Adjustable flexible shell for cooperation and soft robot and control system

Abstract

An adjustable flexible shell used for cooperating with a soft robot is integrally formed by shape memory alloy or intelligent polymer materials, the flexible shell wraps the outer surface of a robot body, a heating wire array is embedded in the flexible shell, and the heating wire array is distributed in a grid shape and fixed in the shell. The flexible shell is provided with a plurality of electric heating wires, the electric heating wires are connected to an external control unit through wires respectively, the flexible shell is further provided with a plurality of partition structures, each partition corresponds to an independent electric heating wire area and is provided with a local temperature feedback element, and after the flexible shell is powered on and heated, local deformation or overall shape reconstruction is achieved. The shell disclosed by the invention adopts a polyester-based polyurethane and shape memory polymer composite system, so that the material has a modulus adjustable range near a glass transition temperature, and excellent ductility, tear resistance and fatigue resistance of a polyurethane matrix are retained, so that the shell can still maintain structural integrity in a repeated soft-hard switching process, and the service life of the shell is prolonged. Embrittlement or cracking caused by phase change circulation is avoided.

Description

Technical Field

[0001] This invention belongs to the field of robot shell control technology, specifically an adjustable flexible shell and control system for collaborative and soft robots. Background Technology

[0002] With the rapid development of collaborative robotics and soft robotics technologies, robots are increasingly being used in unstructured environments, such as medical assistance, home services, and confined space exploration. In these tasks, robots not only need high flexibility and environmental adaptability, but also must meet stringent human-robot safety interaction requirements. While traditional rigid robot shells provide good structural protection, they lack deformation capabilities and are prone to jamming or collision damage in confined spaces. Existing soft robots mostly use homogeneous shells made of passively flexible materials (such as silicone and TPU), which, while possessing a certain degree of compliance, cannot actively adjust their stiffness or shape according to task requirements.

[0003] Currently, some research attempts to introduce smart materials (such as shape memory alloys (SMA) or shape memory polymers (SMP)) to achieve active deformation of the shell. However, existing technologies generally suffer from the following technical problems:

[0004] First, most solutions use only pure SMP or low-content composite materials. Although they can achieve modulus change near the phase transition temperature, repeated heating-cooling cycles can easily cause material fatigue, microcracks, or even brittle fracture. Especially under high-frequency task switching, the shell life is significantly shortened, which cannot meet the industrial-grade reliability requirements.

[0005] Second, existing flexible shells mostly employ overall heating or uniform temperature control in a few large areas, failing to achieve localized differentiated stiffness distribution. For example, when performing complex tasks such as "softening the front end for passage and hardening the rear end for support," traditional structures can only soften or harden globally, sacrificing operational efficiency or structural safety. Simultaneously, control strategies generally rely on preset rules or open-loop commands, unable to dynamically adjust heating parameters based on real-time environmental feedback (such as sudden changes in contact force or obstacle position shifts), leading to delayed response or even failure.

[0006] Third, although some soft robots integrate pressure or temperature sensors, the sensor network layout is unreasonable, the signal transmission is unreliable (e.g., rigid wires are prone to breakage under large deformations), and the perceived data is not effectively associated with the semantics of the high-level task. The control structure often operates independently and cannot receive task type signals from the motion planner (e.g., "precision operation" or "heavy-duty handling"), resulting in a mismatch between the shell state and the actual operation requirements. Summary of the Invention

[0007] In view of the above situation and to overcome the defects of the prior art, the present invention provides an adjustable flexible shell and control system for collaborative and soft robots, so as to at least partially solve the above technical problems.

[0008] The technical solution adopted in this invention is as follows:

[0009] This invention proposes an adjustable flexible shell for collaborative and soft robots, comprising:

[0010] The flexible shell is integrally formed from shape memory alloy or smart polymer material. The flexible shell covers the outer surface of the robot body. An array of heating wires is embedded in the flexible shell. The array of heating wires is distributed in a grid and fixed inside the shell. Each heating wire is connected to an external control unit through a wire.

