, the specific data mentioned in the following description are exemplary and not intended to limit the scope of this disclosure. It is understood that the embodiments described below may also include additional actions not shown and / or actions shown may be omitted, and the scope of this disclosure is not limited in this respect.

[0045] Figure 2 A flowchart of a task execution method 200 according to some embodiments of the present disclosure is shown. Method 200 is applicable to intelligent robots and can be executed by a task execution device, which may be, for example, a standalone device or system. The device can be implemented in software and / or hardware. The method 200 will now be illustrated schematically using a task execution device as an example. Method 200 includes blocks 202, 204, 206, and 208.

[0046] like Figure 2 As shown in box 202, pet status information is obtained, including pet behavior information and pet emotional information. Figure 1As shown, robot 110 can interact and communicate with pet 120 based on the acquired pet status information. In some embodiments, robot 110 can patrol the pet's environment (e.g., a home environment) to acquire pet 120's status information. The pet's status information can include behavioral and emotional information. For example, robot 110 can continuously capture images of its surrounding environment using its equipped high-definition camera and analyze these images in real time using computer vision algorithms (e.g., convolutional neural network algorithms, CNN). With the help of computer vision algorithms, the pet's movement sequences can be identified, thereby continuously analyzing the pet's behavioral information. Taking a home environment as an example, robot 110 can patrol each room in the home environment to obtain behavioral information such as whether pet 120 is standing, sitting, lying down, or moving.

[0047] In some embodiments, the robot 110 can also utilize a sound acquisition device to listen to sounds in the home environment, capturing various sounds such as barking, footsteps, and breathing from the pet via a microphone array in the sound acquisition device. After acquiring these sounds, audio processing technology can be used to analyze the frequency, intensity, and rhythm of the sounds to determine the pet's emotional state and behavioral intentions. For example, if the robot 110 captures an urgent bark from the pet, it can combine this with image information captured by a high-definition camera to determine whether the pet is trapped somewhere or needs urgent attention, thereby determining the pet's emotional state, i.e., identifying the pet's emotional information.

[0048] In box 204, the first machine learning model determines the pet's needs based on the pet's status information. (Reference) Figure 1 For example, if the pet status information mentioned above indicates that pet 120 is lingering near its food bowl, sniffing it, and showing signs of eagerness, then the first machine learning model can infer that pet 120 needs to be fed. As another example, if the pet status information mentioned above indicates that pet 120 is frequently licking its mouth or moving towards its water bowl, then the first machine learning model can infer that pet 120 needs to drink water. As yet another example, if the pet status information mentioned above indicates that pet 120 (such as a cat or dog) is actively approaching robot 110, wagging its tail, or rubbing its head against robot 110, then the first machine learning model can infer that pet 120 needs attention and play. As yet another example, if the pet status information mentioned above indicates that pet 120 (such as a cat or dog) is pacing back and forth or biting its tail, then the first machine learning model can infer that pet 120 needs comfort and relaxation. In some embodiments, the first machine learning model can be a large language model or a multimodal large language model.

[0049] In box 206, the second machine learning model determines the action strategy based on pet needs information, pet attributes, and a pet interaction strategy library. (See reference) Figure 1 In some embodiments, the pet's attributes can be its breed, personality, or other attributes. In some embodiments, the pet interaction strategy library predefines interaction strategies with pet attributes as the primary key. For example, the aforementioned requirement information indicates that pet 120 needs companionship and play. If pet 120 is a dog, and combined with the pet interaction strategy library, the action strategy can be determined to be throwing a ball, playing tug-of-war, or petting its head; if pet 120 is a cat, then the action strategy can be determined to be waving a cat toy, providing catnip toys, etc. In some embodiments, the second machine learning model can be a large-scale action model.

[0050] In box 208, based on the action strategy, the intelligent robot is controlled to perform actions for the pet. For example... Figure 1 As shown, based on the pet's needs, attributes, and pet interaction strategy library, it can be determined that the interaction strategy for the pet 120 is to pat its head. Then, the robot 110 can be controlled to walk to the side of the pet 120 and pat the head of the pet 120 with its equipped mechanical arm.

[0051] According to the method described above 200, pet companion robots can care for pets when humans are unable to be with them, effectively alleviating pets' anxiety caused by loneliness, reducing undesirable behaviors, and improving their quality of life. At the same time, they also make it easier and more convenient for humans to care for pets, enhancing the user experience of using the robot.

[0052] To enable interaction between a robot and a pet, the pet's needs can be determined based on its status information. Then, the pet's needs, attributes, and an interaction strategy library are combined to tailor a suitable interaction strategy for the robot and pet. Finally, the robot executes this strategy to provide companionship to the pet. For example... Figure 3 A schematic diagram of an example flow 300 of task execution according to some embodiments of the present disclosure is shown. Figure 3 The example process 300 shown can be generated by Figure 1 The robot 110 shown is used to perform this action. For ease of explanation, the following description uses the example of the robot interacting with a pet in a home environment. The robot can also interact with pets in an outdoor environment. This disclosure does not limit the interaction environment.

[0053] At step 310, environmental perception and initialization operations are performed. In some embodiments, after the robot starts, it can first use a LiDAR to emit a beam of light to measure the distance between itself and surrounding objects. Simultaneously, a high-definition camera can be used to capture image information of the surrounding environment. By fusing the LiDAR data and image data, and using a Simultaneous Localization and Mapping (SLAM) algorithm, a two-dimensional or three-dimensional map of the home environment can be gradually constructed. For ease of explanation, the following will combine... Figure 4 This describes the execution of environment awareness and initialization operations.

[0054] For example, Figure 4 A schematic diagram illustrating example step 400 of environment perception and initialization according to some embodiments of the present disclosure is shown. Figure 4 As shown in section 311, the robot performs the operation of building and labeling a home map. For example, for a typical family living room, the robot can identify the location and shape of furniture such as sofas, coffee tables, and televisions, and label them on the map.

[0055] At 312, a positioning operation is performed. In some embodiments, to determine the robot's location in the home environment, positioning can be performed using Wi-Fi signals at 312-1. The robot can use the Wi-Fi signal strength in the home for triangulation. The robot can receive signals from multiple Wi-Fi access points and calculate its distance to each access point based on the differences in signal strength, thereby determining its position on the map. Alternatively, at 312-2, positioning can be performed using Bluetooth signals; the robot can also determine its location by receiving signals from Bluetooth beacons.

