Medical service system and method based on humanoid robot

Abstract

The invention relates to the technical field of medical equipment, and discloses a medical service system and method based on a humanoid robot. In the application, the service system comprises a control platform and a humanoid robot responding to the control platform; wherein a control system in the control platform is in communication connection with the robot, and sends a ward round instruction and / or a medicine delivery instruction to the humanoid robot; after the control module of the humanoid robot receives the instruction, a preset instruction set is generated, and the preset instruction set at least comprises a moving instruction, a dialogue instruction and a limb action instruction; the execution mechanism is connected with the control module and used for executing the medical service action corresponding to the preset instruction set. Through a cooperative architecture of a control platform and a humanoid robot, conventional medical tasks arranged by medical staff are realized, service experience is improved through humanoid interaction, and a solution with both efficiency and practicability is provided for intelligent medical treatment.

Description

Technical Field

[0001] The present application relates to the field of medical equipment technology, and in particular to a medical service system and method based on a humanoid robot. Background Art

[0002] Driven by the wave of smart city development, the traditional medical service system and its supporting public medical facilities are no longer able to meet the growing and diverse medical needs of modern people. Smart healthcare, with its intelligent and efficient features, has become an inevitable trend in the transformation and upgrading of the medical industry. In this context, the intelligent innovation of medical facilities and equipment has become a key path to improving the quality of medical services.

[0003] While medical staff's daily ward rounds are crucial for ensuring patient outcomes, they present numerous challenges. Direct contact with patients with infectious diseases or critical illnesses not only exposes medical staff to a high-risk infection environment but also significantly increases the risk of cross-infection. Furthermore, in the often overcrowded environment of large hospitals, medical staff often face multiple tasks simultaneously, fragmenting their attention and preventing them from providing timely and attentive care to every patient. How to effectively reduce the workload and occupational risks of medical staff while ensuring quality care has become a pressing challenge for the healthcare industry.

[0004] With the rapid development of artificial intelligence (AI), intelligent conversational capabilities have become widely used in various devices, including smart robots and smartphones. However, in medical ward rounds, current smart devices on the market still have significant shortcomings. Existing non-humanoid ward round robots rely on nurse station servers and communicate with patients via remote video. These robots lack autonomous movement, making natural interaction difficult, and lack the flexibility to perform physical movements, making them incapable of meeting the complex demands of clinical ward rounds. Summary of the Invention

[0005] The purpose of the implementation method of this application is to provide a medical service system based on a humanoid robot, which realizes automated services in medical scenarios such as ward rounds and medicine delivery, and solves the problems of low efficiency and lack of autonomy in related technologies.

[0006] To solve the above technical problems, a medical service system based on a humanoid robot includes a control platform and a humanoid robot responsive to the control platform; wherein the control platform includes at least a control system, and the control system is communicatively connected to the humanoid robot; the control system is used to send ward round instructions and / or medicine delivery instructions to the humanoid robot; the humanoid robot includes at least a control module and an actuator that are interconnected; wherein the control module is communicatively connected to the control system, and is used to generate a preset instruction set in response to the ward round instructions and / or medicine delivery instructions sent by the control system, wherein the preset instruction set includes at least movement instructions, dialogue instructions, limb movement instructions, and configuration instructions; the actuator is connected to the control module, and is used to execute the medical service actions corresponding to the preset instruction set.

[0007] An embodiment of the present application also provides a medical service method based on a humanoid robot, adopting the above-mentioned medical service system, and the method includes: the control platform sends a ward round instruction and / or a medicine delivery instruction to the humanoid robot; the humanoid robot responds to the ward round instruction and / or the medicine delivery instruction, and performs the medical service action corresponding to the ward round instruction and / or the medicine delivery instruction.

[0008] Compared with the related art, the embodiment of the present application provides a medical service system and method based on a humanoid robot, which includes a control platform and a humanoid robot. The control platform serves as a central system, which receives instructions for carrying out medical service-related tasks, such as ward round instructions, medicine delivery instructions, etc., through the control system within the platform, and sends control instructions to the humanoid robot. The control module of the humanoid robot communicates with the control system and generates a preset instruction set after receiving the instructions. The movement instructions in the instruction set enable the robot to navigate to the target position, the dialogue instructions enable the robot to interact with medical staff or patients, and the body movement instructions control the robot to perform corresponding actions. The robot also includes an actuator for performing corresponding medical service actions according to the instructions generated by the control module (such as completing ward inspections according to ward round instructions and delivering medicines according to medicine delivery instructions). In this way, through the collaborative architecture of "control platform + humanoid robot", routine medical tasks assigned by medical staff are realized, and the service experience is improved through human-like interaction, providing a solution for smart medical care that is both efficient and practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings. These exemplifications do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements. Unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0010] Figure 1Schematic diagram of the structure of the medical service system of the humanoid robot provided in the embodiment of the present application;

[0011] Figure 2 is a structural diagram of a control system provided in another embodiment of the present application;

[0012] Figure 3 is an architectural diagram of a control module of a humanoid robot provided in another embodiment of the present application;

[0013] Figure 4 is a structural schematic diagram of an actuator of a humanoid robot provided in another embodiment of the present application;

[0014] Figure 5 is a flow chart of a medical service method based on a humanoid robot provided in another embodiment of the present application;

[0015] Figure 6 This is another flowchart of a humanoid robot-based medical service method provided in another embodiment of the present application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, each embodiment of the present application will be described in detail below with reference to the accompanying drawings. However, it will be understood by those skilled in the art that in each embodiment of the present application, many technical details are proposed to enable the reader to better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions claimed in the present application can be implemented. The division of the following embodiments is for convenience of description and should not constitute any limitation on the specific implementation of the present application. The various embodiments can be combined and referenced with each other under the premise of no contradiction.

[0017] Analysis reveals that current smart devices on the market still have significant shortcomings in medical ward round scenarios. Existing non-humanoid ward round robots rely on nurse station servers and communicate with patients via remote video. They lack autonomous ward round capabilities and are unable to automatically generate medical records. They also lack the support of large-scale model technology, making natural interaction difficult. Furthermore, they lack flexible body movements, making them incapable of meeting the complex demands of clinical ward rounds.

[0018] Based on this, an embodiment of the present application provides a medical service system based on a humanoid robot, which includes a control platform and a humanoid robot. The system issues control instructions through the control platform, and the humanoid robot executes the control instructions sent by the control platform to perform corresponding medical service actions. In this way, this embodiment constructs a collaborative architecture of "control platform + humanoid robot", realizes routine medical tasks assigned by medical staff, and improves the service experience through human-like interaction, providing a solution for smart medical care that is both efficient and practical.

[0019] To make the purpose, technical solutions, and advantages of the embodiments of the present application more clear, each embodiment of the present application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will appreciate that many technical details are provided in each embodiment of the present application to help readers better understand the present application. However, even without these technical details and various variations and modifications based on the following embodiments, the technical solutions claimed in the present application can still be implemented.

