Intelligent massage manipulation cloud platform and remote teaching system

Abstract

The invention discloses an intelligent massage manipulation cloud platform and a remote teaching system. The system adopts a three-level architecture of a cloud platform, an edge end and a hardware end. The hardware end comprises a teacher end acquisition glove and a student end feedback glove which are respectively used for acquiring manipulation characteristic signals and providing tactile feedback. The edge end is deployed at a user terminal and is responsible for feature event coding and decoding and local real-time deviation comparison. And the cloud platform provides data management and collaborative service. The core of the invention lies in that a'feature event coding 'technology is adopted, key features such as strength level switching, direction inflection points and the like of massage manipulation are converted into lightweight event codes for transmission, and low-delay tactile presence teaching is realized through tactile mapping and local instant feedback of gloves at teacher and student ends. The system effectively solves the problems that a traditional teaching technique is difficult to standardize, remote practical operation feedback is lacked, and an existing technical scheme is high in cost and large in delay, and low-cost and high-reliability massage technique remote teaching is achieved.

Description

Technical Field

[0001] This invention belongs to the field of healthcare informatics technology, specifically a cloud platform for intelligent massage techniques and a remote teaching system. Background Technology

[0002] As an important component of Traditional Chinese Medicine (TCM), the quality of its teaching and transmission directly impacts the development of TCM. Traditional TCM teaching primarily relies on oral instruction and hands-on guidance between master and apprentice. In this model, the teacher demonstrates techniques on-site, while students observe, imitate, and rely on the teacher's direct tactile correction and subjective evaluation to learn the essential techniques. This model heavily depends on the teacher's personal experience and the student's aptitude, and inherently suffers from drawbacks such as difficulty in standardizing the teaching process, quantifying teaching effectiveness, and the geographical limitation of sharing high-quality teaching resources.

[0003] With the development of information technology, some technical solutions have emerged that attempt to digitize the teaching of massage techniques. These existing technical solutions can be broadly divided into two categories: The first type is a remote teaching system based on high-definition video recording or real-time streaming. This type of solution can, to some extent, transmit visual information remotely, allowing students to observe demonstrations by renowned instructors. However, the core essence of massage techniques lies in the application of "force," namely the subtle changes in pressure, rhythm, and depth of penetration during the operation—all of which fall under the realm of touch. Pure video teaching solutions cannot convey tactile information at all; students can only "see the form" but cannot "feel the force," failing to obtain the crucial force sensation and feedback in the operation, thus limiting the teaching effectiveness.

[0004] The second category is interactive teaching systems based on high-precision force sensors and complex force feedback devices. This type of solution attempts to address the problem of tactile transmission. It typically employs high-precision six-dimensional force / torque sensors to collect mechanical data of the teacher's techniques and utilizes expensive robotic arms or specialized force feedback devices to provide students with a reproduction of force sensation. While this type of solution can theoretically reproduce force sensation, it has revealed many serious problems in practical applications: The high cost, with the expensive high-precision sensors and force feedback equipment it relies on, and the cost of a single system typically reaching tens or even hundreds of thousands of RMB, makes it difficult to promote and apply in scenarios such as colleges and training institutions that require large-scale deployment.

[0005] The system is complex and bulky. The complex mechanical structure and the size of the equipment seriously interfere with the practitioner's natural operating experience, which is far removed from the real clinical practice of massage and affects the authenticity of teaching.

[0006] Network latency is a critical factor. The need for real-time transmission of high-frequency continuous force data places extremely high demands on network bandwidth and latency. In ordinary internet environments, it is difficult to guarantee the real-time and synchronous nature of tactile feedback, and network jitter can easily disrupt the "tactile presence," significantly diminishing the teaching effectiveness.

[0007] It failed to focus on the core of teaching, and its technical approach focused on the "precise replication" of continuous force signals rather than optimizing the core points of massage technique teaching - namely, mastering a limited number of key features (such as the switching of force levels, changes in direction, and the feeling of penetration). As a result, the technical complexity was not proportional to the actual teaching benefits.

[0008] In summary, existing massage teaching solutions either fail to convey core tactile information or are difficult to implement and scale due to issues such as cost, complexity, and network requirements, thus failing to effectively meet the current needs for standardized, remote, and universally accessible massage techniques. Therefore, there is an urgent need in this field for an intelligent teaching system that can effectively convey the core tactile characteristics of massage techniques and provide real-time feedback in a low-cost, low-latency, and highly reliable manner. Summary of the Invention

[0009] The purpose of this invention is to provide an intelligent massage technique cloud platform and remote teaching system. Through "feature event encoding" technology, it achieves remote, low-latency, and highly reliable transmission and real-time feedback of the core tactile features of massage techniques at extremely low cost, effectively solving the core pain points of traditional teaching standardization difficulties and the high cost and large latency of existing technologies.