[0056] Meanwhile, during the map building and localization process, the robot can continuously self-calibrate and correct itself to improve location accuracy. For example, if the robot discovers that a location on the map is inconsistent with the actual observed environment while moving, it can rescan the area and update the map.

[0057] In step 313, the operation of retrieving pet information is performed. This information includes the pet's historical health information, pet attributes, and so on. Humans can set the pet's attributes (such as dogs, cats, rabbits, hamsters) in the robot's local storage configuration, and can also set the pet's age. Furthermore, the pet's historical health information can be set. For example, whether the pet has specific diseases (such as diabetes, heart disease, etc.), whether it has food allergies, etc., so that it can play a role in generating interaction strategies later. For example, the pet's dietary attributes can also be set, including preferred food types, feeding times, and food amounts, which helps the robot better provide personalized dietary services for the pet.

[0058] The exemplary environmental perception and initialization step 400 enables the robot to support interactive behaviors with the pet in subsequent steps, thereby enhancing the pet interaction experience.

[0059] Back Figure 3 In step 320, perform a pet status monitoring operation. For clarity, the following will combine... Figure 5 This describes the operations performed to monitor a pet's condition. For example, Figure 5 A schematic diagram of example step 500 for pet status detection according to some embodiments of the present disclosure is shown. Figure 5 As shown in section 321, the robot performs inspections based on preset time intervals. The robot can inspect the home environment at preset time intervals (e.g., once every 10 minutes), and can use motor drive and motion control technology to move along a pre-planned path.

[0060] In section 3211, the robot performs image capture and real-time analysis. During its movement, the high-definition camera continuously captures images of the surrounding environment and analyzes these images in real time using computer vision algorithms. Taking a home environment as an example, the robot can first perform a panoramic scan of each room to identify potential pet locations. For instance, in the bedroom, the robot can check under the bed, next to the wardrobe, and other places the pet likes to hide. If the pet is spotted, the robot can further focus and analyze the pet's posture and movements. For example, object detection algorithms can be used to determine whether the pet is standing, sitting, lying down, or moving.

[0061] In addition to visual perception, the robot can also use a sound acquisition module to listen to surrounding sounds. Therefore, in the 3212, it can perform operations to determine the type of sound through microphones. The microphone array can capture various sounds emitted by the pet, such as barks, footsteps, and breathing. Using audio processing technology, the robot can analyze the frequency, intensity, and rhythm of these sounds, and then combine this with image information to determine if the pet's behavior is abnormal. For example, if the robot hears the pet barking urgently, it can combine this with visual information to determine if the pet is acting abnormally and requires immediate attention.

[0062] By conducting inspections, images and sounds of the pet can be acquired. Based on the collected images and sound information, the pet's behavioral information can be determined. For example, by analyzing consecutive image frames, the sequence of the pet's movements can be identified, and combined with the intensity, rhythm, and frequency of the pet's sounds, behavioral information such as running, jumping, scratching, and rolling can be determined. In some embodiments, the pet's current emotional state can be determined based on its behavioral information. For example, if the pet actively approaches the robot, it indicates that the pet is experiencing an emotion that requires attention. Or, if the pet paces incessantly, it indicates that the pet is anxious.

[0063] The inspection process can also determine the pet's location. Therefore, in section 3213, the operation of determining the pet's location can be performed. That is, based on the collected environmental information (including visual images) and environmental map, the pet's location is determined so that the robot can subsequently perform health checks on the pet or perform actions on the pet.

[0064] The pet's health can also be assessed through inspection. Therefore, at 322, a health status detection operation is performed. At 3221, an infrared detection operation is performed. In some embodiments, an infrared body temperature sensor can be used to measure the pet's body temperature when close to the pet without touching it. At 3222, a visual inspection is performed. In some embodiments, based on the acquired image information, the pet's appearance information can also be determined, and the pet's health information can be determined based on this appearance information. For example, it can be determined from the image whether the pet's fur has become rough and dull, or whether the eyes are red and swollen, or have increased discharge. In some embodiments, the pet's historical health information can be obtained from the robot's configured local storage.

[0065] After obtaining the pet's status information (including behavioral, emotional, and health information), perform a behavior detection operation at step 323. Once the pet's status information is obtained, you can perform an operation at step 3231 to determine if the pet's behavior is abnormal. This involves judging whether the pet's behavior is abnormal based on any one of the following: behavioral, emotional, or health information.

[0066] If an anomaly is detected, an alert will be issued at step 3232 to notify the pet owner that the pet is in an abnormal state. For example, when the robot observes a pet (cat) scratching the sofa, it can record this behavior and determine if it is abnormal. If it is abnormal, the robot can issue an alert to notify the user.

[0067] For example, if a pet's fur becomes rough and dull, or its eyes become red and swollen with increased discharge, the robot can make a preliminary judgment that the pet may have a health problem. It can record these symptoms and notify the user. For another example, when it detects an abnormal rise or fall in the pet's body temperature, the robot can immediately issue an alarm. Furthermore, when it detects that the pet is circling back and forth and making hissing noises, it can determine that the pet's emotional state is abnormal and issue an alert.

[0068] If the pet's status information indicates that everything is normal, then return to... Figure 3 In step 330, the interactive decision-making and action operations are performed. For example, Figure 6 A schematic diagram illustrating example steps 600 of interactive decision-making and action according to some embodiments of the present disclosure is shown. Figure 6 As shown, at 331, a demand determination operation is performed. For example, in some embodiments, a machine learning model can determine the pet's demand information based on the pet's state information (including pet behavior information and pet emotional information). The machine learning model can be a large language model or a multimodal large language model. In some embodiments, the demand information of the pet can be inferred by the machine learning model using the prompt "Please combine the pet's current emotional and behavioral information to determine the pet's demand information" and the aforementioned obtained pet behavior and emotional information.

[0069] For example, if the pet's status information indicates that it is lingering near its food bowl, sniffing it, and showing signs of eagerness, the machine learning model can infer that the pet needs to eat. Similarly, if the pet's status information indicates that it is frequently licking its mouth or walking towards its water bowl, the machine learning model can infer that the pet needs to drink water.