[0020] The following embodiments are divided for the convenience of description and should not constitute any limitation on the specific implementation of the present application. The various embodiments can be combined with each other and referenced to each other without contradiction.

[0021] On the one hand, the embodiments of the present application provide a medical service system based on a humanoid robot. In some embodiments, such as Figure 1 As shown, the system includes a control system and a humanoid robot in communication with the control system. The control platform includes at least the control system, which is in communication with the humanoid robot. The control system is configured to send ward round instructions and / or medication delivery instructions to the humanoid robot. The humanoid robot includes at least a control module and an actuator that are interconnected. The control module is in communication with the control system and is configured to generate a preset instruction set in response to the ward round instructions and / or medication delivery instructions sent by the control system. The preset instruction set includes at least movement instructions, dialogue instructions, body movement instructions, and configuration instructions. The actuator is connected to the control module and is configured to execute the medical service actions corresponding to the preset instruction set.

[0022] Specifically, the system includes a control platform and a humanoid robot that responds to the control platform. The control system of the control platform sends the user's input instructions for ward rounds, medicine delivery, etc. to the humanoid robot. The control module of the humanoid robot (hereinafter referred to as "robot") responds to the instructions to generate a preset instruction set containing instructions such as movement, dialogue, and body movements, and the actuator performs the corresponding actions. This system solves the problems of low efficiency and lack of autonomy in related technologies and realizes automated services in medical scenarios.

[0023] Among them, users can be doctors, nurses and other staff directly involved in medical services. Users can send ward rounds instructions through the control platform in the medical service system of this application, such as having robots replace humans to visit wards to check on the status of patients (especially suitable for patients with high infection risks). They can also issue medicine delivery instructions, and robots will deliver medicines to designated beds according to doctor's orders, reducing the workload of nurses.

[0024] Among them, the control platform can be the hospital's information management platform or backend system, which can batch dispatch multiple robots to perform tasks (such as arranging multiple robots to be responsible for ward rounds in different wards at the same time), and obtain the robot's working status data (such as task completion rate, fault report) from the control system for hospital operation management. In the above-mentioned interaction between the user and the control platform, the user can input text instructions through the control system's operating interface (such as a computer or mobile APP), or issue commands through voice interaction. For example, when the control instruction is a ward round instruction, the doctor in the office says to the control system through voice "Make ward rounds for the patient in Ward 3" → the control system sends the ward round instruction to the robot → the robot enters Ward 3, uses a thermometer to measure the patient's temperature through body movements (such as raising a hand), and transmits the data back to the control system.

[0025] In some embodiments, see Figure 2 , Figure 2 The control system (also called the robot service system) of this application is composed of seven core subsystems. Each module works together through a standardized API interface to form a complete service system covering device management, business configuration, and intelligent interaction. The details are as follows:

[0026] The robot management subsystem is responsible for the full lifecycle management of robotic devices, including assigning unique account identifiers to each robot and remotely performing basic operations such as device restarts, shutdowns, and firmware upgrades. The system also collects real-time robot operation logs and hardware health data, enabling operators to centrally monitor and manage device status.

[0027] The robot configuration subsystem is used to fine-tune the robot's software and hardware parameters. Hardware configuration includes adjusting parameters such as voice volume, walking speed, and sensor sampling frequency. Software configuration involves setting the storage path for audio and video data and adjusting the resolution of monitoring images. Environmental configuration includes importing and updating ward maps and customizing navigation and obstacle avoidance strategies, ensuring the robot's adaptability to the environmental requirements of different medical scenarios.

[0028] The robot business subsystem serves as the logical engine for building medical service scenarios. This subsystem supports multi-dimensional business configuration: it defines different service modes such as ward rounds, education, and guided tours through the scenario management module; uses the task scheduling function to plan the robot's walking route, execution cycle, and priority strategy; and assigns exclusive skill packages to robots in different departments such as cardiology and rheumatology based on the role system. At the same time, it presets a business template library such as standardized ward round processes and education scripts to achieve the process and standardization of medical services.

[0029] The multimodal model subsystem provides AI capability support for the robot. By opening the service interface of the multimodal large model, it integrates a series of intelligent tool sets: a medical knowledge graph that integrates specialized knowledge such as cardiology and neurology, an intelligent interaction engine that supports multi-round dialogue management and emotion recognition, a visual recognition module that realizes drug identification and patient vital signs analysis, and a real-time search service that connects to the medical database, providing underlying intelligent support for the robot's medical services.

[0030] The robot skill set subsystem focuses on building the robot's motion and business capability system. It develops a visual scripting language based on the Unreal Engine to achieve modular encapsulation of skills: a body movement library that includes joint movement scripts such as waving and gait simulation, multimedia skills that integrate functions such as singing and dancing performances, business scripts that encapsulate the logic of ward rounds and medical consultations and the process of educational dialogue, and algorithm modules that embed drug grasping force control algorithms and navigation path optimization models to ensure that the robot has rich execution capabilities.

[0031] The human assistance subsystem is designed to establish an emergency control channel for human-machine collaboration. This subsystem supports manual remote control of the robot: it can take over the robot's movement direction and speed in real time, remotely input answer content and trigger voice broadcast, and interrupt abnormal tasks in emergencies and start troubleshooting processes, realizing flexible switching between manual intervention and automated services.

[0032] The intelligent voice and video subsystem, serving as the perception and expression hub for multimodal interaction, builds a complete audio and video processing chain. For voice interaction, it enables human-computer dialogue through "ASR voice recognition - semantic understanding - TTS text-to-speech." For visual processing, it supports camera image acquisition, face recognition, and scene semantic segmentation. Furthermore, it establishes an encrypted multimedia data archiving and retrieval system to ensure the secure storage and efficient access of audio and video data.

[0033] In practice, the control system invokes multiple subsystems to execute user-entered control commands, forming a closed-loop control system of "business logic-intelligent decision-making-physical execution." This modular architecture offers scalability, enabling the rapid integration of new service modules based on the needs of hospital specialties, enhancing the intelligence and customization of medical services.

[0034] Taking the ward round control instruction as an example, the control system jointly implements the ward round task through the robot business subsystem, robot skill set subsystem, robot configuration subsystem and multimodal model subsystem it contains; among them, the robot business subsystem is used to configure the business configuration corresponding to the ward round instruction, and send the business configuration to the robot skill set subsystem, robot configuration subsystem and multimodal model subsystem, among which the business configuration includes the ward round scene configuration, ward round role configuration and ward round skill configuration; the robot skill set subsystem is used to determine the ward round action script according to the ward round skill configuration; the robot configuration subsystem is used to determine the ward movement path according to the ward round scene configuration; the multimodal model subsystem is used to configure the ward round prompt words of the large language model corresponding to the ward round role; finally, the ward round action script, ward movement path and ward round prompt words are sent to the humanoid robot as control instructions.