[0011] The flexible shell is also provided with several partitioned structures, each partition corresponding to an independent heating wire area and equipped with a local temperature feedback element. After the flexible shell is heated by electricity, the thickness changes or stiffness is adjusted due to the phase change of the material, thereby realizing local deformation or overall shape reconstruction.

[0012] In one embodiment of the present invention, the flexible shell is made of a composite material of polyester-based polyurethane and shape memory polymer, wherein the shape memory polymer content is not less than 30 wt%.

[0013] In one embodiment of the present invention, the heating wire array is a nickel-chromium alloy wire or a carbon nanotube film, which is arranged in a crisscross pattern along the axial and circumferential directions of the flexible shell to form multiple independent heating units. Each heating unit is electrically connected to the control unit through a multi-channel relay module to achieve independent temperature control in different zones.

[0014] In one embodiment of the present invention, the inner surface of the flexible shell is provided with a buffer layer, the buffer layer being made of silicone or thermoplastic elastomer with a thickness of 0.5-2 mm, and being bonded to the flexible shell by bonding or co-extrusion.

[0015] In one embodiment of the present invention, the outer surface of the flexible shell is coated with a smart responsive coating, the coating comprising microencapsulated phase change material or pressure-sensitive color-changing polymer, for assisting environmental perception or human-computer interaction prompts.

[0016] In one embodiment of the present invention, a control system for an adjustable flexible shell for a collaborative and soft robot includes:

[0017] The main controller, located inside the robot body, is used to run the shell state control algorithm;

[0018] The sensing module includes a pressure sensor, a temperature sensor, and an acceleration sensor arranged on the surface and inside the flexible housing. The sensing module is communicatively connected to the main controller via a flexible circuit board.

[0019] The drive module includes a multi-channel constant current source and a relay array. The drive module receives the PWM signal output by the main controller and provides zoned power supply to the heating wire array inside the flexible housing accordingly.

[0020] The power management unit, integrated into the robot body, is used to provide stable power to the main controller, sensing module and drive module.

[0021] In one embodiment of the present invention, the main controller has a built-in state decision model based on reinforcement learning. This model dynamically generates the target temperature curve of each zone heating wire based on the environmental data and task instructions collected in real time by the sensing module, and adjusts the actual output power through PID closed-loop control.

[0022] In one embodiment of the present invention, the pressure sensor is configured as a flexible piezoresistive sensor, embedded in multiple key stress areas of the flexible shell, and each pressure sensor is connected to the flexible circuit board via silver paste wires; the temperature sensor is a miniature thermistor, attached to the center of each heating wire section; the acceleration sensor is a triaxial MEMS inertial measurement unit, fixed on the robot body skeleton and linked with the flexible shell.

[0023] In one embodiment of the present invention, the drive module further includes an over-temperature protection circuit, which is connected in series with each heating wire zone and automatically cuts off the power supply to the zone when a local temperature exceeds a preset threshold; the preset threshold is set according to the glass transition temperature or austenitic phase transformation temperature of the smart material used.

[0024] In one embodiment of the present invention, the main controller is connected to the robot's main motion control system via a CAN bus or EtherCAT interface, receives task type signals, and switches the shell control strategy accordingly.

[0025] The control strategy includes: in precision operation mode, activating the hardening program in the forearm area of ​​the shell; in heavy-duty transport mode, simultaneously thickening the torso and support joint areas; and in narrow space traversal mode, sequentially softening the front, middle and rear shell sections according to the path planning sequence.

[0026] The beneficial effects of the technical solution of this invention are as follows:

[0027] The outer shell of this invention adopts a composite system of polyester-based polyurethane and shape memory polymer, which ensures that the material has an adjustable modulus range near the glass transition temperature, while retaining the excellent ductility, tear resistance and fatigue resistance of the polyurethane matrix. This allows the shell to maintain structural integrity during repeated soft-hard switching, avoiding embrittlement or cracking caused by phase change cycles.