[0070] For example, if the pet's status information indicates that the pet (such as a cat or dog) is actively approaching the robot, wagging its tail, or rubbing its head against the robot, the machine learning model can infer that the pet needs attention and play. As another example, if the pet's status information indicates that the pet (such as a cat or dog) is pacing back and forth or biting its own tail, the machine learning model can infer that the pet needs comfort and relaxation.

[0071] For example, the formula for calculating pet needs information based on pet status information using a machine learning model is shown below:

[0072] D=DEMAND_ANALYSIS_LLM(action,emotion)(1)

[0073] Here, "action" refers to behavioral information, and "emotion" refers to emotional information.

[0074] Once the pet's needs are determined, at step 332, the interaction strategy selection operation can be performed. In some embodiments, a machine learning model can determine the action strategy based on the pet's needs information, pet attributes, and a pet interaction strategy library. The machine learning model can be a Large Action Model (LAM). In some embodiments, the action strategy can be inferred by the Large Action Model by using the prompt "Please combine the pet's current needs information, pet attributes, and interaction strategies to determine the action strategy for the pet," and by combining the aforementioned pet needs information, pet attributes, and interaction strategies.

[0075] In some embodiments, a pet's attributes can include its breed, personality, and other characteristics. In some embodiments, the pet interaction strategy library predefines interaction strategies with pet attributes as the primary key. For example, the aforementioned requirement information indicates that the pet needs companionship and play. Therefore, when selecting a suitable interaction strategy for the pet, the first step is to determine the pet's attributes and then tailor the interaction strategy accordingly.

[0076] If the pet is a dog, and considering the pet interaction strategy library, the robot can determine action strategies such as playing fetch, tug-of-war, or patting its head. For example, some dogs are timid, so the robot can use a gentler approach when interacting with them, avoiding overly aggressive actions and noises. For dogs, interaction strategies might include fetch games, tug-of-war, and training commands. For instance, if the robot determines based on attributes that the dog needs to expend excess energy, it can choose a fetch game, having the dog chase and retrieve the ball. If the dog needs training, the robot can issue commands such as sit, lie down, and wait, and reinforce correct behaviors through reward mechanisms (such as giving treats).

[0077] If the pet is a cat, the appropriate action strategy could be to wave a cat toy, offer catnip toys, etc. For example, some cats are more independent, and the robot can provide interaction when appropriate, without disturbing them too frequently. For instance, if the cat shows interest in playing, the robot can gently wave a cat toy to attract its attention, encouraging it to chase and pounce. If the cat seems relaxed, the robot can emit a soft purring sound to make the cat feel comfortable and secure.

[0078] In some embodiments, pet interaction strategies are pre-defined in the robot's local storage module, and the pet interaction strategy library can be updated at quarterly or monthly intervals.

[0079] For example, the calculation formula for determining pet needs information using a machine learning model based on pet needs information, pet attributes, and an interaction strategy library is shown below:

[0080] S=GENERATION_LLM(D,P,A)(2)

[0081] Where D represents pet needs information, P represents the pet interaction strategy library, and A represents pet attributes.

[0082] Once the interaction strategy is determined, step 333 executes the action. The robot can then perform corresponding actions to soothe the pet. For example, the calculation formula for action execution can be as follows:

[0083] E = EXECUTE_LAM(S)(3)

[0084] Where E represents the action performed by the robot, and S represents the interaction strategy.

[0085] In some embodiments, the interaction strategy instructs the robot to move to the pet's location. The robot can then determine its movement path based on the map, the pet's location information, and its own location information obtained during the environmental perception and initialization phase, enabling it to move to the target location at step 3331. Once the robot has moved to the target location, at step 3332, it can interact with the pet based on the interaction strategy.

[0086] In some embodiments, assuming it is determined that interaction with the pet, refilling food or water is needed, the robot can use a path planning algorithm to calculate the optimal path to the target location. The path planning algorithm can consider obstacles in the home environment (such as furniture, walls, etc.) and the pet's position to select the shortest and safest path. For example, the formula for calculating the optimal path is as follows:

[0087] R=PATH_PLANNING_ALGORITHM(T,H,C)(4)

[0088] Where R is the optimal calculation path, T is the target location, H is the home environment information, and C is the pet's location.

[0089] Home environment information includes data such as the location and shape of obstacles such as furniture and walls. Formula (4) can be used to calculate the best path to the target location based on the target location (such as the location of the pet or where food and water are stored), the various obstacles in the home environment, and the actual location of the pet, using a path planning algorithm.

[0090] In some embodiments, the decision to trigger a path replanning operation can be based on factors such as the specific type of the unexpected event, the robot's current location, the target location, home environment information, and the pet's location. Unexpected event types could include, for example, someone suddenly walking by or a pet moving onto the robot's path. An exemplary formula for calculating the replanning start point is as follows:

[0091] Rtrig=PATH_REPLAN_TRIGGER(U,Pcur,T,H,C)(5)

[0092] Among them, the replanning trigger is Rtrig, the accident type is U, the current location is Pcur, the target location is T, the home environment information is H, and the pet location is C.

[0093] For example, if the pet is on the sofa in the living room, and the food and water are stored in the kitchen, the robot can calculate a path that avoids the coffee table in the middle of the living room and the trash can in the hallway, and quickly move to the kitchen. During the movement, the robot can use lidar and visual sensors to monitor changes in the surrounding environment in real time to avoid collisions. If an unexpected situation occurs (such as someone suddenly walking by or the pet moving into the robot's path), the robot can stop immediately and replan its path.

[0094] Upon reaching the pet, the robot can perform corresponding actions based on the selected interaction strategy. For example, in a fetch game with a dog, the robot can use robotic arm control technology to accurately throw the ball a certain distance and direction, allowing the dog to chase and retrieve it. After the dog retrieves the ball, the robot can detect the dog's position and movements using visual (e.g., camera or radar) and tactile perception (e.g., radar) technologies, and promptly reward the dog (e.g., with a treat or a praising sound). For instance, the formula for calculating the action of throwing the ball is as follows:

[0095] Ethrow=EXECUTE_THROWING_ACTION(Sdog,M)(6)

[0096] In this system, the ball-throwing action is called Ethrow, the dog's interaction strategy is Sdog, and the robotic arm's control parameters are M. The dog's interaction strategy determines the interaction methods, such as playing a ball-throwing game with the dog, while the robotic arm's control parameters are used to adjust the robotic arm's force, angle, and other control data.