[0035] Specifically, the control system, as the core of the medical service system, converts abstract ward round instructions into specific actions that can be executed by robots through the modular collaboration of multiple subsystems: the robot business subsystem is responsible for the overall business configuration of ward round tasks, similar to a project manager planning the overall situation; the robot skill set subsystem is used to disassemble ward round actions and generate mechanical execution scripts, just like an action director designing details; the robot configuration subsystem is used to plan the movement path of the ward, which is equivalent to a navigation engineer optimizing the route; the multimodal model subsystem is used to generate ward round dialogue content, playing the role of a language screenwriter.

[0036] When a user enters a ward round instruction, the business subsystem must complete basic configuration. For example, it defines the physical environment for ward rounds (e.g., the layout of wards in a hospital's internal medicine ward, the location of the nurse station, and the identification of infection zones), creating a cognitive map of the environment for the robot. It also determines the robot's role during the round (e.g., doctor, nurse), corresponds to different interview tones and processes (e.g., using soothing language with patients, using professional terminology with family members), and breaks down the core actions of the round (e.g., taking temperature, checking infusion status, recording data, and inquiring about the patient's condition). After the business configuration is complete, the information is synchronized to other subsystems, and each module is converted into the specific elements of the ward round instruction according to its function. For example, based on the ward round role configuration (e.g., doctor role), a medical-specific large language model interface is called to generate scenario-specific ward round prompts. This mechanism, through the layered collaboration of "business configuration - skill decomposition - path planning - language generation," transforms traditional manual ward rounds into a robotic automated process, providing a practical technical solution for improving efficiency.

[0037] Specifically, the above-mentioned multimodal large model subsystem will fine-tune the model according to different control instructions, use the corresponding historical data for training, and construct prompts for the corresponding instructions so that the model can output the corresponding content.

[0038] For example, let's consider a prompt for a cardiology department: You are a hospital cardiology ward rounds robot. Your primary task is to conduct routine ward rounds on patients with coronary heart disease and output medical records to help doctors and nurses understand their conditions and reduce the tedious manual entry of medical records. Task Name: Coronary Heart Disease Ward Rounds; Task Types: 1. Preoperative Ward Rounds 2. Postoperative Ward Rounds.

[0039] Ward rounds process steps:

[0040] 1. Step 1: Say hello to the patient and confirm whether the patient is <Zhang San> himself. The patient must confirm twice in a row to proceed to the subsequent ward rounds. Otherwise, say goodbye to the patient and end the task.

[0041] 2.Step 2: Ask the patient how long it has been since they were diagnosed with coronary heart disease.

[0042] 3.Step 3: Ask whether the disease can cause chest pain.

[0043] 4. Step 4: If the patient indicates chest pain, please describe the feeling of the chest pain in detail, such as whether it feels like a needle prick, a dull pain, or a strong sense of oppression and pain when pressed?

[0044] 5. Step 5: If the patient indicates chest pain, ask the patient whether the pain occurs throughout the chest or is concentrated in a certain location in the chest.

[0045] 6. Step 6: If the patient indicates chest pain, ask the patient how long the pain typically lasts.

[0046] 7. Step 7: If the patient indicates chest pain, ask the patient to recall previous chest pain episodes and whether they were triggered by a specific factor, such as overwork, emotional excitement, or cold stimulation.

[0047] 8.Step 8: Ask the patient whether he still feels chest tightness, shortness of breath, or difficulty breathing.

[0048] 9. Step 9: Check if the patient still has symptoms of irregular heartbeat such as palpitations, heart palpitations, and feeling that the heart is in the throat.

[0049] 10. Step 10: Ask the patient if they experience any other discomfort besides heart problems during the onset of symptoms, such as throat tightness, neck or arm pain, or stomach pain.

[0050] 11.Step 11: Thank the patient for his / her cooperation and ask whether he / she has taken nitroglycerin sublingually during past symptom attacks.

[0051] 12. Step 12: If the patient has taken nitroglycerin to relieve symptoms, ask the patient how effective it is and how long it usually takes for the symptoms to be relieved.

[0052] 13.Step 13: Ask the patient whether he has ever taken quick-acting heart-saving pills sublingually during previous attacks.

[0053] 14. Step 14: If the patient has taken quick-acting heart-saving pills to relieve symptoms, ask the patient how effective they felt after taking the pills. Generally, how long does it take for the symptoms to be relieved?

[0054] 15.Step 15: Ask the patient whether he has ever had an electrocardiogram in the past.

[0055] 16.Step 16: Ask the patient whether he has had a cardiac ultrasound examination in the past.

[0056] Example of human-computer dialogue during ward rounds for patients with coronary heart disease:

[0057] Robot: Hello, I am the hospital's ward rounds robot. Are you the patient in bed number 8?

[0058] Patient: Yes.

[0059] Robot: Nice to meet you, patient in bed 8. Let me confirm again, you are patient in bed 8, correct?

[0060] Patient: Yes, I am.

[0061] Robot: Okay, thank you for your cooperation. How long ago were you diagnosed with coronary heart disease?

[0062] Patient: Half a year ago.

[0063] Robot: I understand. Does your disease cause chest pain?

[0064] Patient: I have chest pain.

[0065] Robot: I understand that you have chest pain. Could you please describe in detail how the chest pain feels? For example, is it like a needle prick, a dull pain, or a strong sense of oppression and pain when pressed?

[0066] Patient: Yes.

[0067] Robot: Excuse me, I need to confirm. Could you please describe in detail how your chest pain feels, such as a needle-like pain, a dull pain, or a strong sense of oppression and pain when pressed?

[0068] Patient: It hurts when I press it.

[0069] Robot: Okay, sorry to bother you. When you have pain, do you feel it throughout your chest, or is it concentrated in one spot?

[0070] Patient: My entire chest hurts.

[0071] Robot: I understand. How long does this pain usually last?

[0072] Patient: About 10 minutes.

[0073] Robot: I see. Were your previous chest pain episodes triggered by any specific factors, such as overwork, emotional excitement, or exposure to cold?

[0074] Patient: I feel uncomfortable when I am overworked or emotionally excited.

[0075] Robot: Got it. Do you still feel chest tightness and shortness of breath?

[0076] Patient: Yes.

[0077] Robot: I see. Do you still have symptoms of irregular heartbeat such as panic, palpitations, or feeling like your heart is in your throat?

[0078] Patient: I feel anxious.

[0079] Robot: Okay, then, besides your heart, do you experience any other discomfort during an attack? For example, a tightness in your throat, soreness in your neck and arms, or stomach pain?

[0080] Patient: Tightness in throat and soreness in arms.

[0081] Robot: Got it. Thank you very much for your cooperation. In the past, when your symptoms flared up, have you taken nitroglycerin, quick-acting heart pills, or other medications to relieve your symptoms?

[0082] Patient: I have taken nitroglycerin.