[0028] This invention uses an array of heating wires arranged in a grid pattern along the axial and circumferential directions to form a heating unit. In conjunction with a multi-channel relay module, it achieves independent temperature control of zones with millimeter-level spatial resolution. The reinforcement learning state decision model built into the main controller breaks through the limitations of traditional pre-programmed control. It can dynamically generate target temperature curves for each zone based on real-time sensor data (pressure distribution, temperature field, motion state) and high-level task instructions (such as "precision assembly" and "heavy-load handling"). The output power is then precisely adjusted through a PID closed-loop control.

[0029] This invention utilizes a multimodal sensing network comprised of a flexible piezoresistive pressure sensor, a miniature thermistor, and a triaxial MEMS accelerometer. The pressure sensor is embedded in key stress areas and connected to a flexible circuit board via silver paste wires, ensuring signal integrity under large deformations. Temperature sensors monitor the center temperature of each heating zone, providing reliable feedback for closed-loop control. Accelerometers are fixed to the frame and linked to the shell, capturing overall dynamic behavior. All data is rapidly aggregated to the main controller via the flexible circuit board and deeply integrated with the robot's main motion control structure via CAN bus or EtherCAT, achieving seamless mapping from task semantics to the physical state of the shell.

[0030] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0031] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0032] Figure 1 This is a schematic diagram of the adjustable flexible shell for collaborative and soft robots proposed in an embodiment of the present invention;

[0033] Figure 2 This is a structural diagram of the control system for an adjustable flexible shell for collaborative and soft robots proposed in an embodiment of the present invention;

[0034] Figure 3 This is a schematic diagram of the integration of the adjustable flexible shell for collaborative and soft robots proposed in an embodiment of the present invention with the robot system. Detailed Implementation

[0035] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0036] An adjustable flexible shell and control system for collaborative and soft robots according to an embodiment of the present invention will now be described with reference to the accompanying drawings.

[0037] like Figures 1 to 3 As shown, this embodiment of the invention provides an adjustable flexible shell for collaborative and soft robots, comprising: a flexible shell integrally formed from shape memory alloy or smart polymer material, the flexible shell covering the outer surface of the robot body, an array of heating wires embedded in the flexible shell, the array of heating wires being distributed in a grid pattern and fixed inside the shell, and each heating wire being connected to an external control unit via a wire, the flexible shell also having several partition structures, each partition corresponding to an independent heating wire area, and configured with a local temperature feedback element, the flexible shell undergoes thickness changes or stiffness adjustments due to material phase changes after being energized and heated, thereby achieving local deformation or overall shape reconstruction.

[0038] In practical applications of this invention, when the robot starts and enters the working state, the external control unit (usually integrated into the robot's main control structure) generates heating commands for each shell section based on the current task type (e.g., "crossing narrow pipes," "grabbing fragile items," or "carrying heavy objects") and real-time sensing data (from pressure, acceleration, and vision sensors). These commands are output to the drive module in the form of pulse width modulation (PWM) signals. The drive module then heats the heating wires of the designated sections using a multi-channel constant current source. Because the heating wires and the flexible shell material are tightly bonded at the microscopic level, heat is rapidly conducted to the surrounding smart materials, causing their local temperature to rise to the phase transition threshold (e.g., the glass transition temperature Tg of SMP, or the austenite initiation temperature As of SMA). Once the critical point is reached, the material undergoes a reversible phase transition: in shape memory polymers, this manifests as increased molecular chain segment mobility, macroscopically exhibiting softening, stretching, or shrinkage; in shape memory alloys, it triggers a martensitic-austenitic phase transition, inducing a driving force to restore the preset shape, resulting in localized thickening, hardening, or directional bending of the shell.