[0010] The technical solution adopted in this invention is as follows: A cloud platform and remote teaching system for intelligent massage techniques include: The teacher-side data collection glove is used to collect characteristic signals during the teacher's massage operation. The characteristic signals include pressure signals applied to the object being massaged and signals indicating changes in the direction of hand movements. The student-side feedback glove is used to receive feature event codes and convert them into perceptible tactile feedback signals; At the edge, deployed on teacher and student terminals, it is used to encode / decode the feature signals and realize local real-time comparison of student operation data with standard technique data; The cloud platform communicates with the edge device and is used to manage user information, standard method library, teaching collaboration process, assessment data and system data; The teacher-side data acquisition glove, the edge device, the cloud platform, and the student-side feedback glove are sequentially connected to form a closed-loop remote teaching system that integrates technique acquisition with tactile feedback.

[0011] Preferably, the teacher-side data collection gloves include: Flexible glove substrate 1; At least one pressure sensing module is disposed at the corresponding position of the palm and finger area of ​​the flexible glove substrate, for directly detecting the pressure applied to the object during the massage operation and generating a force level signal; At least one attitude sensing module is disposed at a corresponding position on the back of the hand or wrist of the flexible glove substrate, for detecting changes in the angle and direction of hand movement and generating a direction change signal; A first microcontroller, fixed to the wrist or back of the hand of the flexible glove substrate, is electrically connected to the pressure sensing module and the posture sensing module, and is configured to collect the force level signal and the direction change signal, and assemble them into a feature data frame. A first wireless communication module, connected to the first microcontroller, is used to send the feature data frame to the edge terminal where the teacher terminal is located.

[0012] Preferably, the pressure sensing module is a flexible thin-film pressure sensor, which is attached to the thumb pad, thenar eminence, or palm area of ​​the flexible glove substrate; the attitude sensing module is a six-axis inertial measurement unit, used to detect the pitch angle, roll angle, and yaw angle of the hand.

[0013] Preferably, the student-side feedback glove includes: Flexible glove substrate two; The tactile feedback array consists of multiple tactile feedback actuators distributed at different locations on the flexible glove substrate, which are used to generate perceptible tactile cue signals under the drive of the second microcontroller; The second microcontroller is fixed to the wrist or back of the hand of the flexible glove substrate and is used to receive feature event codes from the edge of the student terminal and drive the tactile feedback array to generate corresponding tactile feedback. The second wireless communication module, connected to the second microcontroller, is used to receive data from the edge device.

[0014] Preferably, the tactile feedback actuator in the tactile feedback array is a miniature vibration motor or a linear resonant actuator; the distribution position of the tactile feedback actuator corresponds to the arrangement position of the pressure sensing module on the teacher's end-to-end collection glove, so that the tactile feedback direction perceived by the student is consistent with the part of the teacher applying force.

[0015] Preferably, the student feedback glove further includes an acoustic prompting module disposed on the second flexible glove base. When the system detects a directional deviation in the student's operation, the second microcontroller drives the acoustic prompting module to emit a prompting sound.

[0016] Preferably, the edge end includes: The feature event encoding / decoding module is used to encode the acquired feature signals into lightweight feature event codes and decode the received feature event codes into drive instructions. The local deviation comparison module is used to compare the real-time characteristic event sequence generated by the trainee with the pre-stored standard characteristic event sequence during trainee practice or assessment, and generate deviation information. The audio-video synchronization module is used to match the timestamp of the video stream with the timestamp of the feature event code to ensure the synchronization between video playback and haptic feedback.

[0017] Preferably, the local deviation comparison module adopts the time window comparison method, and sets a time tolerance window. If the triggering time of the student's feature event exceeds the time tolerance window, or the feature type does not match the standard event, it is determined to be an operation deviation. After detecting the deviation, the edge terminal directly sends a tactile feedback command to the student terminal's feedback glove to achieve low-latency local deviation correction.

[0018] A feature-encoding-based remote teaching method for massage techniques, applied to the aforementioned system, the method comprising: Step S1: The teacher wears the teacher-end collection gloves to perform massage operations. The system directly collects the manual feature signals applied to the object of operation, and uploads them to the cloud platform after edge encoding to build or update the standard manual library. Step S2: The student selects learning content from the cloud platform, and the edge device downloads the corresponding standard feature event sequence; Step S3: During remote teaching or practice, the real-time feature event codes on the teacher's end are synchronized to the student's end via the cloud platform; Step S4: The student wears the student-end feedback glove, whose edge receives the real-time feature event code and drives the tactile feedback array on the student-end feedback glove to generate corresponding tactile feedback, so that the student can perceive the teacher's operating force and rhythm in the corresponding part of the hand. Meanwhile, the edge device compares the sequence of characteristic events generated by the student's operation with the standard sequence in real time. Once a deviation is detected, it immediately provides corrective tactile cues to the student through the student's feedback glove.