[0083] Robot: Okay, so how do you feel about the effects of taking nitroglycerin? How long does it usually take for the symptoms to be relieved?

[0084] Patient: It will be relieved in 2 minutes.

[0085] Robot: Thank you for your answer. Have you ever undergone electrocardiogram, cardiac ultrasound and coronary artery CT in the past?

[0086] Patient: Not yet.

[0087] Robot: Got it. Did you undergo coronary angiography during your admission?

[0088] Patient: No angiography was performed when I was admitted to the hospital.

[0089] Robot: OK, I understand. Have you ever had any targeted surgery, such as coronary intervention or heart bypass surgery?

[0090] Patient: No.

[0091] Robot: I see. Do you have any other underlying conditions such as hypertension, hyperlipidemia, or diabetes in your medical history?

[0092] Patient: I have high blood pressure.

[0093] Robot: OK, I understand. Thank you for your cooperation. I will report this information to the doctor. I wish you good health and see you next time.

[0094] Patient: Thank you.

[0095] Robot: You're welcome. I wish you a speedy recovery! If you have any other questions, feel free to ask me.

[0096] Taking the medicine delivery instruction as an example, when the control instruction is the medicine delivery instruction, the robot business subsystem configures the business configuration corresponding to the medicine delivery instruction, and sends the business configuration to the robot skill set subsystem, the robot configuration subsystem and the multimodal model subsystem. Among them, the business configuration includes the medicine delivery scene configuration, the medicine delivery role configuration and the medicine delivery skill configuration; the robot skill set subsystem is used to determine the medicine delivery action script according to the medicine delivery skill configuration; the robot configuration subsystem is used to determine the ward movement path according to the medicine delivery scene configuration; the multimodal model subsystem is used to configure the corresponding large language model prompt words according to the medicine delivery role; finally, the medicine delivery action script, ward movement path and prompt words are sent to the humanoid robot as control instructions.

[0097] Specifically, when a user issues a medicine delivery instruction, the robot service subsystem must first complete core configuration and define the physical environment for medicine delivery (such as the layout of a hospital ward, the location of the pharmacy, elevator access rules, etc.), similar to drawing a task map for the robot. The robot's role in medicine delivery must be determined (such as "nurse assistant" or medicine delivery person), corresponding to different interaction styles (such as using a gentle tone with patients and professional terminology with medical staff). The key actions in the medicine delivery process (such as grabbing the medicine box and handing over the medicine) must be disassembled, and the execution standards and sequence of each action must be clarified.

[0098] For example, based on a medication delivery skill configuration (such as handing medication with both hands and verifying the patient's name), abstract actions are converted into a specific sequence of mechanical instructions: ① Move to the pharmacy pick-up point → ② Grasp the medicine box with dexterous hands using grip force → ③ Move to the patient's bedside → ④ Ask a voice question, "Excuse me, are you Zhang San?" → ⑤ After confirmation, hand over the medication with both hands. The generated "medicine delivery action script" contains precise parameters such as joint movement angles and grip strength, similar to a robot's "action script." The scenario configuration is based on the ward map data (such as corridor width, access control locations, and elevator operation logic) in the medication delivery scenario configuration to calculate the optimal movement path: For example: Nurses Station → Avoid Corridor B in the clean area → Take Elevator No. 3 → Ward 5, avoiding peak patient traffic. Obstacle detection logic is also embedded (such as pausing or detouring when encountering a cart) to ensure a safe and feasible path. Based on the medication delivery role configuration (such as the nurse role), the corresponding large language model interface (such as a medical-specific dialogue model) is used to generate scenario-specific interactive dialogue. Through the above process, abstract medical instructions are converted into physical actions and language interactions that can be understood by the robot, making the drug delivery process intelligent, automated and humanized.

[0099] In some embodiments, in the medical service system, in addition to business instructions such as ward rounds and medicine delivery, users can also input equipment setting instructions, including at least equipment information setting instructions and hardware parameter setting instructions, which are mainly used for equipment management and status monitoring of the robot, similar to "equipment operation and maintenance instructions". Specifically, the control system obtains the equipment setting instructions input by the user, outputs processing result information according to the equipment setting instructions, and returns the processing result information to the user. The robot management subsystem of the control system sets the equipment information of the humanoid robot and generates the current status information of the robot according to the basic control instructions of the equipment setting instructions; the robot configuration subsystem is also used to configure the hardware parameters of the humanoid robot according to the equipment setting instructions, as well as the configuration results; and the control system returns the status information and configuration results to the user as processing result information.

[0100] For example, based on basic user control commands (such as "Set the robot's name to 'Medical Assistant 001'" or "Change its assigned ward to the Internal Medicine Building"), the robot's basic attributes are configured. This information is stored in the system backend and used to distinguish the tasks assigned to different robots. Simultaneously, the system collects the robot's operational data in real time, automatically generates status reports, and integrates the results of processing by the robot management and configuration subsystems.

[0101] In this way, through the closed-loop process of "command input-subsystem processing-response return", the full life cycle management of humanoid robots is realized, providing underlying support for the stable operation of the medical service system.

[0102] In some embodiments, while performing tasks like ward rounds and medication delivery, the humanoid robot uses sensors to collect real-time environmental data. This data is uploaded to a control system for analysis and feedback to the user. The control system then retrieves the data file uploaded by the humanoid robot, generates a recognition result based on the data file, and returns the result to the user. The data file includes at least voice and image data. The control system also includes intelligent voice and video subsystems, which generate corresponding recognition results based on the voice and image data, respectively. The control system then returns the recognition result to the user.

[0103] Specifically, the data file type can be voice data (such as recordings of conversations with patients and environmental sounds); or image data (such as photos of the patient's condition, images of drug labels, and videos of the ward environment). This data can be uploaded after the task is completed (such as uploading all collected data after the ward rounds are completed) or in real time (such as uploading videos and voice immediately in an emergency). The intelligent voice and video subsystem is responsible for structured processing of the data uploaded by the robot. For example, the intelligent voice and video subsystem processes voice data, such as voice recognition (ASR), converting recording files into text (such as converting a patient's words "I feel dizzy today" into text); or semantic understanding, analyzing the meaning of the text, such as extracting key information (such as identifying the patient's possible hypotension symptoms from "dizziness"). It can also be image data processing, analyzing the patient's condition (such as identifying whether the expression is painful through facial images), or scene understanding, such as identifying the ward environment (such as determining whether there are obstacles in the ward or whether the infusion bottle needs to be replaced). The control system integrates the recognition results of the intelligent voice and video subsystem and returns the recognition results to the user.

[0104] Through the above mechanism and the closed loop of "data collection-intelligent recognition-result feedback", the robot becomes the "sensory extension" of medical staff, realizing the automated processing and intelligent transmission of information in medical scenarios, and providing key technical support for the digitalization and intelligence of smart medical care.