[0097] For example, the formula for calculating rewards for dogs is as follows:

[0098] Rdog=REWARD_DECISION_DOG(Pdog,Adog,Rstrat)(7)

[0099] In this system, the dog's reward decision is represented by Rdog, the dog's positional awareness by Pdog, the dog's motion perception by Adog, and the reward policy by Rstrat. The dog's positional awareness can be acquired through visual and tactile perception technologies, as can its motion perception. The reward policy specifies what reward should be given to the dog under what circumstances.

[0100] When interacting with a cat, the robot can gently wave a cat toy, mimicking prey's movements to attract the cat's attention. The robot can adjust the speed and amplitude of the wave based on the cat's reaction to maintain its interest. Simultaneously, the robot can use visual perception technology to observe the cat's expressions and movements, determining whether the cat is tired or satisfied. If so, the robot can stop the interaction and allow the cat to rest. For example, the calculation formula for interacting with a cat is as follows, which represents the wave action of the cat toy based on the interaction strategy and the perceived information about the cat's reaction:

[0101] Edangle=EXECUTE_DANGLE_ACTION(Scat,Vcat,Tcat)(8)

[0102] In this system, the waving motion of the cat toy is represented by Edangle, the cat's interaction strategy is Scat, the visual perception feedback is Vcat, and the tactile perception feedback is Tcat. The cat's interaction strategy determines how to interact with the cat, such as gently waving the cat toy, while the visual and tactile perception feedback are used to obtain information such as the cat's reaction.

[0103] In some embodiments, the robot can also make a decision on whether to stop interacting based on its perception of the cat's facial expressions and movements, as well as strategies for determining whether the cat is tired or satisfied. For example, the formula for calculating whether to stop interaction is as follows:

[0104] Sint=STOP_INTERACTION_DECISION(Ecat,Acat,Fstrat,Sstrat)(9)

[0105] The interaction cessation decision is defined as Sint, cat facial expression perception as Ecat, cat motion perception as Acat, cat fatigue judgment strategy as Fstrat, and cat satisfaction judgment strategy as Sstrat. Cat facial expression perception and cat motion perception can be achieved using visual perception technology. The cat fatigue judgment strategy and the cat satisfaction judgment strategy define the criteria for judging whether a cat is fatigued or satisfied.

[0106] In some embodiments, when the robot determines that the pet needs to eat or drink, it can move to the food and water storage area. If the food is pre-packaged dry food, the robot can use its robotic arm to open the packaging and pour an appropriate amount of food into the pet's bowl. If the food is wet food or food that requires special handling, the robot can operate according to a pre-set program to ensure the freshness and hygiene of the food.

[0107] To add water, the robot can use a pipe or container connected to a water source to inject an appropriate amount of water into the pet's water bowl. During the process, visual perception technology helps the robot confirm that food and water have been added to the correct locations, preventing spills. Simultaneously, the robot can monitor the water level and food quantity to ensure the pet has a sufficient supply of food and water. In some embodiments, the water addition operation can be performed based on the water source connection method, water container control parameters, and water level control strategy. For example, the formula for calculating water addition is as follows:

[0108] Ewater=EXECUTE_WATER_ADDITION(Wcon,Ccon,Lstrat)(10)

[0109] The water addition operation is designated as Ewater, the water source connection method as Wcon, the water container control parameters as Ccon, and the water level control strategy as Lstrat. The water source connection method refers to whether the robot obtains water through a pipe or a container connected to a water source. The water container control parameters indicate the data related to controlling the water pouring action, and the water level control strategy indicates the appropriate water level to be added to the pet's water bowl.

[0110] In some embodiments, a decision on whether to ensure a pet has sufficient food and water supply can be made based on perceptions of food and water provision, monitoring data on water levels and food quantities, and reference to supply standards. For example, the decision on food and water supply is calculated using the following formula:

[0111] Msup=SUPPLY_MONITORING_DECISION(Vsup,Lsup,Fsup,Ssup)(11)

[0112] The supply monitoring decision is Msup, visual perception data is Vsup, water level monitoring data is Lsup, food quantity monitoring data is Fsup, and supply standard is Ssup. Visual perception data is used to confirm whether food and water are added to the correct locations. Water level monitoring data and food quantity monitoring data are used to monitor the water level in the pet's water bowl and the food quantity in the pet's food bowl, respectively. The supply standard specifies the concrete criteria for ensuring that the pet has a sufficient supply of food and water.

[0113] In this way, pet companion robots can care for pets when humans are unable to be with them, providing playtime, water, food, and other forms of companionship, thus improving the pets' quality of life. At the same time, they also make it easier and more convenient for humans to care for their pets, enhancing the human experience of using the robots.

[0114] Back Figure 3 At 340, data recording and feedback operations are performed. To enhance the comfort, feasibility, and stability of the robot-pet interaction, detailed data such as the pet's status, the services provided, and the pet's reactions can be recorded, and the robot can be adjusted based on these records or feedback from the pet owner. For example, Figure 7 A schematic diagram of example steps 700 for data logging and feedback according to some embodiments of the present disclosure is shown.

[0115] like Figure 7 As shown, in some implementations, at 341, a recording operation is performed. That is, the robot can record in detail the pet's observed state 3411, the service data provided 3412, and the pet's reaction data 3413 each time.

[0116] In some embodiments, pet status observation data includes the pet's location, posture, behavior, sounds, physiological indicators, etc. For example, the robot can record the pet's specific location at a certain point in time (such as on a living room sofa), its posture at that time (such as lying down, sitting, etc.), its ongoing behavior (such as playing, sleeping, etc.), the type of sounds it makes (such as barking, purring, etc.), and the measured physiological indicators (such as body temperature, heart rate, etc.).

[0117] In some embodiments, service data includes interaction methods, food and water amounts, training time, etc. For example, the robot can record the type of interactive game played with the pet (such as a ball-throwing game), the duration of the interaction, the type and amount of food added, the time and amount of water added, and the specific content and duration of training.

[0118] In some embodiments, response data may also be recorded, including the pet’s response to the interaction (e.g., active participation, lack of interest, etc.), acceptance of food and water (e.g., eating with relish, refusing to eat, etc.), and mastery of training (e.g., correctly following instructions, needing more practice, etc.).