[0039] Meanwhile, based on the mechanism of independent temperature control in different zones, the flexible shell can exhibit differentiated mechanical states in different areas at the same time. For example, when the robot arm needs to pass through a hole with a diameter of only 50mm, the control structure first activates the heating circuit of the front shell zone, softening it and reducing its thickness (e.g., from 3mm to 2mm), while maintaining the rigidity of the torso to support the internal transmission mechanism. When performing high-load handling tasks, the heating wires of multiple load-bearing areas in the shoulder, elbow, and wrist are activated simultaneously, hardening the shell material and slightly expanding it (e.g., increasing it to 5mm), improving structural rigidity and impact resistance, and effectively protecting the internal motors, reducers, and cables from overload damage. In addition, local temperature feedback elements continuously transmit the actual temperature of each zone back to the control unit, forming a closed-loop temperature control to prevent material performance degradation or failure due to overheating.

[0040] Furthermore, the flexible shell can be deeply coupled with the robot's own motion planning structure. For example, during the path planning phase, if a complex obstacle is detected ahead, the structure can pre-activate the softening program in the corresponding shell area, ensuring the robot has sufficient deformation margin before contacting the obstacle, thus achieving "active compliance" rather than "passive collision." Simultaneously, the shell area around the end effector can harden upon contact with the workpiece, forming a temporary rigid support surface, improving the accuracy of repeated positioning; after the task is completed, it immediately softens and resets, restoring compliant interaction characteristics.

[0041] In one possible implementation, the flexible shell is made of a composite material of polyester-based polyurethane and shape memory polymer, wherein the shape memory polymer content is not less than 30 wt%, and the heating wire array is made of nickel-chromium alloy wire or carbon nanotube film, which is arranged in a cross pattern along the axial and circumferential directions of the flexible shell to form multiple independent heating units. Each heating unit is electrically connected to the control unit through a multi-channel relay module to achieve independent temperature control in different zones.

[0042] In practical applications, when the robot performs a task, the control structure generates a zoned heating strategy based on the current working conditions (such as spatial constraints, load size, or human-machine interaction safety level). This strategy acts on an electrothermal network embedded within the shell—a network composed of nickel-chromium alloy wires or carbon nanotube films, arranged crisscrossingly along the shell's axial direction (e.g., from the shoulder to the wrist) and circumferentially (around the limb's circumference), forming a grid-like topological structure. Simultaneously, the network is oriented according to the principal deformation directions of key degrees of freedom in the robot's kinematic model, ensuring that each potential deformation area is covered by at least one independent heating unit. Each grid area enclosed by intersecting heating wires constitutes a functional heating unit, completely electrically isolated from each other. The heating units are connected to the central control unit via a multi-channel relay module. The relay module receives digital commands from the control unit, switching the power supply circuit of a specific unit on or off as needed, thereby achieving millisecond-level independent temperature control within each zone. For example, when the robotic arm needs to bend laterally to pass through an obstacle, the control structure only activates a few heating units on the outer arc surface, raising its local temperature to above the Tg of the SMP (e.g., 65°C). At this time, the composite material in the region changes from a glassy state to a highly elastic state, the modulus drops sharply, and it is easy to stretch and deform. Meanwhile, the inner and back regions remain in a low-temperature rigid state, providing the necessary structural support force to prevent overall collapse.

[0043] Furthermore, due to the reversibility of the SMP phase transition, once the task is completed or the obstacle is removed, the control unit immediately cuts off the power supply to the corresponding area. The shell rapidly cools down to below Tg through natural heat dissipation or active air cooling, the material re-hardens and "locks" in its current shape, providing a stable mechanical basis for subsequent operations. If it is necessary to restore the original geometry, the shape memory effect can be triggered by a pre-programmed global heating program—this relies on the permanent network structure endowed to the SMP during the molding stage, which allows it to spontaneously shrink back to its initial configuration at high temperatures.

[0044] In one possible implementation, the inner surface of the flexible shell is provided with a buffer layer, which is made of silicone or thermoplastic elastomer and has a thickness of 0.5-2 mm. The buffer layer is bonded to the flexible shell by bonding or co-extrusion. The outer surface of the flexible shell is coated with a smart response coating, which contains microencapsulated phase change material or pressure-sensitive color-changing polymer for assisting environmental perception or human-computer interaction prompts.