[0105] During the robot's medical service delivery process, it interacts with the user. The control system disclosed in this embodiment retrieves the data files uploaded by the humanoid robot, generates a response message based on the data files, and returns the response message to the humanoid robot. Specifically, in the medical service system, the response message is an "intelligent feedback instruction" generated by the control system based on the robot's uploaded data, used to guide the robot's further actions. Its processing flow is different from the second response message (returned to the user), forming a closed loop of "robot data upload → system intelligent analysis → robot execution feedback."

[0106] In some embodiments, the control system calls the intelligent voice and video subsystem to send the recognition results to the multimodal model subsystem, so that the multimodal model subsystem can select a large language model based on the recognition results, and return the recognition results to the humanoid robot.

[0107] The intelligent voice and video subsystems first recognize the data (voice or image data) and then send the results to the multimodal model subsystem. The multimodal model subsystem uses a large language model to analyze the data. In addition to generating recognition results, it can also generate command feedback. Command feedback can be voice interaction, action interaction, or path adjustment. After the robot receives the command feedback, the control module parses it and generates a preset command set (movement, dialogue, etc.), which drives the actuator to execute it.

[0108] In this way, through the process of "data recognition-model analysis-command feedback", the robot is upgraded from a "passive execution tool" to an "intelligent decision-making subject", achieving more flexible and safer automated responses in medical services, and promoting the development of smart medical care towards autonomy and intelligence.

[0109] In some embodiments, the user can use the control system to obtain the data files uploaded by the humanoid robot to generate the required content. For example, in the robot service system, multiple subsystems are used to realize the generation of the ward round medical record by the large model after the ward round is completed.

[0110] The intelligent voice and video subsystems analyze uploaded data (voice or image data). For example, during ward rounds, a microphone array captures the voice signals of conversations between the patient and the robot in real time. ASR speech recognition technology converts these signals into text data, accurately recording the conversation content. Cameras simultaneously capture visual information such as the patient's facial expressions and body movements. While not directly used to generate medical records, these information helps the robot understand the patient's condition and improves the quality of interaction. The multimodal model subsystem uses a large model to perform deep semantic analysis on the text data transmitted by the intelligent voice and video subsystems, identifying key medical information such as symptoms, medical history, and medication use. The system also integrates its own medical knowledge graph to classify and integrate the extracted information. For example, symptoms mentioned by the patient can be categorized as current medical history, and past illnesses can be classified as past medical history. Based on a pre-set medical record template (provided by the robot service subsystem), the system automatically populates the corresponding fields with the organized information to generate a structured ward round record. If necessary, a manual assistance subsystem will be introduced to facilitate human intervention and review. If questions or inaccuracies are found in the generated medical records, medical staff can use this subsystem to manually review, supplement information, correct errors, and improve the ward round medical record. These subsystems work together, each performing its own function, to achieve a complete process from ward round conversation data collection and processing to the final ward round medical record generation.

[0111] Here's an example of a ward round medical record generated using a large model: Ward Round Medical Record Summary: Patient Information: Name: Neurology Patient, Bed 7, Gender: Male, Age: 60, Chief Complaint: Carotid Artery Stenosis --- Chief Complaint: The patient presented with left leg numbness for three months. --- Present Medical History: 1. The patient reports mild numbness in the left leg, which has persisted for approximately three months. 2. The patient has found no effective treatment to alleviate the symptoms. 3. The patient denies dizziness, headaches, hearing loss, or vision impairment, and has normal mobility. 4. The patient's sleep and bowel movements are normal, with no significant weight changes. --- Past Medical History: 1. The patient denies any history of hypertension, hyperlipidemia, or hyperglycemia. 2. The patient denies any history of infectious diseases such as hepatitis and tuberculosis. 3. The patient denies any history of cardiovascular or cerebrovascular disease. 4. The patient denies any history of surgery. 5. The patient has taken oral medications such as aspirin, clopidogrel bisulfate tablets, and ticagrelor. 6. The patient denies any food or drug allergies. --- Personal History: 1. The patient denies any chronic smoking or alcohol consumption. ---Marital and reproductive history: 1. The patient is married with three children. 2. The patient's spouse and children are in good health. ---Family history: 1. The patient's parents are alive and deny any family history of cardiovascular or cerebrovascular disease. 2. The patient denies any family history of any specific genetic disease.

[0112] In the aforementioned interaction between the control system and the humanoid robot, the control system achieves deep interaction with the humanoid robot through systematic configuration. Specifically, the system allows for comprehensive configuration of the robot's account, role, scenario, skills, ward map, and other elements. Role settings are closely tied to medical business scenarios, such as a cardiology ward round robot or a rheumatology ward round robot. Once assigned a specific role, the robot automatically invokes the corresponding skills to perform tasks such as ward round dialogues, achieving precise business functions. Scenario configuration defines the robot's work environment, including ward rounds, tours, and education. Each scenario is matched with a dedicated skill. Furthermore, general skills such as chatting, singing, and dancing can be flexibly applied to various scenarios. This configuration system, based on digital technology, deeply integrates medical business logic with the robot's physical capabilities, creating a "define and execute" intelligent service model. It not only enables efficient execution of medical tasks but also provides solid technical support for multi-task collaboration in smart healthcare scenarios, significantly improving the intelligence and convenience of medical services.

[0113] In this embodiment, a humanoid robot is also provided, which includes at least a control module and an actuator that are interconnected; wherein the control module is communicatively connected to the control system, and is used to generate a preset instruction set in response to control instructions sent by the control system, wherein the preset instruction set includes at least movement instructions, dialogue instructions, limb movement instructions, and configuration instructions; the actuator is connected to the control module, and is used to execute medical service actions corresponding to the preset instruction set.

[0114] See also Figure 3 , Figure 3 A diagram of the robot hardware control module architecture is provided. In some embodiments, the control module includes at least a controller and a control unit. The control unit is communicatively connected to the control system and is configured to generate a preset instruction set in response to control instructions generated by the control system. The controller is connected to the control unit and is configured to obtain the instruction set generated by the control unit and control the actuators according to the instructions.

[0115] Specifically, the controller is mainly responsible for sensor data acquisition and motor control tasks, responsible for real-time processing of environmental information fed back by various sensors, and accurately regulating the motor to perform corresponding actions. The robot control unit, as the intelligent center of the system, is built based on the Android operating system, integrating LTE / 5G and Wifi dual network modules, and supports interaction with the robot service system (i.e., the control system disclosed in this application) through network communication. Its core function is to realize application deployment and intelligent interaction-it can install various APP applications and deploy intelligent body programs in the form of APP. These programs establish communication with the controller through the USB interface, and drive the execution of skills by calling the interface, such as controlling the movement of the robotic arm or triggering sensor data collection. The two units are physically connected through the USB interface to build a closed-loop communication link of "control instruction issuance-execution data return", ensuring that the robot can achieve efficient collaboration in the entire process of perception, decision-making, and execution, and provide stable underlying architecture support for intelligent application scenarios.

