An exoskeleton fingertip integrating fingertip sensing and vibration feedback capabilities

Abstract

This invention discloses an exoskeleton fingertip integrating fingertip sensing and vibration feedback capabilities, providing stable fingertip pressure sensing and corresponding vibration feedback. The fingertip interaction force sensing component consists of a rigid fingertip pressing plate, a soft silicone pad, and a piezoresistive thin-film sensor array. The vibration feedback component of the fingertip module provides vibration feedback via a vibration motor. By controlling the frequency, amplitude, and duty cycle of the vibration motor, the magnitude of tangential force, sliding speed, and texture sensations during interaction are simulated. This design provides more stable fingertip force and tactile feedback for the hand exoskeleton, enabling more stable and precise hand teleoperations and improving the user experience.

Description

Technical Field

[0001] This invention relates to an underactuated force feedback exoskeleton, and more particularly to an exoskeleton fingertip that integrates fingertip sensing function and vibration feedback capability. Background Technology

[0002] Hand exoskeletons are essentially robotic technologies worn on the human hand, combining exoskeleton bionics and information technology. The hand is one of the most important motor organs in humans; most daily tasks are performed using dexterous hands, such as dressing, carrying plates, and picking up food. A large portion of the information in human interaction with the environment comes from tactile perception, which helps humans achieve finer manipulation.

[0003] In the field of hand exoskeletons, appropriate force feedback needs to be applied to the wearer's hand during teleoperation. Hand exoskeletons using fingertip force feedback can directly apply pressure or resistance to the fingertips, simulating the stiffness, weight, and elasticity of virtual objects with extremely high fidelity. Users can realistically feel the difference between "pinching" a hard ball or "squeezing" a sponge. In teleoperation or precision assembly tasks, fingertip force feedback constructs a closed-loop human-machine control system, ensuring the accuracy of operations and the stability and effectiveness of interactive information. Hand exoskeletons with stable and accurate fingertip force feedback capabilities can be applied to skill data acquisition, hazardous environment operations, remote surgery, and other fields within embodied intelligence.

[0004] However, force feedback hand exoskeletons that utilize fingertip force feedback face significant challenges, primarily in their sensing, perception, and control systems.

[0005] On the one hand, to achieve a truly closed-loop force control system, it is essential to acquire real-time information about the actual force exerted when the fingertip comes into contact with the external environment. Therefore, integrating a high-precision force sensor into the fingertip becomes a direct and effective solution. However, currently commercially available high-precision multi-dimensional force sensors (such as six-dimensional force / torque sensors) are typically large, complex in structure, and expensive. This not only makes them difficult to integrate within the confined space of the fingertip but also significantly increases the overall weight and energy consumption of the system, limiting their application in lightweight wearable devices. In contrast, while flexible thin-film pressure sensors offer advantages such as thinness, good flexibility, and ease of integration, they generally suffer from insufficient sensitivity, significant hysteresis, severe nonlinearity, poor repeatability, and poor long-term stability. These shortcomings make it difficult to meet the requirements of high resolution, high dynamic response, and high stability in scenarios requiring precise force control.

[0006] Furthermore, most current fingertip exoskeleton devices can only output normal pressure, i.e., force feedback in a single direction perpendicular to the fingertip surface, lacking the ability to simulate tangential force (friction) and its dynamic changes. Real-life tactile experiences are often the result of the combined effects of multi-dimensional force information, such as the continuous micro-force changes generated when a finger gently strokes an object's surface, or the slight slippage tendency of an object along the fingertip due to gravity when holding a heavy object. This complex mechanical information is crucial for creating realistic and nuanced tactile perception. If tangential force and its changing trends cannot be effectively simulated, users will find it difficult to perceive key tactile characteristics such as "smoothness" or "slippage," thus severely limiting the realism and immersion of the tactile feedback system.