[0019] This invention relates to an intelligent massage technique cloud platform and remote teaching system. Compared with the closest existing technology, this invention, through comprehensive innovation in technical conception, system architecture, and specific implementation methods, has produced a series of significant advancements and beneficial effects that are interconnected and synergistic, as detailed below: First, at the technical principle level, this invention, by establishing the "feature event encoding" theory, achieves a fundamental shift in the paradigm of tactile information transmission, effectively overcoming the technical bottlenecks of high bandwidth dependence and high latency in distance learning. Existing technologies generally focus on the high-fidelity acquisition and reproduction of continuous mechanical signals in massage operations. This "precise replication" path inevitably leads to a large amount of data and high transmission bandwidth requirements, making it difficult to achieve real-time and stable tactile presence transmission in conventional network environments, thus significantly reducing the immersiveness and effectiveness of distance learning. This invention abandons the above-mentioned traditional approach and creatively proposes to deconstruct the continuous and complex dynamic process of massage techniques into a series of discrete "key feature events" with clear clinical significance. By focusing on the accurate perception, efficient encoding, and reliable transmission of these feature events, rather than redundantly replicating continuous force curves, this invention significantly reduces the data transmission frequency required by the system from over 100Hz in traditional solutions to an average of less than 10Hz, achieving an order-of-magnitude reduction in data transmission volume (typically exceeding 90%). This fundamental shift makes it possible to achieve latency-free or low-latency "tactile" synchronous teaching in ordinary commercial Wi-Fi or 4G / 5G mobile network environments, thus fundamentally solving the core technical problem of the break in tactile presence caused by network latency.

[0020] Secondly, regarding economic viability and scalability, this invention, based on the aforementioned theoretical innovations, employs a highly simplified and low-cost hardware design, overcoming the cost barriers of high-end force feedback devices and achieving feasibility for large-scale deployment. Since complex force reproduction is not required, the hardware of this invention can be entirely constructed using mature, inexpensive, general-purpose commercial components (such as thin-film pressure sensors, inertial measurement units, micro-vibration motors / linear resonant actuators, and general-purpose microcontrollers). Through ingenious circuit design and firmware algorithms, the material cost of a single student-end device is controlled at an extremely low level while ensuring the reliability of core functions. Compared to existing force feedback devices that cost tens or even hundreds of thousands of yuan per set, the cost is reduced by two orders of magnitude. This disruptive cost advantage significantly lowers the procurement and deployment threshold for TCM colleges, vocational training institutions, and grassroots medical units, providing a solid economic foundation for the widespread dissemination and large-scale application of high-quality massage teaching resources, and significantly improving the technology's accessibility and affordability.

[0021] Third, in terms of teaching effectiveness and user experience, this invention constructs a multimodal immersive learning environment and a real-time closed-loop feedback mechanism, achieving a qualitative leap in teaching quality from "observable" to "perceptible, practiceable, and evaluable." Existing remote teaching solutions are mostly limited to one-way audio and video transmission, preventing students from obtaining crucial tactile information and lacking real-time guidance during practical operations. This invention, through precise mapping of gloves on both the teacher's and student's ends, enables students to synchronously perceive the tactile mirror image of the teacher's operation on corresponding parts of their hands, establishing an intuitive connection of "where force is applied, where sensation is felt." Crucially, through edge-localized real-time deviation comparison technology, the system can detect deviations in force, direction, and timing of students' operations with millisecond-level latency during practice, and immediately provide feedback through differentiated tactile vibration signals. This is equivalent to achieving remote and automated "hands-on" correction by a master teacher. Furthermore, the system can automatically and quantitatively score practical assessments, generating detailed reports and completely changing the traditional evaluation model that relies on subjective experience. This teaching loop, which integrates "visual observation, tactile perception, instant feedback, and objective evaluation," greatly enhances learning efficiency and depth.

[0022] Fourth, regarding system reliability and robustness, the "cloud-edge-hardware" three-tier collaborative architecture adopted in this invention effectively reduces the system's dependence on network stability and ensures the continuity of the teaching process. Traditional centralized architectures place a large number of computing tasks in the cloud, requiring extremely high network connectivity. This invention pushes core functions with stringent real-time requirements (such as feature event encoding / decoding, local deviation comparison, and instant haptic feedback triggering) down to the user terminal (edge) for processing, while the cloud focuses on resource management, data storage, and non-real-time analysis. This architecture allows critical follow-up and correction functions to operate independently and smoothly even in the event of network fluctuations or brief interruptions, and student operation data can be temporarily stored locally and synchronized after the network is restored. This design significantly enhances the system's adaptability and service continuity in complex network environments.

[0023] Fifth, in terms of industry empowerment and technological advancement, this invention transcends the scope of a single teaching tool, constructing a sustainable and evolving digital ecosystem of massage techniques. This provides strong support for the standardization and data-driven development of massage techniques. The four-dimensional standard technique library established by this invention, consisting of "feature event sequences + video + text analysis + anatomical atlas," is a crucial infrastructure for the digitization and standardization of massage techniques. Simultaneously, the massive amounts of teaching and operational data generated during system operation provide valuable data resources for big data-based teaching effectiveness evaluation, personalized learning path recommendation, and even clinical research on the correlation between "technique-symptom-efficacy." This not only serves current teaching but will also propel the entire massage discipline towards data-driven, evidence-based precision medicine.