[0045] In practical applications, when the flexible shell deforms or adjusts its stiffness due to task requirements—for example, softening and shrinking locally when traversing narrow spaces, or hardening and thickening during heavy-duty operations—the embedded drive and sensing network within it induces volume changes and stress redistribution. Simultaneously, the buffer layer, composed of silicone or thermoplastic elastomer (TPE), possesses high damping, low modulus, and excellent energy absorption characteristics. It is firmly bonded to the flexible shell through bonding or co-extrusion processes, forming a continuous composite structure without the risk of interface delamination. During shell deformation, the buffer layer effectively isolates the micro-strain generated by the phase transition of the smart material from being transmitted to internal robot components (such as motors, encoders, cable bundles, or circuit boards), preventing fatigue damage or loosening of connections caused by repeated stretching-compression cycles. Furthermore, in the event of an accidental collision between the robot and an external object, the buffer layer can absorb impact energy, reduce peak force, and improve the safety of human-robot collaboration.

[0046] Meanwhile, the outer surface of the flexible shell is not a static covering layer, but a dynamic intelligent interface that interacts with the environment. Its intelligent responsive coating contains microencapsulated phase change materials (mPCMs) or pressure-sensitive chromic polymers. For example, when the robot's end effector contacts a high-temperature workpiece, the mPCMs in a localized area absorb heat and undergo a solid-liquid phase transition, stabilizing the coating surface temperature. Simultaneously, its thermochromic components trigger a color change (e.g., from blue to red), providing operators with an intuitive thermal risk warning. During precision assembly, if the applied contact force exceeds a preset threshold, the molecular configuration of the pressure-sensitive chromic polymer changes, causing a shift in the reflectance spectrum and revealing visible color spots in the contact area, thus achieving "force visualization." This assists operators in adjusting the force in real time to prevent workpiece damage.

[0047] Furthermore, the control structure not only adjusts the shell stiffness based on internal sensor data, but also incorporates color / temperature feedback from the outer coating for strategy correction. For example, when exploring an unknown environment, if the coating in a certain area changes color due to continuous pressure, the structure can determine that there is a potential obstacle and activate the softening program of the corresponding shell partition in advance, achieving integrated "perception-decision-response".

[0048] In one possible implementation, the control structure for the adjustable flexible shell of a collaborative and soft robot includes:

[0049] The main controller, located inside the robot body, is used to run the shell state control algorithm;

[0050] The sensing module includes a pressure sensor, a temperature sensor, and an acceleration sensor arranged on the surface and inside the flexible housing. The sensing module is connected to the main controller via a flexible circuit board.

[0051] The drive module includes a multi-channel constant current source and a relay array. The drive module receives the PWM signal output by the main controller and provides zoned power supply to the heating wire array inside the flexible housing accordingly.

[0052] The power management unit, integrated into the robot body, provides stable power to the main controller, sensing module, and drive module.

[0053] In practical applications of this invention, after the structure is started, the main controller first loads a shell state control algorithm preset for the current task scenario. This algorithm is not a static rule base, but a multimodal decision engine that integrates task type recognition, environmental risk assessment, and material dynamics models. Simultaneously, sensing modules distributed throughout the flexible shell surface and interior continuously collect multidimensional physical information: pressure sensors (typically flexible piezoresistive or capacitive arrays) monitor the magnitude and distribution of contact forces between different areas of the shell and the external environment in real time; temperature sensors (such as miniature thermistors or infrared thermometers) capture the actual temperature rise curve of the material after heating with a heating wire; and acceleration sensors (three-axis MEMS inertial units) sense the overall motion state of the robot, including vibration frequency, impact intensity, and attitude changes. The sensors are interconnected at high density via embedded flexible circuit boards, ensuring not only the reliability of signal transmission but also preventing the breakage of traditional rigid circuits due to repeated shell deformation, thus ensuring the continuity and integrity of the sensed data.