[0007] On the other hand, at the structural design level, hand exoskeletons typically employ underactuated, multi-degree-of-freedom, multi-link mechanisms to accommodate different hand sizes and limited wearing space. This design allows for the adaptive matching of multiple finger joint movements with a relatively small number of actuators, offering advantages such as compact structure, light weight, and relatively controllable cost. However, because the number of actuators in an underactuated system is less than the number of degrees of freedom, some joints are in a passive state, making it impossible to apply an active torque independently to each joint. The force distribution of the system depends on the structural coupling relationship and contact conditions. While this characteristic offers advantages in grip adaptability, it is significantly limited when precise control of fingertip output force (especially pressing force of a specific direction and magnitude) is required.

[0008] Specifically, when the system cannot independently control each passive joint, the fingertip output force is affected by the coupling effect of the entire mechanism's posture and contact state, making it difficult to accurately predict and control the output force, thus affecting the stability and accuracy of the closed-loop control algorithm. While a fully actuated design, with independent actuators and transmission units for each degree of freedom, can achieve more precise force / position control, it significantly increases system weight, size, and energy consumption, while also increasing the complexity of the mechanical structure and the difficulty of control. This is a significant disadvantage for hand exoskeletons that emphasize lightweight, portability, and long-term wearing comfort. Therefore, achieving a balance between the lightweight advantages of underactuated structures and the need for precise force control has become one of the key challenges in current hand exoskeleton design.

[0009] Furthermore, tactile feedback from the fingertips is not solely determined by a single normal pressure, but rather is the result of the coordinated action of multiple mechanoreceptors. Human skin contains four main types of tactile sensory corpora: Meckel's corpora are sensitive to light touches and low-frequency vibrations of approximately 10–100 Hz, playing a crucial role in perceiving sliding and changes in initial contact; Pacinian corpora, located deep within the skin, are sensitive to high-frequency vibrations of approximately 100–300 Hz, capable of sensing subtle textures or transient impacts; Rufini corpora primarily respond to continuous skin traction and pressure changes, playing a significant role in regulating grip stability; and Merkel discs are sensitive to static pressure below 10 Hz and surface shape and texture features. The complementarity of different receptors in frequency and time response allows humans to obtain rich and nuanced tactile experiences. Therefore, relying solely on single-pressure feedback is insufficient to fully activate all sensory channels; compensation through multimodal stimulation is necessary. Considering that the human body's perception of stimulus intensity and frequency is approximately logarithmic, the amplitude and frequency of vibration signals in different frequency bands can be modulated to encode and simulate information such as texture density, tangential force changes, slight contact and slippage trends, thereby improving the realism and layering of tactile feedback under hardware limitations.

[0010] In summary, realizing an exoskeleton fingertip module that integrates fingertip sensing and vibration feedback capabilities is of high value. Summary of the Invention

[0011] The purpose of this invention is to address the shortcomings of existing technologies by proposing an exoskeleton fingertip that integrates fingertip sensing function and vibration feedback capability.

[0012] The objective of this invention is achieved through the following technical solution: an exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability, comprising:

[0013] The fingertip module base has a vibration motor at its bottom to simulate tactile sensation from the exoskeleton. A thin-film pressure sensor array is installed inside the fingertip module base. The thin-film pressure sensor array is covered with a silicone pad. A rigid pressing plate that is directly connected to the fingertip module base is set on the silicone pad. The upper layer of the rigid pressing plate has a smooth, rigid surface for contact with the fingers, and a raised structure is set at the center of the lower layer to concentrate the force and transmit it to the soft silicone pad.

[0014] Furthermore, the fingertip module base has a groove inside, and the thin-film pressure sensor array and silicone pad are installed inside the groove of the fingertip mounting base.

[0015] Furthermore, the thin-film pressure sensor array needs to be individually calibrated under the silicone layer before use. The calibration process includes: placing the thin-film sensor in a 3D printed mold with the same internal geometry as the mounting base to ensure that its force boundary conditions are consistent with the final assembly state; then applying standard loads at certain force intervals to each sensing unit and recording the corresponding voltage output data simultaneously; and based on the collected force and voltage data, performing curve fitting and parameter calibration on each sensor to establish its force-voltage mapping relationship under actual structural conditions.