[0024] In summary, through its unique technical concept and systematic implementation, this invention has produced positive effects and significant progress far exceeding existing technologies in several key aspects, such as tactile transmission efficiency, system manufacturing cost, teaching practice effectiveness, service reliability, and industry empowerment depth. It possesses outstanding substantive characteristics and significant advancements. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the overall architecture of the system of the present invention; Figure 2 The teacher's end collects a structural diagram and component connection diagram of the gloves; Figure 3 Provide the student with a diagram showing the structure and component connections of the glove; Figure 4 A microservice architecture diagram for a cloud platform; Figure 5 A typical workflow diagram for remote teaching and assessment in a system; Figure 6 This is a flowchart of the virtual time integrator algorithm in Example 2; Figure 7 This is a schematic diagram of the tactile feedback waveform in Example 2 (a is the first penetrating force waveform, b is the second penetrating force waveform). Figure 8 This is a logic block diagram for force regulation monitoring and warning in Example 3.

[0026] 100. Teacher-side data acquisition glove; 110. Flexible glove substrate one; 120. Pressure sensing module; 130. Posture sensing module; 140. First microcontroller; 150. First wireless communication module; 160. Surface electromyography signal sensing module; 200. Edge end; 210. Feature event encoding / decoding module; 220. Local deviation comparison module; 230. Audio and video synchronization module; 300. Cloud platform; 400. Student-side feedback glove; 410. Flexible glove substrate two; 420. Tactile feedback array; 421. Tactile feedback actuator; 430. Dedicated tactile alert device; 440. Second microcontroller; 450. Second wireless communication module; 460. Acoustic prompting module. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0028] Example 1, see Figure 1The intelligent massage technique cloud platform and remote teaching system described in this invention adopts a three-tiered distributed architecture: hardware, edge, and cloud platform. The hardware layer, serving as the system's physical sensing and execution layer, includes a teacher-side data acquisition glove 100 and a student-side feedback glove 400. The edge layer 200 runs as client software on the teacher's and student's terminal devices (PCs, tablets, smartphones) and is the system's local intelligent processing unit. The cloud platform 300, deployed on public cloud infrastructure, is the system's control center and resource hub. All three components interact via wireless communication protocols (Bluetooth BLE 5.0) and standard network protocols.

[0029] See Figure 2 The teacher-side data collection glove 100 specifically includes a flexible glove substrate 110, a pressure sensing module 120, an attitude sensing module 130, a first microcontroller 140, and a first wireless communication module 150.

[0030] The flexible glove substrate 110 is made of elastic spandex blend material, which has good elasticity and breathability. The inner side of the main contact area of ​​the palm (thumb pad, thenar eminence, palm center) is laminated with a medical-grade silicone layer with a thickness of 0.5mm to 1.0mm to increase the coefficient of friction, simulate the feel of real operation, and provide a uniform force surface for the pressure sensor.

[0031] The pressure sensing module 120 includes multiple flexible thin-film pressure sensors, which are precisely attached to the thumb pad, thenar eminence, and center of the palm of the flexible glove substrate 110, respectively. Each pressure sensor is connected to a signal conditioning circuit, typically a series voltage divider circuit, which converts the sensor's resistance change into a voltage signal output. The analog-to-digital converter integrated within the first microcontroller 140 acquires this voltage signal at a predetermined sampling frequency. Before leaving the factory, the system calibrates each sensor using a standard pressure calibration device, setting voltage thresholds for three pressure levels: light, medium, and heavy (e.g., corresponding to pressure values ​​of 50-80g, 150-200g, and 300-350g, respectively). The firmware of the first microcontroller 140 operates in an interrupt-driven manner: when the ADC sampled value continuously exceeds a certain threshold for a set anti-shake time, a pressure level characteristic event is determined to have occurred, and the event type and precise timestamp are recorded.

[0032] The attitude sensing module 130 is preferably a six-axis inertial measurement unit (IMU), which is rigidly fixed in a dedicated compartment on the back of the glove hand. This IMU is used to detect the three-dimensional attitude angles of the hand (pitch, roll, and yaw). The first microcontroller 140 reads the raw data from the IMU via the I2C bus and runs a sensor fusion algorithm (complementary filtering algorithm or Kalman filtering algorithm) to obtain a stable, drift-free attitude estimate. When the algorithm detects that the hand rotation angle exceeds a preset threshold and the angular velocity exceeds a set value, it determines that a directional inflection point feature event has occurred.

[0033] The first wireless communication module 150 is preferably a Bluetooth Low Energy module, which is connected to the first microcontroller 140 via a UART interface. When a feature event is detected, the first microcontroller 140 packages the event type, associated sensor identifier, force level or direction information, and high-precision timestamp into a lightweight data frame in a custom format, and sends it to the teacher's terminal device (i.e., edge device 200) via the first wireless communication module 150.

[0034] See Figure 3 The student-end feedback glove 400 includes a second flexible glove substrate 410, a tactile feedback array 420, a second microcontroller 440, and a second wireless communication module 450.

[0035] The second flexible glove substrate 410 is made of a material similar to that of the teacher's end, but the silicone friction layer can be omitted.