[0116] Please continue reading Figure 3 , Figure 3 The core components and interaction logic of the robot hardware system are demonstrated. The entire system is centered on the controller, which collects information through sensors, executes component actions, and connects control units to form a complete closed-loop control system.

[0117] Within the core control system, the controller serves as the core of the system, processing data from various sensors and modules and coordinating the robot's movements. Sensors include various types: ultrasonic sensors detect the distance to objects by transmitting and receiving ultrasonic waves, enabling close-range obstacle avoidance; infrared sensors detect infrared signals from objects, assisting with detection and navigation; depth cameras acquire scene depth information for 3D modeling and object positioning; cameras capture images and support visual functions such as facial recognition; microphones collect sound signals for voice recognition; lidar creates environmental maps through laser scanning, supporting autonomous navigation; and inertial measurement units (IMUs) provide motion status data to assist with balance control.

[0118] In the execution and operation section, the active wheels act as moving components, receiving controller commands to achieve movement and steering. Joints are driven by motors, giving the limbs movement capabilities. The dexterous hand simulates the functions of a human hand, enabling grasping and manipulation. A speaker converts electrical signals into sound, enabling voice interaction. In the power and connection section, the motor driver converts controller signals into motor drive signals, and the power module provides power to all components. The robot control unit in the system connection section acts as a bridge between the service system and the controller, transmitting commands and providing feedback on status information.

[0119] Each part interacts with data through hardware connections, forming a complete system for the robot from environmental perception, decision-making and control to action execution, providing infrastructure support for the application of robots in different scenarios.

[0120] See also Figure 4 , Figure 4 The components of a humanoid robot's actuator are presented. In some embodiments, the dexterous hands and joints are used to execute corresponding actions according to body movement commands; the mobile chassis is used to move the humanoid robot to a target location according to movement commands; and the microphone and speaker are used to execute dialogue commands, acquire user language, and generate dialogue speech.

[0121] The specific composition of the robot's actuator may also include other components, such as Figure 4 As shown in the figure, the hardware system of the humanoid robot is composed of perception, movement and interaction components, which work together to realize the intelligent functions of the robot.

[0122] In terms of perception, infrared sensors are usually installed in places such as the head, and detect obstacles and approaching human bodies by sensing the infrared radiation of objects; depth cameras can obtain scene depth information for three-dimensional modeling and object positioning; microphone arrays are distributed on the head or chest to collect sound signals for voice recognition; ordinary cameras capture images and support face recognition and object detection; lidars are mostly installed on the bottom or around the body, and create environmental maps through laser scanning to assist autonomous navigation; ultrasonic radars are distributed on the bottom or edge, and use ultrasonic echolocation to detect close-range obstacles.

[0123] Movement is performed by a mobile chassis and joint components. The mobile chassis, which can be wheeled or tracked, enables walking and steering. Joints in the shoulder, elbow, wrist, hip, and knee give the robot limbs movement capabilities. The dexterous hand mimics the structure of a human hand, possessing multiple joints and degrees of freedom, enabling it to grasp and manipulate objects.

[0124] Human-machine interaction is primarily achieved through a speaker, which converts text into speech output for playing prompts, conversations, and educational messages. These hardware components work together to support the application of humanoid robots in healthcare, service, and other scenarios.

[0125] The humanoid robot executes system commands through a control module and actuators, achieving mobile navigation and body movement functions. For example, the robot utilizes sophisticated SLAM (Simultaneous Localization and Mapping) technology for autonomous movement. This technology allows the robot to simultaneously build a map and localize itself in unknown environments. In a ward scenario, manual control of the robot's movement is initially required. A navigation sensor system, including lidar, ultrasonic radar, and IMU, collects environmental data to construct a ward map, laying the data foundation for subsequent autonomous navigation and dynamic obstacle avoidance functions. This SLAM-based navigation solution enables the robot to accurately identify paths and obstacles in complex medical environments, achieving stable movement. Furthermore, the humanoid robot is equipped with a multi-degree-of-freedom arm, which achieves flexible movement in three-dimensional space through the coordinated operation of multiple articulated joints. When no voice command input is available, the robot can simulate human body movements, such as waving and raising hands, based on pre-set commands. This natural movement expression enhances the user-friendly interaction and optimizes the user experience. This movement execution mechanism combines mechanical structure and control algorithms to ensure smooth movement and interactive adaptability.

[0126] Another aspect of the present application is to provide a medical service method based on a humanoid robot. This method relies on a specific medical service system to implement the process from command input to robot execution, aiming to use the humanoid robot to assist or replace some medical service work. Figure 5 , Figure 5 A flowchart of a medical service method based on a humanoid robot is given. As shown in the figure, the method includes the following steps:

[0127] Step 501: The control platform sends a ward round instruction and / or a medicine delivery instruction to the humanoid robot.

[0128] Step 502 : The humanoid robot responds to a ward round instruction and / or a medicine delivery instruction and performs a medical service action corresponding to the ward round instruction and / or the medicine delivery instruction.

[0129] Specifically, the ward rounds and / or medication delivery instructions here are instructions issued by users (such as medical staff) to the system to perform medical service-related tasks. The control platform is responsible for receiving these instructions and transmitting them to the humanoid robot, which is a key link in the instruction flow. After receiving the instructions, the humanoid robot performs the corresponding medical service actions according to the instructions and utilizes the various functional systems of the medical service system's control platform (such as the action scripts determined by the skill set subsystem and the paths planned by the configuration subsystem). For example, it completes ward inspections according to ward rounds instructions and delivers medicine according to medication delivery instructions. In this way, through the collaborative architecture of "control platform + humanoid robot", routine medical tasks assigned by medical staff are completed, and the service experience is enhanced through human-like interaction, providing a solution for smart healthcare that combines efficiency and practicality.

[0130] In one embodiment, see Figure 6 After step 502, the method may further include the following steps:

[0131] Step 503: The humanoid robot collects and generates a data file;

[0132] Step 504: The control platform generates a recognition result based on the data file and returns the recognition result to the user.

[0133] Specifically, the humanoid robot uses a variety of sensors, including microphone arrays, cameras, and lidar, to collect data and perform interactive operations. During ward rounds, the microphone array collects patient voice information, while the camera captures visual data such as facial expressions and movements. During movement, lidar and ultrasonic radar collect environmental information.

[0134] The collected raw data must be organized and encoded according to a unified format standard: voice information is converted into audio files, visual data is converted into image / video formats, and environmental information is generated into structured coordinate files. After receiving these data files, the control platform uses an algorithmic model to extract key information. This involves using speech recognition and natural language processing technologies to analyze the symptoms and requirements expressed during ward rounds, and using environmental data to determine the robot's position and obstacle conditions.