[0016] Furthermore, the thin-film pressure sensor array is a 2×2 thin-film sensor array. By modeling the propagation model of force in the flexible film as a two-dimensional Gaussian function, the force value information collected by the thin-film pressure sensor array is combined with information such as the location of the thin-film pressure sensor array to simultaneously estimate the magnitude and location of the pressing force applied to the silicone pad.

[0017] Furthermore, the setting of the vibration motor to simulate tactile sensation on the exoskeleton includes:

[0018] Texture feedback is achieved by controlling the output frequency of the vibration motor to activate different types of tactile corpora. The interactive tangential force obtained when the dexterous hand interacts with the object is mapped to the vibration motor amplitude through the human sensory level. The interactive tangential force is fed back by controlling the amplitude of the vibration motor, and the texture density perception when the dexterous hand interacts with the object is controlled by controlling the output duty cycle of the vibration motor.

[0019] Furthermore, the method of activating different types of tactile corpora by controlling the output frequency of the vibration motor to achieve texture feedback includes: selecting a frequency of 10-40 Hz for rough stone texture, a frequency of 60-120 Hz for fine wood grain texture, a frequency of 150-250 Hz for writing paper texture, a frequency of 300-450 Hz for velvet texture, and a short frequency of 150-200 Hz for metal impact texture.

[0020] Furthermore, the step of mapping the interactive tangential force obtained when the dexterous hand interacts with the object to the vibration motor amplitude through the human body's perception level includes: first, processing the tangential force information into a stepped discrete tangential force; then, based on the logarithmic relationship between the human body's perceived vibration amplitude and the vibration perception level according to Weber's theorem in psychology, as well as the relationship between the tangential force, the motor's basic amplitude, and the perception level, obtaining the relationship between the human body's perceived amplitude and the motor's basic amplitude under different tangential forces.

[0021] Furthermore, the basic amplitude of the motor At different frequencies The value below is based on the minimum amplitude that a human hand can perceive when using a vibrating motor at a frequency of 0Hz. The minimum amplitude of a vibration motor that can induce tactile sensation at 250 Hz, the frequency most sensitive to the human hand. set up: .

[0022] The beneficial effects of this invention are as follows: This invention constructs a fingertip force sensing structure that is force-concentrated, stably transmitted, and calibrable through the collaborative design of an array-type thin-film sensor, an elastic silicone force-transmitting layer, and a raised-dot pressing plate. Combined with a two-dimensional Gaussian function reconstruction algorithm, it achieves continuous, smooth, and high-precision estimation of the fingertip normal contact force. Simultaneously, by integrating a vibration motor and its parametric modeling control at the bottom of the module, it achieves effective simulation of tactile information such as texture, slippage, and tangential force changes. This solution improves the stability and reliability of force sensing while ensuring a compact and lightweight structure, enhances the closed-loop control performance of the exoskeleton system in dynamic environments, and significantly improves the realism and interactive experience for users during operation. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the fingertip tactile module proposed in this invention.

[0024] Figure 2 This is an exploded view of the fingertip tactile module proposed in this invention.

[0025] Figure 3 This is a schematic diagram of the tactile sensing part of the fingertip tactile module proposed in this invention.

[0026] Figure 4 This is a real diagram of the fingertip tactile module proposed in this invention.

[0027] Figure 5 This is a diagram showing the relationship between the human hand's perception of vibrations at different frequencies, provided by this invention. Detailed Implementation

[0028] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0029] like Figure 1 As shown, embodiments of the present invention provide a fingertip tactile module that integrates tactile sensing capabilities and vibration feedback, suitable for sensing precise and stable force values ​​and generating a suitable vibration tactile simulation during the implementation of a force feedback exoskeleton.