[0036] The tactile feedback array 420 consists of multiple miniature tactile feedback actuators 421, which are preferably linear resonant actuators or miniature vibration motors. Crucially, the distribution of the tactile feedback actuators 421 strictly corresponds to the arrangement of the pressure sensing module 120 on the teacher's end-feedback glove 100. That is, actuators are also installed at the thumb pad, thenar eminence, and palm of the student's glove. Each actuator is connected to a drive circuit, which is controlled by the second microcontroller 440.

[0037] The second microcontroller 440 receives instructions from the student terminal (edge ​​terminal 200) via the second wireless communication module 450 (also a BLE module). When an event code characterizing the teacher's technique is received, the second microcontroller 440 parses the instruction and controls the tactile feedback actuator 421 at the corresponding position to generate vibrations of a specific intensity and pattern (e.g., vibration intensity controlled by a PWM wave), so that the student can perceive the position and intensity of the teacher's force on the corresponding part of their hand, forming a precise tactile mirror. When a deviation correction instruction is received, the actuator is driven to generate a prompting vibration pattern that differs from the mirror feedback.

[0038] The edge terminal 200 runs as client software on the user terminal, and its core functions are implemented by the following software modules: Feature Event Encoding / Decoding Module 210: This module defines a streamlined communication protocol for converting raw signals acquired by the hardware into lightweight feature event codes. For example, a 16-bit custom encoding format is used, including device type, force / direction information, low-order timestamp bits, and check bits. This encoding significantly reduces the data volume, allowing ordinary wireless networks to meet real-time transmission requirements. The decoding process, conversely, parses the received feature event codes into executable feedback instructions.

[0039] Local Deviation Comparison Module 220: This module is the core of achieving low-latency real-time feedback. In student practice or assessment modes, this module compares the real-time characteristic event sequence generated by the student with the teacher's standard characteristic event sequence pre-downloaded from the cloud. The comparison algorithm uses a time window comparison method, setting a dynamic tolerance window (e.g., ±200ms) for each expected event. If the student's event does not match the standard sequence in time, type, or location, it is immediately determined as an operational deviation (such as timing deviation, type deviation, or omission deviation). Crucially, after deviation determination, the generation and sending of feedback instructions are completed entirely locally, without uploading to the cloud and then distributing, thus minimizing feedback latency (typically less than 50ms) and ensuring that core teaching functions remain available even during network interruptions.

[0040] Audio / Video Synchronization Module 230: This module is responsible for coordinating the synchronized playback of the video stream and haptic feedback signals on the teacher's end. It uses the Network Time Protocol to synchronize the time on each end and adds synchronization timestamps to video frames and feature events. On the student's end, the player precisely triggers the corresponding haptic feedback based on the timestamp of the current video frame, ensuring a perfect match between the image the student "sees" and the touch they "feel".

[0041] See Figure 4 The cloud platform 300 adopts a microservice architecture and mainly includes the following service modules: User and Permission Management Module: Based on the RBAC (Role-Based Access Control) model, it manages the accounts, permissions, and course associations of different roles such as administrators, teachers, students, and clinicians.

[0042] Standard Technique Library Module 310: Stores structured digital resources for massage techniques. Each standard technique includes a sequence of characteristic events (core digital standards), high-definition instructional videos (shot from multiple angles), textual explanations (key points of the movements, precautions), and SVG format anatomical atlases (annotated acupoints, muscles, and meridians), forming a multi-dimensional teaching resource encompassing "tactile-visual-theoretical" elements.

[0043] Remote Teaching Collaboration Module 320: Based on real-time communication protocols such as WebSocket, it manages remote teaching sessions initiated by teachers. It is responsible for receiving and forwarding audio and video streams and characteristic event codes on the teacher's end, as well as managing and monitoring the status of online students.

[0044] Intelligent Assessment Module 330: Provides fully automated management of the remote practical assessment process. Teachers can create assessment tasks and set scoring rules (such as a weighted scoring model based on hierarchical accuracy, directional accuracy, and temporal synchronization). The system automatically collects student assessment data, performs objective scoring according to the rules, and generates detailed scoring reports and learning portfolios.

[0045] Data Platform and Analytics Module: Constructs a unified data warehouse to integrate data from across the entire system. Utilizes big data analytics technologies (such as statistical analysis and machine learning algorithms) to evaluate teaching effectiveness, identify high-frequency error points, recommend personalized learning content to students, and provide data insights for optimizing clinical techniques.

[0046] Taking a typical remote teaching session as an example, the workflow is as follows: S01: Teachers wear teacher-end data collection gloves 100, log in to the client, create or select a course, and then initiate remote teaching.

[0047] S02: The teacher's end collects the hand technique feature signal in real time using the gloves 100, and the edge end 200 encodes it into feature event code, which is then uploaded to the cloud platform 300 along with the video stream.

[0048] S03: The cloud platform 300 will distribute audio and video streams and feature event codes to the learning participants' terminals in real time.