[0054] All raw sensor data is aggregated by the flexible circuit board and then transmitted via high-speed serial communication (such as I...). 2The data (via C, SPI, or a custom CAN protocol) is uploaded to the main controller in real time. The controller then performs data fusion processing: for example, if the pressure sensor detects abnormal high pressure in a certain area, and the acceleration sensor simultaneously records a violent impact, the structure determines it to be a sudden collision event and immediately triggers a local softening protection strategy; if the temperature sensor reports that the heating rate of a certain zone is too slow, the subsequent PWM duty cycle is automatically adjusted to compensate for heat loss. The decision results are output to the drive module in the form of digital control commands. The core of the drive module consists of a multi-channel constant current source and a solid-state relay array. It can receive high-resolution PWM signals generated by the main controller and implement millisecond-level, zone-independent power supply control for the heating wire array embedded in the flexible shell. Each heating wire zone corresponds to an independent drive channel, ensuring that different areas can synchronously execute differentiated thermal excitation—for example, the front area heats up rapidly to soften and pass through obstacles, while the rear area maintains low-temperature rigidity to support the load. The constant current source effectively overcomes the power fluctuation problem caused by the heating wire resistance changing with temperature, ensuring the linear controllability of the heating process; the relay array provides electrical isolation and overload protection to prevent local short circuits from affecting the overall structural safety.

[0055] In one possible implementation, the main controller has a built-in reinforcement learning state decision model. The model dynamically generates the target temperature curves of each heating wire zone based on the environmental data and task instructions collected in real time by the sensing module, and adjusts the actual output power through PID closed-loop control. The pressure sensor is set as a flexible piezoresistive sensor, which is embedded in multiple key stress areas of the flexible shell. Each pressure sensor is connected to the flexible circuit board through silver paste wires. The temperature sensor is a miniature thermistor, which is attached to the center of each heating wire zone. The acceleration sensor is a three-axis MEMS inertial measurement unit, which is fixed on the robot body skeleton and linked with the flexible shell.

[0056] In practical applications of this invention, after the structure is activated, the main controller first receives instructions from the robot's high-level task planning module (such as "grab a glass", "cross a pipe", or "human-robot handshake"), while continuously acquiring multi-dimensional real-time environmental data provided by the sensing module. The flexible piezoresistive pressure sensor is strategically embedded in the stress-bearing areas of the flexible shell—such as the end effector contact surface, the outer side of joint bends, and the torso load-bearing band—identified as high-risk or highly sensitive areas in historical task data. When the robot interacts with the outside world, the pressure sensor transmits the local contact force distribution to the flexible circuit board with millisecond-level response speed via silver paste wires, which are then aggregated to the main controller. Due to its excellent ductility and conductivity stability, the silver paste wires maintain signal integrity during repeated stretching and bending of the shell, avoiding sensing failure caused by fatigue fracture of traditional metal wires.

[0057] Meanwhile, miniature thermistors attached to the center of each heating zone continuously monitor the actual material temperature, providing high-precision local thermal feedback. A three-axis MEMS inertial measurement unit fixed to the robot's skeleton simultaneously captures the overall motion state, including acceleration abrupt changes, attitude tilt angles, and vibration spectra. Together, these data form a comprehensive understanding of the current operating conditions: pressure distribution reveals contact intensity and the point of impact; temperature data reflects the material's phase transition progress; and acceleration information characterizes external impacts or motion dynamics. The fusion of these three data points provides high-quality state input to the reinforcement learning state decision model in the main controller.

[0058] Furthermore, during each task execution, the model evaluates the long-term benefits of different temperature strategies based on the current state (such as comprehensive indicators like "operational success rate," "energy consumption," and "structural safety"), and dynamically outputs the optimal target temperature curve. For example, when grasping fragile items, the model prioritizes raising the target temperature of the fingertip area to slightly above the glass transition temperature (Tg) of the shape memory polymer, softening it to achieve a compliant envelope; and when it detects a tendency for the item to slip, it quickly lowers the area temperature setting, triggering local hardening to enhance the clamping force. The generated target temperature curve is then input to the independent PID closed-loop controller for each zone. The PID controller compares the target temperature with the actual temperature fed back by the thermistor in real time, calculates the required PWM duty cycle, and drives a multi-channel constant current source to apply precise power output to the corresponding heating wire zone. Since each zone is equipped with a dedicated temperature feedback and control loop, the structure can achieve millimeter-level spatial resolution thermal field control, allowing the flexible shell to exhibit a heterogeneous mechanical distribution of both rigidity and flexibility at the same time. For example, when traversing narrow passages, the front shell section is heated and softened to shrink and deform, while the central support area maintains low-temperature rigidity to protect the internal transmission mechanism. The non-uniform state is entirely dynamically planned by the reinforcement learning model based on real-time obstacle perception.