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0031] Please refer to Figure 1 The present invention discloses an exoskeleton fingertip module that simultaneously possesses fingertip sensing and vibration feedback capabilities. For example... Figure 2 The device includes a 2*2 thin-film sensor array 1-4, a silicone pad with a hardness of 20 degrees 5, a fingertip mounting base 6, a vibration motor with a size of 21mm*30mm 7, and a 3D-printed rigid pad with bumps on the bottom 8.

[0032] like Figure 1 and Figure 4 As shown, the 2×2 thin-film sensor array is installed inside the groove of the fingertip mounting base. This structural design provides stable mechanical constraints on the sensor array, preventing displacement or loss of position of the sensors due to finger movement or external disturbances during the operation of the hand exoskeleton. This ensures the stability and reliability of the sensors throughout the entire usage process, guaranteeing good consistency and accuracy of the collected force data.

[0033] A soft silicone pad is placed above the 2×2 thin-film sensor array, and this silicone pad is also embedded in the groove of the fingertip base. The silicone material has good flexibility and cushioning properties. When external force is applied to the contact surface, the silicone pad can evenly distribute the pressure, so that the pressing force applied to a single contact surface can be effectively transmitted to multiple sensor units in the array. This allows each thin-film sensor to participate in force measurement, improving the stability and reliability of the overall measurement results.

[0034] When using an underactuated, high-degree-of-freedom hand exoskeleton to achieve fingertip force feedback, the user's hand is usually suspended in the air, and the fingertips have a distinct soft tissue structure. This leads to a phenomenon where the perceived force may be the same, but due to differences in contact pattern and pressure distribution, the actual force measured by the sensor can vary significantly. To address this issue, the system further designs a... Figure 3 The rigid gasket structure shown has a smooth, rigid upper layer and a raised structure at the center of the lower layer.

[0035] This structural design serves two purposes. First, the smooth and rigid upper surface creates a more complete and stable contact interface with the fingertip, allowing the finger to receive a relatively uniform force contact area during pressing, thus reducing uneven local force distribution caused by finger slippage during movement. Second, when the finger applies pressure, the force first acts on the upper surface of the rigid pad and is then concentrated and transmitted to the soft silicone pad through the raised structure beneath it. Subsequently, the soft silicone pad diffuses and transmits this concentrated pressure to the 2×2 thin-film sensor array below.

[0036] This multi-layered structure design, consisting of a rigid pad, a soft silicone pad, and a thin-film sensor array, not only ensures effective acquisition of convex surface pressure but also fully leverages the advantages of the thin-film sensor array in data fusion. Compared to traditional methods that directly apply pressure to a single thin-film sensor, this approach averages and corrects data through collaborative measurements from multiple sensor units. This effectively reduces measurement noise, improves the stability of force acquisition, and minimizes force unevenness caused by finger slippage or changes in contact position.

[0037] Because the stress state of the thin-film pressure sensor varies under different installation conditions, its force-voltage output curve is inconsistent when it is placed under the silicone layer versus when it is completely exposed. Therefore, it needs to be calibrated individually under the actual assembly structure conditions. Specifically, the thin-film sensor is placed in a 3D-printed mold with the same internal geometry as the mounting base 6, ensuring that its stress boundary conditions are consistent with the final assembly state. Then, standard loads are applied to each sensing unit at certain force intervals, and the corresponding voltage output data is recorded simultaneously. Based on the collected force and voltage data, curve fitting and parameter calibration are performed on each sensor to establish its force-voltage mapping relationship under actual structural conditions, thereby improving the accuracy and consistency of subsequent force calculations.

[0038] After each group of thin-film sensors has completed curve fitting and calibration, when a force is applied to the tactile module, the reading value of each thin-film sensor can be obtained. The force applied to the silicone pad and transmitted to each thin-film sensor conforms to a two-dimensional Gaussian function.