[0049] S04: The edge device 200 on the student's end receives data, plays the video synchronously, and drives the student's feedback glove 400 to generate corresponding tactile feedback, so that the student can perceive the teacher's operation.

[0050] S05: At the same time, the local deviation comparison module 220 of the edge terminal 200 of the student terminal compares the feature sequence generated by the student's own operation with the teacher's standard sequence in real time. Once a deviation is detected, a tactile prompt is immediately given through the student terminal feedback glove 400.

[0051] S06: Teaching process data and assessment results are continuously recorded and uploaded to the cloud platform 300 for analysis, generating various reports for teaching optimization.

[0052] Example 2, see Figure 6 In this embodiment, a virtual time integrator algorithm is embedded in the first microcontroller 140 of the teacher's end data acquisition glove 100. The execution steps of this algorithm are as follows: S701: System initialization. Set gravity threshold, first integral threshold I.level1 Second integration threshold I level2 (I) level2 >I level1 Initialize the integrator state to idle, integration register I. reg Reset to zero.

[0053] S702: Real-time monitoring of pressure sensor output. The first microcontroller 140 samples the pressure sensor signal at a fixed frequency.

[0054] S703: Determine the current pressure value P current Is it greater than F? threshold If not, proceed to S708; if yes, proceed to the points process.

[0055] S704: Determine the integrator status. If it is in an idle state, set the status to integrating and start the high-precision timer.

[0056] S705: Perform integration calculation. Calculate the force value exceeding the threshold at the current sampling point: ΔF = P current -F threshold Then update the integration register: I reg =I reg +ΔF*Δt, where Δt is the sampling time interval.

[0057] S706: Integral value judgment and event triggering. (The rest of the text appears to be incomplete and requires further context.) reg Compared with the preset threshold: If I reg ≥I level1 If the first-level event has not yet been triggered, then generate and send the first penetration force characteristic event code, and mark that the first level has been triggered.

[0058] If I reg ≥I level2 If the second-level event has not yet been triggered, then generate and send the second penetration force characteristic event code, and mark that the second level has been triggered.

[0059] S707: Continue monitoring the pressure value, repeat S705-S706, until the pressure value is below the threshold.

[0060] S708: When the pressure value P current <F threshold When the integrator is reset, set its state to idle and clear the integration register I. reg Reset the event trigger flag to prepare for the next press.

[0061] See Figure 7 On the student's end, when the penetrating force event code is received, the second microcontroller 440 drives the haptic feedback actuator 421 to generate a specific waveform. Figure 7a shows the first tactile feedback waveform 800a corresponding to the first penetrating force characteristic event code. Its envelope shows that the intensity gradually increases from zero to the peak region A1 in stage T1, maintains the peak region A1 in stage T2, and then gradually decays to zero in stage T3. Figure 7 Figure b shows the second tactile feedback waveform 800b corresponding to the second penetrating force characteristic event code. It starts as a short pulse with a high amplitude P1, and then transitions to a stable segment P2 with a longer duration but slightly lower amplitude. Through the significant difference between the first tactile feedback waveform 800a and the second tactile feedback waveform 800b, the trainee can clearly distinguish the different degrees of penetrating force sensation.

[0062] In Example 3, the teacher's end-feeding glove 100 incorporates a surface electromyography (EMG) signal sensing module 160 in its wrist module. The EMG signal sensing module 160 includes at least one pair of differential electrodes and a reference electrode, integrated into a medical-grade silicone pad on the inner side of the wristband to ensure good contact with the skin. The electrodes are preferably attached to the muscle belly of the forearm flexor muscles. The EMG signal sensing module 160 also includes a pre-amplifier and filter circuit for preliminary processing of weak EMG signals.

[0063] See Figure 8 It demonstrates the logical process of strengthening standardized monitoring and warning: S801: Signal Acquisition and Preprocessing. The surface electromyography signal sensing module 160 acquires the raw electromyography signal, which is then amplified, bandpass filtered, and subjected to power frequency notch filtering before being sampled by the ADC of the first microcontroller 140.

[0064] S802: Feature extraction. The first microcontroller 140 calculates the root mean square (RMS) value of the signal from the surface electromyography (EMG) sensor module within the sliding time window and normalizes this RMS value as a percentage of the wearer's maximum voluntary contraction force.

[0065] S803: Standardized Comparison. The local deviation comparison module 220 of the edge terminal 200 compares the real-time calculated %MVC value with the standard electromyographic activity range corresponding to the current learning technique downloaded from the cloud platform 300. This standard range defines the relaxation threshold, the reasonable range of the exertion phase, and the recovery time requirement.

[0066] S804: Deviation Judgment. If the real-time %MVC value is consistently higher than the relaxation threshold during the relaxation period, or the activity pattern during the exertion period does not conform to the standard, or the muscle recovery is slow after the technique ends, it is judged as a deviation in exertion standardization.

[0067] S805: Generate and send deviation signal. Once a deviation is determined, the local deviation comparison module 220 generates a force application standard deviation signal and sends it to the student end through the edge terminal 200.