[0059] In one possible implementation, the drive module further includes an over-temperature protection circuit connected in series with each heating wire zone. This circuit automatically cuts off power to the zone when a local temperature exceeds a preset threshold. The preset threshold is set based on the glass transition temperature or austenitic phase transformation temperature of the smart material used. The main controller communicates with the robot's main motion control structure via a CAN bus or EtherCAT interface, receives task type signals, and switches shell control strategies accordingly. These strategies include: activating the hardening program in the forearm area of ​​the shell in precision operation mode; simultaneously thickening the torso and support joint areas in heavy-load handling mode; and sequentially softening the front, middle, and rear shell zones according to the path planning sequence in narrow space traversal mode.

[0060] In specific applications of this invention, when a collaborative or soft robot starts a task, the main controller first establishes a real-time data channel with the robot's main motion control structure through a high-speed industrial communication interface (such as CAN bus or EtherCAT). The channel not only transmits conventional position and speed commands, but also receives semantic task type signals from the high-level task planning module—such as "precision assembly," "heavy-duty handling," or "narrow space traversal."

[0061] In precision operation mode, the structure recognizes the task's requirements for end-effector stiffness and vibration suppression, and immediately activates the hardening program in the forearm and wrist areas: the drive module applies heating power to the corresponding heating wire zones, raising the local temperature slightly above the glass transition temperature (Tg) or austenitic phase transformation temperature (As) of the smart material, triggering a leap in material modulus and forming a temporary rigid support structure, thereby improving the accuracy of repeated positioning and suppressing micro-vibrations; while in heavy-duty handling mode, the structure simultaneously activates the electrothermal circuits in the key load-bearing areas of the torso, shoulder, and hip joints, causing the shell material to undergo a phase transformation... Controlled expansion or crystallization strengthening increases overall thickness and rigidity, effectively distributing the mechanical load of the internal transmission mechanism and preventing motor stalling or structural fatigue due to overload. When entering narrow space traversal mode, the structure softens the front detection area, middle transition section, and tail following area in a time-space sequence according to the spatial sequence output by the path planning algorithm, achieving "wave-like" deformation propulsion: the front softens and contracts first to probe into the gap, the middle deforms and follows, and the tail finally leaves its original position. The whole process is like biological peristalsis, greatly improving the robot's ability to pass through unstructured and restricted environments.

[0062] Throughout the entire control process, each heating wire zone is connected in series with an independent over-temperature protection circuit. The circuit employs dual redundancy: on one hand, the main controller implements software temperature limiting via a temperature sensor; on the other hand, a hardware-level comparator monitors the relationship between zone current and temperature in real time. If the local temperature exceeds a preset safety threshold (strictly set within the reversible phase transition window of the intelligent material, typically slightly below the material's thermal decomposition temperature, and dynamically calibrated based on the specific material's Tg or As), the over-temperature protection circuit will automatically cut off the zone's power supply within milliseconds to prevent material carbonization, coating peeling, or thermal runaway. After power failure, the structure records the faulty zone number and disables or degrades its use in subsequent tasks. Simultaneously, a warning message is sent to the host computer via the communication bus for timely intervention by maintenance personnel.