[0039]

[0040] in It is the location of the pressure point. The parameters of the Gaussian function represent the standard deviation of the force model. This refers to the actual pressure applied by the fingertip, measured by the force values ​​collected from four thin-film sensors. Then, the actual force applied to the silicone pad can be reliably estimated, which can then be used in hand exoskeletons based on fingertip force feedback. For each thin film sensor The coordinates.

[0041] The vibration feedback of the fingertip module is achieved using a vibration motor, which is fixed to the fingertip module. By controlling the vibration frequency, vibration amplitude, and control duty cycle of the vibration motor, the sensory feedback of information such as texture information, interaction force, and roughness when the end device interacts with the environment is achieved. This interaction information comes from the force or texture information when the dexterous hand or other end effector interacts with the environment. In order to realize the feedback from the hand exoskeleton through the tactile sensation of the dexterous hand, the interaction information between the dexterous hand and the object needs to be converted into interaction information that the hand exoskeleton can process. For example, the magnitude of the interaction force is converted into the vibration amplitude to a certain extent. As shown in Table 1, when humans receive low-frequency vibration signals, they can feel the sensation of an object slowly moving or pressing on their skin; when humans receive mid-frequency signals, they can feel a sensation similar to friction on a rough surface; when the vibration frequency reaches the high-frequency range of 100-400Hz, humans can feel a smooth texture on a flat surface.

[0042] Table 1. Comparison of Characteristics and Functions of Finger Tactile Receptors

[0043] Receptor name Frequency range Main functions and sensing characteristics Merkel touches the plate 0-100Hz Responsible for distinguishing the hardness, edges, and roughness of an object. Meissner bodies 10-60Hz It is highly sensitive to dynamic deformation of the skin and is primarily used for gripping control. Ruffini's body <10hz It can sense the posture of the hand and the direction of finger movement, and is sensitive to lateral shear force. Pacinian bodies 60-1000Hz Perceiving subtle textures (such as silk) or vibrations transmitted through tools.

[0044] Amplitude reflects the physical magnitude of skin deformation, while in psychology, Weber's theorem corresponds to the perceived intensity—the amplitude of vibration felt by the human body. Vibration perception level It approximates a logarithmic relationship and uses a linear relationship. Right now,

[0045]

[0046] in It is the amplitude output by the vibration motor. This refers to the sensitivity level of the fingertips to stimuli. The amplitude of the vibration motor output from the fingertip module of the hand exoskeleton. It should be in tangential interaction with the end device and the environment. Relatedly, in order to simultaneously differentiate tangential forces at different levels from those of a dexterous hand... It can be considered Levels of human perception There exists a linear relationship. Simply put, the tangential force information obtained from the end device is processed into five levels, with the goal of enabling the exoskeleton wearer to distinguish these force levels, thereby allowing them to perceive continuous tangential force from the end device. It was transformed into discrete step-like forces. Thus, the level of stimulation perception is obtained. :

[0047]

[0048] in This sets a minimum perceptible value for the output of the vibration motor; the user will only experience stimulation when the vibration motor provides a value greater than this threshold. It is the minimum perceived value Tangential force step value Convert to stimulation level A linear transformation parameter. Perception-based hierarchical feedback demonstrates superior practicality in human-machine loop systems by aligning with human psychophysical thresholds, reducing cognitive bandwidth consumption, and prioritizing the delivery of executable information over raw data. For example... Figure 5 As shown, the human hand has a minimum amplitude threshold for vibrations at different frequencies. We can set the basic output amplitude of the vibration motor at different frequencies. With frequency The relationship is:

[0049]

[0050] in It is the smallest amplitude that can cause tactile sensation when the frequency of the vibrating motor is 0Hz. This refers to the minimum amplitude of the vibration motor at 250Hz, which is the frequency at which human touch is most sensitive. Combining this with the relationship between the human body's perceived vibration amplitude and vibration perception level, we have:

[0051]

[0052] Table 2 Comparison of Tactile Sensation Types and Frequency

[0053] Target texture Recommended frequency range Amplitude (Intensity) Recommendations Rough stone 10 - 40 Hz High amplitude (strong impact) Fine wood grain 60 - 120 Hz medium amplitude Writing paper 150 - 250 Hz Low amplitude (varying with moving speed) Velvet / Silk 300 - 450 Hz Extremely low amplitude Metal impact Short bursts of 150-200 Hz Rapid decay after a momentary high amplitude

[0054] The table above allows for the simulation of hand tactile sensation using different frequencies of a vibrating motor.