[0119] This data can be stored in the robot's local database and can also be periodically uploaded to a cloud server for long-term analysis and tracking. In some embodiments, users can view this data at any time through a mobile application to understand their pet's daily activities and health status.

[0120] As mentioned earlier, in some embodiments, the robot can also use Wi-Fi or Bluetooth communication technology to promptly send important information to the user's mobile phone or other smart devices. For example, if the robot detects abnormal behavior in a pet (such as prolonged fasting, excessive scratching, etc.), insufficient food or water, or suspected health problems (such as abnormal body temperature), it can immediately send a notification to the user. The notification can be sent in the form of push notifications, SMS messages, or emails to ensure that the user is informed in a timely manner.

[0121] Alternatively, users can view detailed pet data and the robot's work logs through a mobile application. The application displays information such as the pet's activity trajectory, changes in food and water intake, and interaction patterns in intuitive charts and reports, providing a clear overview for users. Users can also remotely control the robot to perform actions such as increasing food intake, changing interaction methods, and activating emergency rescue mode. If users have any questions or suggestions regarding the robot's operation, they can also communicate and provide feedback to the robot's developers through the application.

[0122] Continue to refer to Figure 3 At level 350, continuous learning and optimization are performed. As mentioned earlier, the machine learning model can determine the pet's needs and action strategies. Based on these strategies, the robot can then perform actions for the pet. Therefore, after the robot completes its actions for the pet, it can obtain the pet's reaction information and update the machine learning model accordingly, thereby providing the pet with a better interactive experience.

[0123] For example, Figure 8 A schematic diagram illustrating example step 800 of continuous learning and optimization according to some embodiments of the present disclosure is shown. Reference Figure 8 In step 351, the operation of obtaining the pet's reaction is performed. Then, in step 353, a comparison operation is performed. This involves comparing the pet's reaction after the action with the labeled reaction 352 to obtain a comparison result 354. This comparison result 354 can then be used to adjust the second machine learning model 355 and the first machine learning model 356 in reverse.

[0124] In some embodiments, strategies in the pet interaction strategy library can also be evaluated and optimized based on the pet's reactions to different interaction methods. At 357, a positive judgment operation is performed. If the pet's reaction to the interaction strategy is positive, then at 358, an operation to increase the weight of that interaction strategy can be performed to improve the pet's interaction experience in subsequent processes. If the pet's reaction to the interaction strategy is negative, then at 359, an operation to decrease the weight of that interaction strategy can be performed.

[0125] Alternatively, the robot can periodically send newly collected pet data to a cloud server, where machine learning algorithms can be used to update the first machine learning model, the second machine learning model, and the interaction strategy library. The cloud server can analyze and process large amounts of pet data, extracting valuable information and patterns. For example, by clustering new behavioral data, new behavioral patterns or preferences of pets can be discovered. If it is found that a pet frequently performs a certain behavior at a specific time and place, the robot can adjust its observation and interaction strategies based on this pattern to better meet the pet's needs.

[0126] In addition to updates to the cloud server, the robot can also learn and adjust locally in real time. For example, when the robot finds that a certain interaction method is not working in the current situation, it can immediately try other methods and adjust based on the pet's reaction. This real-time learning capability allows the robot to better adapt to different scenarios and the changing needs of the pet.

[0127] As pets age, their health changes, and their behavior evolves, robots can personalize their management and training plans. For example, as a pet grows, its dietary needs and activity levels can change accordingly. The robot can adjust the type and amount of food, as well as the intensity of interaction and training, based on the pet's age and weight. If a dog grows from puppy to adult, the robot can increase the amount of food and provide more challenging training games to meet the intellectual and physical needs of an adult dog.

[0128] If a pet is sick or injured, the robot can adjust its care methods according to its specific needs. For example, if a pet has a fracture, the robot can be more careful when moving and interacting with the pet to avoid causing further injury. If a pet requires special diet or medication, the robot can strictly manage the treatment according to the veterinarian's advice, ensuring that the pet takes its medication on time and eats the appropriate food.

[0129] Furthermore, robots can be personalized based on a pet's behavioral habits. For example, if a pet is accustomed to resting at specific times and places, the robot can avoid disturbing the pet during those times and provide more interaction and attention at other times. If a pet has a special reaction to a certain sound or color, the robot can utilize these characteristics during interactions to increase the pet's engagement and satisfaction.

[0130] Through this continuous adjustment and optimization, pets can have a better companionship experience, thereby enhancing the user experience of humans using robots.

[0131] The above Figures 3-8This paper describes an example of using robots to interact with pets, and also describes ways to improve the effectiveness of this example based on the pet's interaction experience. Figures 3-8 As described above, pet companion robots can care for pets when humans are unable to be with them, improving the pets' quality of life. At the same time, they also make it easier and more convenient for humans to care for their pets, enhancing the user experience of using the robots.

[0132] Figure 9 A block diagram of a task execution apparatus 900 according to some embodiments of the present disclosure is shown. Figure 9 As shown, device 900 includes a pet status information acquisition module 902, configured to acquire pet status information, including pet behavior information and pet emotional information. Device 900 includes a pet demand information determination module 904, configured to determine pet demand information based on the pet status information using a first machine learning model. Device 900 includes an action strategy determination module 906, configured to determine an action strategy based on the pet demand information, pet attributes, and a pet interaction strategy library using a second machine learning model. Device 900 includes a control module 908, configured to control an intelligent robot to perform actions for the pet based on the action strategy.

[0133] Figure 10 A block diagram of a device 1000 capable of implementing various embodiments of the present disclosure is shown. For example... Figure 10 As shown, device 1000 includes a central processing unit (CPU) and / or a graphics processing unit (GPU) 1001, which can perform various appropriate actions and processes according to computer program instructions stored in read-only memory (ROM) 1002 or loaded from storage unit 1008 into random access memory (RAM) 1003. Various programs and data required for the operation of device 1000 can also be stored in RAM 1003. The CPU / GPU 1001, ROM 1002, and RAM 1003 are interconnected via bus 1004. Input / output (I / O) interface 1005 is also connected to bus 1004. Although not shown in... Figure 10 As shown, device 1000 may also include a coprocessor.