[0068] S806: Specific Tactile Alert. Upon receiving the deviation signal, the second microcontroller 440 of the student-side feedback glove 400 drives a dedicated tactile alert 430 (such as an eccentric rotor motor) on the back of the wrist to generate a low-frequency, irregular vibration signal. This alert signal has a completely different tactile texture from the vibrations used by the tactile feedback array 420 in the palm area to characterize the force applied in hand movements. Therefore, without interfering with the main learning task, it effectively reminds the student to adjust their force application method.

[0069] See Figure 6 The teaching method adopted in this invention includes the following steps in sequence: The first stage, teaching preparation and resource construction, is a collaborative effort between the system administrator and the instructor. The administrator initializes the system on the cloud platform, including creating teacher and student accounts, assigning role permissions, and establishing virtual teaching classes. The instructor then designs the course based on a pre-built standard technique library on the cloud platform. This library stores characteristic event sequences of various massage techniques, high-definition demonstration videos, textual explanations, and anatomical atlases. The instructor selects target teaching content from the library, sets teaching parameters including demonstration duration, practice mode, deviation tolerance threshold, and assessment standards, thereby generating a structured electronic course.

[0070] Second, the multimodal remote teaching and synchronous perception stage is the core of the teaching process. The instructor logs into the system and initiates a remote teaching session. Wearing a teacher-end data acquisition glove, the instructor performs standard techniques on the object being manipulated. The glove collects the mechanical and kinematic signals generated by the operation in real time and converts the continuous signals into discrete force level feature events and direction inflection point feature events through a built-in feature event extraction algorithm. Optionally, it can also generate penetration force level feature events and force application standard feature events. These feature events are encoded into lightweight data frames and distributed in real time to the online students' terminal devices via a cloud platform, along with the synchronously acquired teacher operation video stream. Students wear student-end feedback gloves, and their terminal devices receive the feature event codes and video stream. Through an audio-visual synchronization module, the tactile feedback array of the feedback gloves generates tactile stimuli corresponding to the teacher's operation, while simultaneously playing the teaching video, providing students with a multimodal immersive sensory experience that synchronizes vision and touch.

[0071] Third, in the real-time follow-up practice and localized deviation correction stage, the system activates real-time follow-up mode while the student perceives the teacher's standard operation. The student imitates the teacher's techniques, and the sensor built into the feedback glove they wear simultaneously collects the student's operation data. The edge processing module running on the student's terminal device compares the student's real-time generated feature event sequence with the pre-downloaded teacher's standard feature event sequence in real time through its local deviation comparison unit. The comparison uses a time window comparison method. If a deviation is detected in the student's operation in terms of event type, trigger time, or spatial location, the comparison unit immediately generates a correction instruction and sends it directly to the student's feedback glove. After receiving the instruction, the feedback glove drives a specific tactile actuator to generate a warning feedback that is different from the teaching tactile mirror, thereby providing the student with millisecond-level instant correction guidance without relying on cloud communication.

[0072] Fourth, in the teaching assessment and data-driven optimization stage, after the teaching unit ends, teachers create practical assessment tasks through the system's intelligent assessment module; students complete the assessment within a set time, and the system automatically collects the characteristic event sequence and video of their entire operation process; the assessment module automatically scores the students' operation in terms of hierarchical accuracy, directional accuracy, temporal synchronization, and operational completeness based on a preset weighted scoring model, and generates an assessment report containing detailed scores and visual playback of deviation points; all teaching processes and assessment data are uploaded to the cloud platform's data middle platform; the middle platform uses big data analysis technology to analyze group learning data to optimize teaching strategies and analyzes individual student data to build their learning profiles, and then uses recommendation algorithms to push personalized reinforcement training content to them, forming a closed loop of data feedback driving continuous improvement in teaching effectiveness.

[0073] The beneficial effects of this invention lie in transforming the abstract teaching of massage techniques into a quantifiable, traceable, and reproducible standardized process through the aforementioned coherent, technology-driven teaching steps. This method not only breaks through the spatial limitations of traditional teaching but, more importantly, significantly improves the quality and efficiency of remote teaching through precise tactile transmission and immediate feedback mechanisms, as well as data-driven personalized guidance. It provides an effective technical path for the large-scale, standardized transmission of massage techniques.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A cloud platform and remote teaching system for intelligent massage techniques, characterized in that, include: The teacher-side data acquisition glove (100) is used to acquire characteristic signals during the teacher's massage operation. The characteristic signals include pressure signals applied to the object being massaged and signals indicating changes in the direction of hand movements. The student-side feedback glove (400) is used to receive feature event codes and convert them into perceptible tactile feedback signals; Edge terminal (200), deployed on teacher terminal and student terminal, is used to encode / decode the feature signals and realize local real-time comparison of student operation data and standard technique data; The cloud platform (300) is communicatively connected to the edge terminal (200) and is used to manage user information, standard method library, teaching collaboration process, assessment data and system data; The teacher-side data acquisition glove (100), the edge device (200), the cloud platform (300), and the student-side feedback glove (400) are sequentially connected to form a closed-loop remote teaching system from manual technique acquisition to tactile feedback.