[0055] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

[0056] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this application. This application is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. An exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability, characterized in that, include: The fingertip module base has a vibration motor at its bottom to simulate tactile sensation from the exoskeleton. A thin-film pressure sensor array is installed inside the fingertip module base. The thin-film pressure sensor array is covered with a silicone pad. A rigid pressing plate that is directly connected to the fingertip module base is set on the silicone pad. The upper layer of the rigid pressing plate has a smooth, rigid surface for contact with the fingers, and a raised structure is set at the center of the lower layer to concentrate the force and transmit it to the soft silicone pad. The method of setting up a vibration motor to simulate tactile sensation on the exoskeleton includes: Texture feedback is achieved by controlling the output frequency of the vibration motor to activate different types of tactile corpora. The interactive tangential force obtained when the dexterous hand interacts with the object is mapped to the amplitude of the vibration motor through the human sensory level. The interactive tangential force is fed back by controlling the amplitude of the vibration motor. The texture density perception when the dexterous hand interacts with the object is controlled by controlling the output duty cycle of the vibration motor. The method of mapping the interactive tangential force obtained when a dexterous hand interacts with an object to the vibration motor amplitude through the human body's perception level includes: first, processing the tangential force information into a stepped discrete tangential force; then, based on the logarithmic relationship between the human body's perceived vibration amplitude and the vibration perception level according to Weber's theorem in psychology, as well as the relationship between the tangential force, the motor's basic amplitude, and the perception level, obtaining the relationship between the human body's perceived amplitude and the motor's basic amplitude under different tangential forces.

2. The exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability according to claim 1, characterized in that, The fingertip module base has a groove inside, and the thin-film pressure sensor array and silicone pad are installed inside the groove of the fingertip mounting base.

3. The exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability according to claim 1, characterized in that, Before use, the thin-film pressure sensor array needs to be individually calibrated under the silicone layer. The calibration process includes: placing the thin-film sensor in a 3D printed mold with the same internal geometry as the mounting base to ensure that its force boundary conditions are consistent with the final assembly state; then applying standard loads to each sensing unit at predetermined force intervals and recording the corresponding voltage output data simultaneously; and based on the collected force and voltage data, performing curve fitting and parameter calibration on each sensor to establish its force-voltage mapping relationship under actual structural conditions.

4. The exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability according to claim 1, characterized in that, The thin-film pressure sensor array is a 2×2 thin-film sensor array. By modeling the propagation model of force in the flexible film as a two-dimensional Gaussian function, the force value information collected by the thin-film pressure sensor array is combined with the location information of the thin-film pressure sensor array for fitting, so as to simultaneously estimate the magnitude and location of the pressing force applied to the silicone pad.

5. The exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability according to claim 1, characterized in that, The method of activating different types of tactile corpora by controlling the output frequency of the vibration motor to achieve texture feedback includes: selecting a frequency of 10-40 Hz for rough stone texture, a frequency of 60-120 Hz for fine wood texture, a frequency of 150-250 Hz for writing paper, a frequency of 300-450 Hz for velvet texture, and a short frequency of 150-200 Hz for metal impact texture.

6. The exoskeleton fingertip integrating fingertip sensing function and vibration feedback capability according to claim 1, characterized in that, The motor's basic amplitude At different frequencies The value below is based on the minimum amplitude that a human hand can perceive when using a vibrating motor at a frequency of 0Hz. The minimum amplitude of a vibration motor that can induce tactile sensation at 250 Hz, the frequency most sensitive to the human hand. set up: .

Images