[0135] Based on data analysis results, the control platform generates responses according to pre-set rules and the knowledge base. If a patient inquires about their condition, a professional explanation is generated based on the medical knowledge base; if the robot encounters an environmental issue, an operational instruction is generated. Ultimately, the response is fed back to the user via the network: for the patient, it is presented in voice or text form, while for the robot, it is converted into control instructions to drive the next action. The entire process forms a closed-loop interactive mechanism of "data collection - processing and analysis - decision-making and response."

[0136] The steps of the various methods above are divided only for the purpose of clear description. During implementation, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0137] In some specific examples, the medical service actions supported by the above medical service system may include key links such as ward rounds, patient education, ward tours, and ward medication delivery. The following further describes the specific execution process of these specific medical service actions in conjunction with the various subsystems in the control system in the above embodiment.

[0138] When a medical service action involves ward rounds, and a doctor issues a voice command such as "Go to bed number 8 for rounds," the intelligent voice and video subsystem performs speech recognition and intent classification, transmitting the round task type and the target location of bed number 8 to the robot service subsystem. This subsystem then configures parameters such as the round scenario, role, and skills, and synchronizes these parameters with the robot skill set subsystem (which generates the action script), the robot configuration subsystem (which plans the navigation path), and the multimodal model subsystem (which prepares the dialogue script). The robot configuration subsystem generates the optimal path based on the ward map, and the robot control unit drives the mobile chassis to perform navigation. During navigation, the intelligent voice and video subsystem processes the environmental image in real time, identifies obstacles, and triggers the obstacle avoidance algorithm. The robot configuration subsystem dynamically adjusts the path accordingly. Upon reaching the target location, the multimodal model subsystem invokes the medical large language model to generate opening lines such as "Hello, I am a ward round robot." This is then synthesized and played back using the text-to-speech module of the intelligent voice and video subsystem. When the patient responds, the system performs real-time speech recognition, and the multimodal model subsystem generates follow-up questions based on the context, such as "What is your temperature today?" Simultaneously, the robot's control unit controls limb movements according to a pre-set action script, achieving anthropomorphic gestures such as arm swings. After the conversation concludes, the multimodal model subsystem integrates the patient's responses and sensor data such as body temperature and facial expression recognition to generate a structured ward round record, which is submitted to the robot service subsystem for storage as an electronic medical record. The intelligent voice and video subsystem records the entire process and stores it in a path specified by the robot configuration subsystem (such as a hospital server / cloud storage). Upon completion of the ward round, the robot service subsystem marks the task status as "Completed," the robot configuration subsystem plans the return path, and the robot autonomously navigates to its standby point. If the doctor issues a "Cancel the ward round" command during the round, the intelligent voice and video subsystem recognizes this intention and notifies the robot service subsystem. This subsystem then interrupts the current task, and the robot configuration subsystem replans the return path. The robot terminates the conversation and returns to its standby point, marking the task status as "Cancelled."

[0139] When a medical service action involves patient education, such as a healthcare provider issuing a voice command like "Go to bed 8 for preoperative education," the robot's intelligent voice and video subsystem first recognizes the voice command, accurately extracts the task type ("preoperative education") and the target location of bed 8, and transmits this information to the robot's business subsystem. Based on the task type, this subsystem then calls the corresponding preoperative education template, covering topics such as preoperative fasting instructions and postoperative recovery essentials. This template information is then synchronized with the multimodal model subsystem and the robot's skill set subsystem to prepare for subsequent task execution. Next, the robot's configuration subsystem calculates the optimal route to bed 8 based on the constructed ward map and transmits this route information to the robot's control unit. During the robot's movement, the intelligent voice and video subsystem processes environmental data in real time. If an obstacle, such as a cart, is detected in the path, the robot's obstacle avoidance algorithm is triggered, causing the robot to automatically circumvent it to ensure safe arrival at the target location. Upon arrival at bed 8, the robot begins presenting the educational content. The multimodal model subsystem breaks down the educational text into multiple voice segments, which are synthesized into voice for playback through the TTS module of the intelligent voice and video subsystem. At the same time, the robot skill set subsystem generates a matching body movement script. For example, when explaining the surgical process, it controls the robot to make corresponding gestures to assist the patient in understanding the educational content, which is driven and executed by the robot control unit. After the education task is completed, the robot business subsystem updates the task status to "completed" and generates an execution log to record key information such as the education time and content summary. Subsequently, the robot configuration subsystem plans the return path, and the robot autonomously navigates along the path and returns to the initial standby point, completing the entire patient education task process.

[0140] When the medical service action involves a ward tour, during the command parsing and task planning phase, the robot uses the control platform's intelligent voice and video subsystems to recognize the command and identify the "ward tour" task type. The robot's business subsystem loads a list of target locations, such as the nurse station, pharmacy, treatment room, and restroom, from a pre-set template and configures the corresponding explanation content. The robot's configuration subsystem generates an optimal tour route based on the ward map and synchronizes it with the control unit. During navigation and point-to-point movement, the robot's control unit drives the mobile chassis to each location along the planned path. During movement, the intelligent voice and video subsystem processes environmental data in real time. For example, detecting a pedestrian triggers the obstacle avoidance algorithm to cause the robot to pause and wait, ensuring safe movement. Upon arrival at the target location, the robot enters the explanation interaction phase. The multimodal model subsystem invokes the corresponding explanation text, such as "This is the treatment room. Routine nursing procedures are performed every morning at 9:00 AM." The intelligent voice and video subsystem converts the text into speech output, while the robot's skill set subsystem generates supporting body movements, such as pointing the device or performing hand gestures. If the patient asks a question, the multimodal model subsystem generates a real-time response and provides feedback through the voice and video systems. After completing the explanation of all points, the robot business subsystem updates the task status to "Completed", the robot configuration subsystem plans the return path, the robot autonomously navigates to the standby point, and uploads the execution log containing information such as the stay time and number of interactions, forming a closed task loop.

[0141] When a medical service action involves delivering medicine to a ward, if a healthcare provider issues a voice command like "Go to bed 8 and deliver medicine, two tablets at a time, three times a day, after meals," the robot will execute the medicine delivery task in sequence. During the command parsing and task configuration phase, the intelligent voice and video subsystem first recognizes the voice command, accurately extracts the medicine delivery task type, the target location of bed 8, and the medication order information, and passes it to the robot's business subsystem. This subsystem then creates the medicine delivery task, associates it with the corresponding electronic medical order, and synchronizes the task information with the robot configuration subsystem and the multimodal model subsystem to prepare for subsequent execution. During the drug acquisition and verification phase, the multimodal model subsystem utilizes a visual language model, combined with camera-generated image data, to identify the medication packaging provided by the nurse. The robot's control unit then activates its dexterous hand to grasp the medication, using tactile and force sensors to ensure a stable grip. Simultaneously, the intelligent voice and video subsystem verifies the name, dosage, and other information on the medication label against the medical order, and feeds the results back to the business subsystem. During navigation and path planning, the robot configuration subsystem generates the optimal route to bed 8 based on the ward map and sends it to the robot control unit. During movement, the intelligent voice and video subsystem continuously monitors the environment. If an obstacle, such as a stretcher, is detected, the obstacle avoidance algorithm is triggered, controlling the robot to circumvent it. Upon reaching the target location, the drug delivery and patient interaction phase begins. The multimodal model subsystem generates voice prompts, such as "Bed 8, your medication has arrived...", which are played back through the intelligent voice and video subsystem. The system uses cameras and depth sensors to locate the patient's hand. The robot control unit activates the dexterous hand to deliver the medication, and force sensors sense the patient's movement to confirm the delivery. Upon completion, the robot service subsystem updates the task status to "Completed" and records the delivery time. The robot configuration subsystem plans the return path, and the robot autonomously navigates to the standby point. It also uploads an execution log containing drug verification results and patient interaction videos, completing the entire drug delivery task loop.