2. The intelligent massage technique cloud platform and remote teaching system according to claim 1, characterized in that, The teacher-side data collection gloves (100) include: Flexible glove substrate 1 (110); At least one pressure sensing module (120) is disposed at the corresponding position of the palm and finger area of ​​the flexible glove substrate (110) for directly detecting the pressure applied to the object during the massage operation and generating a force level signal; At least one attitude sensing module (130) is disposed at the corresponding position on the back of the hand or wrist of the flexible glove substrate (110) for detecting changes in the angle and direction of hand movement and generating a direction change signal; The first microcontroller (140) is fixed to the wrist or back of the hand of the flexible glove substrate (110), and is electrically connected to the pressure sensing module (120) and the posture sensing module (130). It is configured to collect the force level signal and the direction change signal and assemble them into a feature data frame. The first wireless communication module (150), connected to the first microcontroller (140), is used to send the feature data frame to the edge terminal (200) where the teacher terminal is located.

3. The intelligent massage technique cloud platform and remote teaching system according to claim 2, characterized in that, The pressure sensing module (120) is a flexible thin-film pressure sensor, which is attached to the thumb pad, thenar eminence or palm area of ​​the flexible glove substrate (110); the attitude sensing module (130) is a six-axis inertial measurement unit, used to detect the pitch angle, roll angle and yaw angle of the hand.

4. The intelligent massage technique cloud platform and remote teaching system according to claim 1, characterized in that, The student-side feedback gloves (400) include: Flexible glove substrate 2 (410); The tactile feedback array (420) consists of multiple tactile feedback actuators (421) distributed at different locations on the flexible glove substrate (410); The second microcontroller (440) is fixed to the wrist or back of the hand of the flexible glove base (410) and is used to receive feature event codes from the edge end (200) where the student terminal is located, and drive the tactile feedback array (420) to generate corresponding tactile feedback. The haptic feedback array (420) is used to generate perceptible haptic cue signals under the drive of the second microcontroller (440); The second wireless communication module (450), connected to the second microcontroller (440), is used to receive data from the edge terminal (200).

5. The intelligent massage technique cloud platform and remote teaching system according to claim 4, characterized in that, The tactile feedback actuator (421) in the tactile feedback array (420) is a micro vibration motor or a linear resonant actuator; the distribution position of the tactile feedback actuator (421) corresponds to the arrangement position of the pressure sensing module (120) on the teacher's end collection glove (100), so that the tactile feedback direction perceived by the student is consistent with the part of the teacher exerting force.

6. The intelligent massage technique cloud platform and remote teaching system according to claim 5, characterized in that, The student feedback glove (400) also includes an acoustic prompting module (460) disposed on the flexible glove base (410). When the system detects a directional deviation in the student's operation, the second microcontroller (440) drives the acoustic prompting module (460) to emit a prompting sound.

7. The intelligent massage technique cloud platform and remote teaching system according to claim 1, characterized in that, The edge end (200) includes: The feature event encoding / decoding module (210) is used to encode the acquired feature signals into lightweight feature event codes and decode the received feature event codes into drive instructions; The local deviation comparison module (220) is used to compare the real-time characteristic event sequence generated by the trainee with the pre-stored standard characteristic event sequence in real time during the trainee's follow-up practice or assessment, and generate deviation information. The audio-visual synchronization module (230) is used to match the timestamp of the video stream with the timestamp of the feature event code to ensure the synchronization of video playback and haptic feedback.

8. The intelligent massage technique cloud platform and remote teaching system according to claim 7, characterized in that, The local deviation comparison module (220) adopts the time window comparison method and sets a time tolerance window. If the triggering time of the student's feature event exceeds the time tolerance window, or the feature type does not match the standard event, it is determined to be an operation deviation. After detecting the deviation, the edge end (200) directly sends a tactile feedback command to the student end feedback glove (400) to achieve low-latency local deviation correction.

9. The intelligent massage technique cloud platform and remote teaching system according to any one of claims 1-8, characterized in that, The system's remote teaching method for massage techniques includes: Step S1: The teacher wears the teacher-end collection gloves (100) to perform massage operations. The system directly collects the manual feature signals applied to the object of operation, and uploads them to the cloud platform (300) after encoding by the edge end (200) to build or update the standard manual library. Step S2: The student selects learning content from the cloud platform (300), and the edge terminal (200) downloads the corresponding standard feature event sequence; Step S3: During remote teaching or practice, the real-time feature event codes of the teacher's end are synchronized to the student's end via the cloud platform (300); Step S4: The student wears the student-end feedback glove (400), and its edge end (200) receives the real-time feature event code and drives the tactile feedback array (420) on the student-end feedback glove (400) to generate corresponding tactile feedback, so that the student can perceive the teacher's operating force and rhythm in the corresponding part of the hand. Meanwhile, the edge end (200) compares the characteristic event sequence generated by the student's operation with the standard sequence in real time. Once a deviation is found, it immediately provides corrective tactile prompts to the student through the student end feedback glove (400).

Images