For VR hand exoskeleton systems, a wearable hand exoskeleton system with force controllable actuator modules was developed to ensure free finger movements and force mode control. To overcome the weakness of the previous system, a wearable hand exoskeleton system with finger motion measurement and force feedback was developed and evaluated in terms of user experience (UX).
Motivation
The Hand
Because complex finger movements are possible according to the anatomy of the hand, people can hold objects of various shapes stably.
Applications of the hand exoskeleton systems
- Virtual Reality
- Rehabilitation
In the first group, the central nervous system (CNS) is affected by diseases (usually strokes) [27]. In the second group of patients, muscles or wrists are damaged in an accident or surgery.
State of the Art
Hand Force (Kinesthetic) Feedback Systems
In stroke patients, motor signals are disrupted due to abnormalities in the brain that cause muscle spasticity or fatigue, although the muscles are normal. As described above in the context of VR, establishing clear criteria for evaluating system performance is challenging.
![Figure 1-3: The hand force feedback systems for VR [1–8]](https://thumb-ap.123doks.com/thumbv2/123dokinfo/10484723.0/20.892.133.779.157.590/figure-1-3-hand-force-feedback-systems-vr.webp)
Hand Rehabilitation Systems
Various types of systems are used to support and rehabilitate ADL hand movements. Additionally, because the Exo-Glove's actuator module is far from the hand, the patient must carry a heavy box.
Thesis Overview
Introduction
To minimize the interference problem, the cable structure was put on the back of the fingers, so the system size is quite large. In addition, the motors located on the outside of the hand impaired the mobility of the system.
Structure Design
- Design of the Structure
- Kinematic Analysis of the Structure
Consequently, all splice positions of linkage 1 were calculated from the four-strand linkage analysis. The working space of the link structure can be increased by increasing the length of the structure and the size of the system as shown in Fig.
![Figure 2-2: Anatomy of the finger (edited from [20])](https://thumb-ap.123doks.com/thumbv2/123dokinfo/10484723.0/28.892.182.725.158.352/figure-2-2-anatomy-finger-edited-20.webp)
Actuator Module Design
- Series Elastic Actuator (SEA) Mechanism
- Actuator Module Control
The frequency response of the motor with the friction compensation algorithm is shown in Fig. The control input weighting factor R was obtained experimentally to successfully control the motor position while preventing saturation of the control input.
Implementation of the Hand Exoskeleton
- Analysis of Force Distribution
- Force Transmission Experiment
The only colored area shows the fingertip force in the workspace of the proposed structure. Also, without moving the MCP link, the force also decreases as the angle of the PIP link changes. The performance to transmit force feedback to the fingertip was also tested using the prototype exoskeleton system.
As shown by the results of the kinematic analysis and experiment, the fingertip force decreases as the joint angles of the finger increase.
Summary
The applied force on the fingertip, Ftip, is highly dependent on the joint angles of the finger; as the joint angles increase, the fingertip force also decreases. As the finger is bent, the angle between the actuator module and the normal direction of the fingertip increases and the normal force of the fingertip decreases. Similar to the results of the kinematic analysis in the previous section, the delivered force is reduced when the finger is bent.
However, the normal fingertip force was less than that generated due to the drastic change in the normal direction of the fingertip when the finger is bent.
Design of the Exoskeleton System for VR
Because the finger structure is designed as a structure in which two curved links are connected by one rotating joint and, unlike the previous system, it is supported only by the fingertips, a full joint ROM can be guaranteed while avoiding the collision between the links and the users is minimized with different movements. Manual users can use the system without replacing the parts. Moreover, although the portability and portability of the system was only described by qualitative description in a previous study, the system performance in this study was also evaluated through UX evaluation. The system has simple structures worn on the fingertips and palm, ensuring full ROM for natural finger movements.
The system performance for finger movement measurement and interaction with VR was verified through the experiment.
Finger Motion Measurement
- Workspace of the Fingertip
- Calculation of Fingertip Position
- Calibration Process
- Joint Angle Measurement of the Index Finger
- Joint Angle Measurement of the Thumb
The position of the tip of the index finger relative to {0}. the frame is calculated using forward kinematics as follows:. In the case of the thumb, a method to convert the position of the fingertip relative to the {0}thframe to the position relative to the CMC joint was investigated in the next section. Thus, the measured thumb tip orientation is used to convert the position of the thumb tip relative to the motor to the position relative to the CMC joint.
Thus, as with the algorithm for the index finger, the joint angles of the thumb are also calculated based on the thumb phalanx lengths and the fingertip position with respect to the CMC joint.