[0134] Multiple components in device 1000 are connected to I / O interface 1005, including: input unit 1006, such as keyboard, mouse, etc.; output unit 1007, such as various types of monitors, speakers, etc.; storage unit 1008, such as disk, optical disk, etc.; and communication unit 1009, such as network card, modem, wireless transceiver, etc. Communication unit 1009 allows device 1000 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0135] The various methods or processes described above can be executed by CPU / GPU 1001. For example, in some embodiments, the methods may be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage unit 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on device 1000 via ROM 1002 and / or communication unit 1009. When the computer program is loaded into RAM 1003 and executed by CPU / GPU 1001, one or more steps or actions in the methods or processes described above can be performed.

[0136] In some embodiments, the methods and processes described above can be implemented as a computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of this disclosure.

[0137] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example, but not limited to, electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0138] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network (LAN), a wide area network (WAN), and / or a wireless network, to an external computer or external storage device. The network may include copper cables, fiber optic cables, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0139] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​and conventional procedural programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0140] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processing unit of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner. Thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0141] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0142] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of devices, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0143] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

[0144] The following are some example implementations of this disclosure.

[0145] Example 1. A task execution method, characterized in that it is applicable to intelligent robots and includes:

[0146] Obtain pet status information, which includes pet behavior information and pet emotional information;

[0147] The first machine learning model determines the pet's needs information based on the pet's status information;

[0148] The second machine learning model determines the action strategy based on pet needs information, pet attributes, and a pet interaction strategy library; and

[0149] Based on the aforementioned action strategy, the intelligent robot is controlled to perform actions targeting the pet.

[0150] Example 2. According to the method described in Example 1, obtaining pet status information includes:

[0151] Based on a first preset time interval and a map, acquire images and sounds of the pet;

[0152] Based on the image and the sound, determine the pet's behavioral information; and

[0153] Based on the pet's behavioral information, the pet's emotional information is determined.

[0154] Example 3. According to any one of Examples 1-2, the pet status information further includes pet health information, which includes first pet health information and second pet health information, and further includes:

[0155] Based on the image, determine the pet's appearance information; and

[0156] Based on the appearance information, the second pet health information is determined, and the first pet health information indicates the pet's historical health information.

[0157] Example 4. The method according to any one of Examples 1-3 further includes:

[0158] Record and determine whether any item in the pet status information is abnormal, the pet status information including the behavioral information, the pet's emotional information, and the health information; and

[0159] An abnormality warning is issued in response to the detection of abnormal pet status information.

[0160] Example 5. The method according to any one of Examples 1-4, wherein determining the pet's demand information by the first machine learning model based on the pet's state information includes:

[0161] The first machine learning model determines the pet's needs information based on the first preset prompt, the pet's behavior information, and the pet's emotional information.

[0162] Example 6. According to any one of Examples 1-5, the pet interaction strategy library includes multiple pet interaction strategies corresponding to pet attributes, and further includes:

[0163] The pet interaction strategy library is retrieved based on a second preset time interval.

[0164] Example 7. According to any one of Examples 1-6, the pet interaction strategy library includes a first strategy library and a second strategy library, wherein the action strategy determined by the second machine learning model based on pet demand information, pet attributes, and the pet interaction strategy library includes:

[0165] Determine whether the pet's attribute is the primary or secondary attribute; and

[0166] In response to the detection of the first attribute, the second machine learning model determines a first action strategy based on a second preset prompt, the pet's needs, the pet's attributes, and the first strategy library; or

[0167] In response to the detection of the second attribute, the second machine learning model determines the second action strategy based on the second preset prompt, the pet's needs, the pet's attributes, and the second strategy library.

[0168] Example 8. The method according to any one of Examples 1-7 further includes:

[0169] Obtain environmental information about the pet's location, including a map of the environment; and

[0170] Based on the environmental information, the location information of the pet and the intelligent robot is determined.

[0171] Example 9. The method according to any one of Examples 1-8, wherein controlling the intelligent robot to perform actions for the pet based on the action strategy includes:

[0172] In response to the detection that the action strategy includes a movement strategy, a movement path is determined based on the map, the pet's location information, and the intelligent robot's location information; and

[0173] Based on the movement path, the intelligent robot is controlled to move to the pet.

[0174] Example 10. According to any one of Examples 1-9, determining the movement path based on the map, the pet's location information, and the intelligent robot's location information includes:

[0175] Based on the movement path and the environmental information, a collision strategy is determined; and

[0176] The movement path is adjusted based on the collision strategy.

[0177] Example 11. The method according to any one of Examples 1-10, wherein the action on the pet includes at least one of feeding action, watering action, and soothing action.

[0178] Example 12. The method according to any one of Examples 1-11 further includes:

[0179] After the intelligent robot performs an action on the pet, it acquires the pet's reaction information; and

[0180] Based on the reaction information, the first machine learning model and the second machine learning model are adjusted in reverse.

[0181] Example 13. The method according to any one of Examples 1-12, wherein the first machine learning model and the second machine learning model are at least one of a large action model or a large language model.

[0182] Example 14. According to any one of Examples 1-13, the first machine learning model and the second machine learning model are configured in the same machine learning model.

[0183] Example 15. A task execution device, characterized in that it is suitable for intelligent robots and comprises:

[0184] The pet status information acquisition module is configured to acquire pet status information, which includes pet behavior information and pet emotional information.

[0185] The pet demand information determination module is configured to determine pet demand information based on the pet status information by a first machine learning model;

[0186] The action strategy determination module is configured to determine the action strategy using a second machine learning model based on pet demand information, pet attributes, and a pet interaction strategy library; and

[0187] The control module is configured to control the intelligent robot to perform actions on the pet based on the action strategy.

[0188] Example 16. The apparatus according to any one of Examples 15, wherein the pet status information acquisition module comprises:

[0189] The first acquisition module is configured to acquire images and sounds of the pet based on a first preset time interval and a map.

[0190] The first determining module is configured to determine the pet's behavior information based on the image and the sound; and

[0191] The second determining module is configured to determine the pet's emotional information based on the pet's behavior information.

[0192] Example 17. The apparatus according to any one of Examples 15-16, wherein the pet status information further includes pet health information, the pet health information including first pet health information and second pet health information, and further includes:

[0193] The third determining module is configured to determine the pet's appearance information based on the image; and

[0194] The fourth determining module is configured to determine the second pet health information based on the appearance information, wherein the first pet health information indicates the pet's historical health information.