[0142] In this application, the control platform serves as the central system. It receives instructions for carrying out medical service-related tasks, such as ward rounds instructions, medicine delivery instructions, etc., through the control system within the platform, and sends control instructions to the humanoid robot. The control module of the humanoid robot communicates with the control system and generates a preset instruction set after receiving the instructions. The movement instructions in the instruction set enable the robot to navigate to the target position, the dialogue instructions enable the robot to interact with medical staff or patients, and the body movement instructions control the robot to perform corresponding actions. The robot also includes an actuator to perform corresponding medical service actions according to the instructions generated by the control module (such as completing ward inspections according to ward rounds instructions and delivering medicines according to medicine delivery instructions). Through the collaborative architecture of "control platform + humanoid robot", routine medical tasks assigned by medical staff are realized, and the service experience is improved through human-like interaction, providing a solution for smart medical care that is both efficient and practical.

[0143] Those skilled in the art will appreciate that the above embodiments are specific embodiments for implementing the present application, and that in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present application.

Claims

1. A medical service system based on a humanoid robot, characterized in that: comprising a control platform, and a humanoid robot responsive to the control platform; wherein, The control platform includes at least a control system, and the control system is communicatively connected to the humanoid robot; The control system is used to send ward round instructions and / or medicine delivery instructions to the humanoid robot; The humanoid robot comprises at least a control module and an actuator connected to each other; wherein, The control module is in communication with the control system, and is configured to respond to ward round instructions and / or medicine delivery instructions sent by the control system and generate a preset instruction set, wherein the preset instruction set includes at least movement instructions, conversation instructions, body movement instructions, and configuration instructions; The execution mechanism is connected to the control module and is used to execute the medical service action corresponding to the preset instruction set.

2. The medical service system according to claim 1, characterized in that: The control system includes at least a robot business subsystem, a robot skill set subsystem, a robot configuration subsystem and a multimodal model subsystem; The ward round instruction at least includes a ward round action script, a ward movement path, and a ward round prompt word; The robot business subsystem is configured to configure a business configuration corresponding to the ward round instruction and send the business configuration to the robot skill set subsystem, the robot configuration subsystem, and the multimodal model subsystem, wherein the business configuration includes a ward round scenario configuration, a ward round role configuration, and a ward round skill configuration; The robot skill set subsystem is used to determine the ward round action script according to the ward round skill configuration, and send the ward round action script to the humanoid robot; The robot configuration subsystem is configured to determine the ward movement path according to the ward rounds scene configuration, and send the ward movement path to the humanoid robot; The multimodal model subsystem is used to configure the ward round prompt words of the large language model corresponding to the ward round role and send the ward round prompt words to the humanoid robot.

3. The medical service system according to claim 2, characterized in that: The medicine delivery instruction at least includes a medicine delivery action script, a ward movement path, and a medicine delivery prompt word; The robot business subsystem is further configured to configure a business configuration corresponding to the medicine delivery instruction and send the business configuration to the robot skill set subsystem, the robot configuration subsystem, and the multimodal model subsystem, wherein the business configuration includes a medicine delivery scenario configuration, a medicine delivery role configuration, and a medicine delivery skill configuration; The robot skill set subsystem is used to determine a medicine delivery action script according to the medicine delivery skill configuration, and send the medicine delivery action script as a medicine delivery instruction to the humanoid robot; The robot configuration subsystem is configured to determine a ward movement path according to the medicine delivery scenario configuration, and send the ward movement path to the humanoid robot; The multimodal model subsystem is used to configure the medicine delivery prompt words of the large language model corresponding to the medicine delivery role and send the medicine delivery prompt words to the humanoid robot.

4. The medical service system according to claim 2, characterized in that: The control system further includes a robot management subsystem connected to the humanoid robot; The robot management subsystem sets the device information of the humanoid robot according to the device information setting instruction input by the user, generates the current status information of the robot, and returns the status information to the user; The robot configuration subsystem is further configured to configure the hardware parameters of the humanoid robot according to the hardware parameter setting instructions input by the user, generate a configuration result, and return the configuration result to the user.

5. The medical service system according to claim 2, characterized in that: The control system also includes an intelligent voice and video subsystem connected to the humanoid robot; The intelligent voice and video subsystem is used to obtain the data files obtained by the humanoid robot when executing the ward rounds instruction and / or medicine delivery instruction, generate recognition results based on the data files, and feed back the recognition results to the user.

6. The medical service system according to claim 5, characterized in that: The intelligent voice and video subsystem is further configured to send the recognition result to the multimodal model subsystem; The multimodal model subsystem is also used to generate a response message based on the recognition result based on the large language model, and return the response message to the humanoid robot so that the humanoid robot can interact with the patient based on the response message.

7. The medical service system according to claim 1, characterized in that: The control module at least includes a controller and a control unit; wherein, The control unit is in communication with the control system and is configured to generate a preset instruction set in response to a ward round instruction and / or a medicine delivery instruction sent by the control system; The controller is connected to the control unit and is used to obtain the instruction set generated by the control unit and control the actuator according to the instructions.

8. The medical service system according to claim 7, characterized in that: The actuator comprises at least a dexterous hand connected to the controller, a joint, a microphone, a speaker and a mobile chassis; The dexterous hand and the joints are used to perform corresponding actions according to the body movement instructions; The mobile chassis is used to move the humanoid robot to a target position according to the movement instruction; The microphone and the speaker are used to execute the dialogue instruction to obtain user language and generate dialogue voice.

9. A medical service method based on a humanoid robot, characterized in that: Using the medical service system according to any one of claims 1 to 8, the method comprises: The control platform sends ward round instructions and / or medicine delivery instructions to the humanoid robot; The humanoid robot responds to the ward round instruction and / or the medicine delivery instruction and performs the medical service action corresponding to the ward round instruction and / or the medicine delivery instruction.

10. The medical service method based on a humanoid robot according to claim 9, characterized in that: After the humanoid robot performs the medical service action corresponding to the ward round instruction and / or the medicine delivery instruction, the method further includes: The humanoid robot collects and generates data files; The control platform generates a recognition result based on the data file and returns the recognition result to the user.

Images