[0063] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0064] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. An adjustable flexible shell for collaborative and soft robots, characterized in that, include: The flexible shell is integrally formed from shape memory alloy or smart polymer material. The flexible shell covers the outer surface of the robot body. An array of heating wires is embedded in the flexible shell. The array of heating wires is distributed in a grid and fixed inside the shell. Each heating wire is connected to an external control unit through a wire. The flexible shell is also provided with several partitioned structures, each partition corresponding to an independent heating wire area and equipped with a local temperature feedback element. After the flexible shell is heated by electricity, the thickness changes or stiffness is adjusted due to the phase change of the material, thereby realizing local deformation or overall shape reconstruction.

2. The adjustable flexible shell for collaborative and soft robots according to claim 1, characterized in that, The flexible shell is made of a composite material of polyester-based polyurethane and shape memory polymer, wherein the shape memory polymer content is not less than 30 wt%.

3. The adjustable flexible shell for collaborative and soft robots according to claim 1, characterized in that, The heating wire array is made of nickel-chromium alloy wire or carbon nanotube film, and is arranged in a crisscross pattern along the axial and circumferential directions of the flexible shell to form multiple independent heating units. Each heating unit is electrically connected to the control unit through a multi-channel relay module to achieve independent temperature control in different zones.

4. The adjustable flexible shell for collaborative and soft robots according to claim 1, characterized in that, The inner surface of the flexible shell is provided with a buffer layer, which is made of silicone or thermoplastic elastomer, with a thickness of 0.5-2mm, and is bonded to the flexible shell by bonding or co-extrusion.

5. The adjustable flexible shell for collaborative and soft robots according to claim 1, characterized in that, The outer surface of the flexible shell is coated with a smart response coating, which contains microencapsulated phase change materials or pressure-sensitive color-changing polymers to assist in environmental perception or human-computer interaction prompts.

6. A control system for an adjustable flexible shell for collaborative and soft robots according to any one of claims 1-5, characterized in that, include: The main controller, located inside the robot body, is used to run the shell state control algorithm; The sensing module includes a pressure sensor, a temperature sensor, and an acceleration sensor arranged on the surface and inside the flexible housing. The sensing module is communicatively connected to the main controller via a flexible circuit board. The drive module includes a multi-channel constant current source and a relay array. The drive module receives the PWM signal output by the main controller and provides zoned power supply to the heating wire array inside the flexible housing accordingly. The power management unit, integrated into the robot body, is used to provide stable power to the main controller, sensing module and drive module.

7. The control system for the adjustable flexible shell of a collaborative and soft robot according to claim 6, characterized in that, The main controller has a built-in reinforcement learning state decision model. Based on the environmental data and task instructions collected in real time by the sensing module, the model dynamically generates the target temperature curve of the heating wire in each zone and adjusts the actual output power through PID closed-loop control.

8. The control system for the adjustable flexible shell of a collaborative and soft robot according to claim 6, characterized in that, The pressure sensor is a flexible piezoresistive sensor, embedded in multiple key stress areas of the flexible shell. Each pressure sensor is connected to the flexible circuit board via silver paste wires. The temperature sensor is a miniature thermistor, attached to the center of each heating wire section. The acceleration sensor is a triaxial MEMS inertial measurement unit, fixed to the robot's frame and linked with the flexible shell.

9. The control system for the adjustable flexible shell of a collaborative and soft robot according to claim 6, characterized in that, The drive module also includes an over-temperature protection circuit, which is connected in series with each heating wire zone and automatically cuts off the power supply to the zone when the local temperature exceeds a preset threshold. The preset threshold is set according to the glass transition temperature or austenitic phase transformation temperature of the smart material used.

10. The control system for the adjustable flexible shell of a collaborative and soft robot according to claim 6, characterized in that, The main controller communicates with the robot's main motion control system via a CAN bus or EtherCAT interface, receives task type signals, and switches the shell control strategy accordingly. The control strategy includes: activating the hardening process in the forearm region of the outer shell in precision operation mode; In heavy-duty transport mode, the torso and supporting joint areas are thickened simultaneously; in narrow space traversal mode, the front, middle and rear shell sections are softened sequentially according to the path planning order.

Images