[0195] Example 18. The apparatus according to any one of Examples 15-17 further includes:

[0196] A recording module is configured to record and determine whether any item in the pet status information is abnormal, the pet status information including the behavioral information, the pet's emotional information, and the health information; and

[0197] The issuing module is configured to issue an abnormality warning in response to detecting an abnormality in the pet's status information.

[0198] Example 19. The apparatus according to any one of Examples 15-18, wherein the pet demand information determining module comprises:

[0199] The fifth determining module is configured to determine the pet's demand information based on the first machine learning model using a first preset prompt, the pet's behavior information, and the pet's emotional information.

[0200] Example 20. The apparatus according to any one of Examples 15-19, wherein the pet interaction strategy library includes a plurality of pet interaction strategies corresponding to pet attributes, and further includes:

[0201] The second acquisition module is configured to acquire the pet interaction strategy library based on a second preset time interval.

[0202] Example 21. The apparatus according to any one of Examples 15-20, wherein the pet interaction strategy library includes a first strategy library and a second strategy library, wherein the action strategy determination module includes:

[0203] The sixth determining module is configured to determine whether a pet's attribute is the primary attribute or the secondary attribute; and

[0204] The seventh determining module is configured to, in response to the detection of a first attribute, determine a first action strategy by the second machine learning model based on a second preset prompt, the pet's needs, the pet's attributes, and the first strategy library; or

[0205] The eighth determining module is configured to, in response to the detection of a second attribute, determine a second action strategy by the second machine learning model based on the second preset prompt, the pet's needs, the pet's attributes, and the second strategy library.

[0206] Example 22. The apparatus according to any one of Examples 15-21 further includes:

[0207] The third acquisition module is configured to acquire environmental information about the pet's location, including a map of the environment; and

[0208] The ninth determining module is configured to determine the pet's location information and the intelligent robot's location information based on the environmental information.

[0209] Example 23. The apparatus according to any one of Examples 15-22, wherein the control module comprises:

[0210] The ninth determining module is configured to, in response to detecting that the action strategy includes a movement strategy, determine a movement path based on the map, the pet's location information, and the intelligent robot's location information; and

[0211] The movement control module is configured to control the intelligent robot to move to the pet based on the movement path.

[0212] Example 24. The apparatus according to any one of Examples 15-23, wherein the ninth determining module comprises:

[0213] The tenth determining module is configured to determine a collision strategy based on the movement path and the environmental information; and

[0214] The adjustment module is configured to adjust the movement path based on the collision strategy.

[0215] Example 25. The apparatus according to any one of Examples 15-24, wherein the action toward the pet includes at least one of feeding, watering, and soothing actions.

[0216] Example 26. The apparatus according to any one of Examples 15-25 further includes:

[0217] The fourth acquisition module is configured to acquire the pet's reaction information after the intelligent robot performs an action on the pet; and

[0218] The reverse adjustment module is configured to reverse adjust the first machine learning model and the second machine learning model based on the reaction information.

[0219] Example 27. The apparatus according to any one of Examples 15-26, wherein the first machine learning model and the second machine learning model are at least one of a large action model or a large language model.

[0220] Example 28. The apparatus according to any one of Examples 15-27, wherein the first machine learning model and the second machine learning model are configured in the same machine learning model.

[0221] Example 29. An electronic device comprising:

[0222] Processor; and

[0223] A memory coupled to the processor, the memory having instructions stored therein, which, when executed by the processor, cause the electronic device to perform actions, the actions including:

[0224] Obtain pet status information, which includes pet behavior information and pet emotional information;

[0225] The first machine learning model determines the pet's needs information based on the pet's status information;

[0226] The second machine learning model determines the action strategy based on pet needs information, pet attributes, and a pet interaction strategy library; and

[0227] Based on the aforementioned action strategy, the intelligent robot is controlled to perform actions targeting the pet.

[0228] Example 30. The electronic device according to Example 29, wherein obtaining pet status information includes:

[0229] Based on a first preset time interval and a map, acquire images and sounds of the pet;

[0230] Based on the image and the sound, determine the pet's behavioral information; and

[0231] Based on the pet's behavioral information, the pet's emotional information is determined.

[0232] Example 31. The electronic device according to any one of Examples 29-30, wherein the pet status information further includes pet health information, the pet health information including first pet health information and second pet health information, and further includes:

[0233] Based on the image, determine the pet's appearance information; and

[0234] Based on the appearance information, the second pet health information is determined, and the first pet health information indicates the pet's historical health information.

[0235] Example 32. The electronic device according to any one of Examples 29-31 further includes:

[0236] Record and determine whether any item in the pet status information is abnormal, the pet status information including the behavioral information, the pet's emotional information, and the health information; and

[0237] An abnormality warning is issued in response to the detection of abnormal pet status information.

[0238] Example 33. An electronic device according to any one of Examples 29-32, wherein determining the pet's needs information by a first machine learning model based on the pet's state information includes:

[0239] The first machine learning model determines the pet's needs information based on the first preset prompt, the pet's behavior information, and the pet's emotional information.

[0240] Example 34. An electronic device according to any one of Examples 29-33, wherein the pet interaction strategy library includes multiple pet interaction strategies corresponding to pet attributes, and further includes:

[0241] The pet interaction strategy library is retrieved based on a second preset time interval.

[0242] Example 35. An electronic device according to any one of Examples 29-34, wherein the pet interaction strategy library includes a first strategy library and a second strategy library, wherein the action strategy determined by the second machine learning model based on pet demand information, pet attributes, and the pet interaction strategy library includes:

[0243] Determine whether the pet's attribute is the primary or secondary attribute; and

[0244] In response to the detection of the first attribute, the second machine learning model determines a first action strategy based on a second preset prompt, the pet's needs, the pet's attributes, and the first strategy library; or

[0245] In response to the detection of the second attribute, the second machine learning model determines the second action strategy based on the second preset prompt, the pet's needs, the pet's attributes, and the second strategy library.

[0246] Example 36. The electronic device according to any one of Examples 29-35 further includes:

[0247] Obtain environmental information about the pet's location, including a map of the environment; and

[0248] Based on the environmental information, the location information of the pet and the intelligent robot is determined.

[0249] Example 37. An electronic device according to any one of Examples 29-36, wherein controlling the intelligent robot to perform actions toward the pet based on the action strategy includes:

[0250] In response to the detection that the action strategy includes a movement strategy, a movement path is determined based on the map, the pet's location information, and the intelligent robot's location information; and

[0251] Based on the movement path, the intelligent robot is controlled to move to the pet.

[0252] Example 38. According to any one of Examples 29-37, the electronic device, wherein determining the movement path based on the map, the pet's location information, and the intelligent robot's location information includes:

[0253] Based on the movement path and the environmental information, a collision strategy is determined; and

[0254] The movement path is adjusted based on the collision strategy.

[0255] Example 39. The electronic device according to any one of Examples 29-38, wherein the action for the pet includes at least one of feeding action, watering action, and soothing action.

[0256] Example 40. The electronic device according to any one of Examples 29-39 further includes:

[0257] After the intelligent robot performs an action on the pet, it acquires the pet's reaction information; and

[0258] Based on the reaction information, the first machine learning model and the second machine learning model are adjusted in reverse.

[0259] Example 41. An electronic device according to any one of Examples 29-40, wherein the first machine learning model and the second machine learning model are at least one of a large action model or a large language model.

[0260] Example 42. An electronic device according to any one of Examples 29-41, wherein the first machine learning model and the second machine learning model are configured in the same machine learning model.

[0261] Example 43. A computer-readable storage medium having stored thereon computer-executable instructions, wherein the computer-executable instructions are executed by a processor to implement an electronic device according to any one of Examples 1 to 14.

[0262] Example 44. A computer program product tangibly stored on a computer-readable medium and comprising computer-executable instructions that, when executed by a device, cause the device to perform an electronic device according to any one of Examples 1 to 14.

[0263] Although this disclosure has been described using language specific to structural features and / or the logical operation of electronic devices, it should be understood that the subject matter defined in any of the appended examples is not necessarily limited to the specific features or operations described above. Rather, the specific features and operations described above are merely exemplary forms of implementing the claims.

Claims

1. A task execution method, characterized in that, Suitable for intelligent robots, including: Obtain pet status information, which includes pet behavior information and pet emotional information; The first machine learning model determines the pet's needs information based on the pet's status information; The second machine learning model determines the action strategy based on pet needs information, pet attributes, and a pet interaction strategy library; and Based on the aforementioned action strategy, the intelligent robot is controlled to perform actions targeting the pet.

2. The method according to claim 1, wherein obtaining pet status information includes: Based on a first preset time interval and a map, acquire images and sounds of the pet; Based on the image and the sound, the pet's behavioral information is determined; as well as Based on the pet's behavioral information, the pet's emotional information is determined.

3. The method according to claim 2, wherein the pet status information further includes pet health information, the pet health information including first pet health information and second pet health information, and further includes: Based on the image, determine the pet's appearance information; as well as Based on the appearance information, the second pet health information is determined, and the first pet health information indicates the pet's historical health information.

4. The method according to claim 3, further comprising: Record and determine whether any item in the pet status information is abnormal, the pet status information including the behavioral information, the pet emotional information, and the health information; as well as An abnormality warning is issued in response to the detection of abnormal pet status information.

5. The method according to claim 1, wherein determining the pet's needs information by the first machine learning model based on the pet's state information includes: The first machine learning model determines the pet's needs information based on the first preset prompt, the pet's behavior information, and the pet's emotional information.

6. The method according to claim 1, wherein the pet interaction strategy library includes multiple pet interaction strategies corresponding to pet attributes, and further includes: The pet interaction strategy library is retrieved based on a second preset time interval.

7. The method according to claim 6, wherein the pet interaction strategy library includes a first strategy library and a second strategy library, wherein the second machine learning model determines the action strategy based on pet demand information, pet attributes, and the pet interaction strategy library, including: Determine whether the pet's attribute is the primary attribute or the secondary attribute; as well as In response to the detection of the first attribute, the second machine learning model determines the first action strategy based on the second preset prompt, the pet's needs, the pet's attributes, and the first strategy library; or In response to the detection of the second attribute, the second machine learning model determines the second action strategy based on the second preset prompt, the pet's needs, the pet's attributes, and the second strategy library.

8. The method according to claim 2, further comprising: Obtain environmental information about the pet's location, including a map of the environment; as well as Based on the environmental information, the location information of the pet and the intelligent robot is determined.

9. The method of claim 8, wherein controlling the intelligent robot to perform actions on the pet based on the action strategy comprises: In response to the detection that the action strategy includes a movement strategy, a movement path is determined based on the map, the pet's location information, and the intelligent robot's location information; as well as Based on the movement path, the intelligent robot is controlled to move to the pet.

10. The method according to claim 9, wherein determining the movement path based on the map, the pet's location information, and the intelligent robot's location information includes: Based on the movement path and the environmental information, a collision strategy is determined; as well as The movement path is adjusted based on the collision strategy.

11. The method according to claim 1, wherein the actions performed on the pet include at least one of feeding, watering, and comforting actions.

12. The method according to claim 1, further comprising: After the intelligent robot performs an action on the pet, it acquires the pet's reaction information; as well as Based on the reaction information, the first machine learning model and the second machine learning model are adjusted in reverse.

13. The method according to claim 1, wherein the first machine learning model and the second machine learning model are at least one of a large action model or a large language model.

14. The method of claim 13, wherein the first machine learning model and the second machine learning model are configured in the same machine learning model.

15. A task execution device, characterized in that, Suitable for intelligent robots, including: The pet status information acquisition module is configured to acquire pet status information, which includes pet behavior information and pet emotional information. The pet demand information determination module is configured to determine pet demand information based on the pet status information by a first machine learning model; The action strategy determination module is configured to determine the action strategy using a second machine learning model based on pet demand information, pet attributes, and a pet interaction strategy library; and The control module is configured to control the intelligent robot to perform actions on the pet based on the action strategy.

16. An electronic device comprising: processor; as well as A memory coupled to the processor, the memory having instructions stored therein, which, when executed by the processor, cause the electronic device to perform the method according to any one of claims 1 to 14.

17. A computer program product tangibly stored on a non-transitory computer-readable medium and comprising machine-executable instructions that, when executed, cause a machine to perform the method according to any one of claims 1 to 14.

Images