Force Feedback
Although there are 2 unknown angles in the case of the index finger, there are 3 unknown angles in the case of the thumb and a total of 27 position candidates are considered. The fingertip position relative to the motor is converted to that relative to the CMC joint by using matrix[R] and vector[T], and the fingertip position is converted to 5 DOF of finger joint angles through the estimation algorithm using 27 pose candidates.
Experimental Verification
- Measurement of the Fingertip Position
- Pinch Motion
- Estimation of Finger Joint Angles
- Performance Evaluation as a Haptic Device
2-33 shows the experimental setup and the results of the pinching movement using the thumb and index finger. The performance of the finger joint angle estimation was verified by comparing with a motion capture system. Furthermore, the measurement performance of the proposed system for the thumb was also a) MCP joint angle in F/E motion.
Thus, it was verified that the proposed system accurately captures the angles of the thumb and index finger joints.
Summary
Because other movements were linked to the CMC F/E movement, the 3 independent movements (i.e., CMC F/E, CMC A/A, IP F/E movements) of the 5 DOF movements were verified in this experiment. The finger joint angles measured by the proposed system are very similar to those of the motion capture system.
User Experience (UX) Evaluation
- Introduction
- Virtual Reality Program
- Evaluation Framework for Wearable Hand Systems
- Experimental Setup
- Result & Discussion
- User Experience (UX) Evaluation
- System Satisfaction
- Summary
For the UX evaluation, the subjective evaluation of the subject using the system was obtained. However, the evaluation framework should be revised according to the characteristics of the system even though this system is also a portable device. Similar to UX evaluation factor, the satisfaction score for system design space was assessed to capture the influence of the system design on the evaluation factor.
The evaluation criteria of the haptic interface were determined according to the characteristics of the proposed system.
![Figure 2-36: Evaluation of a haptic interface [21]](https://thumb-ap.123doks.com/thumbv2/123dokinfo/10484723.0/70.892.241.675.153.363/figure-2-36-evaluation-of-haptic-interface-21.webp)
Summary
In addition to these advantages, users also reported some disadvantages, namely that the system was heavy and uncomfortable, and that the feel of the force feedback was different from the feel of real objects. Thus, the patient feels uncomfortable due to the fixed parts in these systems which are made as table-mounted or splint-mounted systems to distribute the weight of the system. Although this method can reduce the weight of the system in the hand, the external case for the actuator and sensor modules degrades the portability of the system.
In addition, the high portability system reduces short preparation time and reduces fatigue during rehabilitation exercises.
The Exoskeleton Design
Optimized Design Structure
- Hand flexion/extension experiment
- Optimization Algorithm
- Spring mechanism
- Force distribution analysis
-8, the angles of the finger joints from the exoskeleton structure obtained through the optimization algorithm were compared with those of the desired trajectory. In addition, the potential field is calculated using the spring force around the desired finger joint trajectory. The transmitted spring forces on the finger joints are proportional to the deflected movements of the finger.
Figures 3-11 (b) and (c) show the sensitivity of the moment transmitted to the MCP and PIP joints, respectively.
Performance Evaluation
Implementation of the exoskeleton system for the hand
The figure shows the moment sensitivity of the joint, as the cross-sectional area of the selected connection varies from 9 mm2 to 36 mm 2 , while the area of the other connections is constant at 9 mm 2 . Notably, the transmitted moment at the PIP joint decreases dramatically as the cross-sectional area of link3 increases and link4 decreases. As connection3 becomes thicker, the inertia of the structure around the MCP joint increases, increasing the moment transmitted to the MCP joint.
Similarly, the moment transferred to the PIP joint is highly dependent on the thickness of link 4.
Finger Motion Experiment
In addition, a potentiometer embedded in the actuator measures the length of the motor stroke to estimate the posture of four fingers. The magnitude of the spring force is calculated from the deflected length, measured by a potentiometer mounted on the spring box. A small-sized linear motor was attached to the dorsum of the hand to guide the fingers (L12-P, Actuonix, Canada [63]), and the exoskeleton structure was fabricated using 3D printing technology.
The figure shows that the exoskeleton system for the hand can effectively guide the patient's fingers along the desired exoskeleton trajectory.
Force Distribution Experiment
Summary
The fingertip workspace of the joint structure was evaluated kinematically and compared with that of the functional ROM. Since the structural workspace allows a ROM that is 90% of the natural value, the user can use different postures when interacting with objects. Using the proposed finger motion measurement algorithm, the system accurately measured 5 DOF motions of the thumb and 4 DOF motions of the index and middle fingers after a single calibration, unlike previous systems.
The average satisfaction score was 5.87/7; most subjects had a positive attitude towards the system and many felt that the learning ability was excellent.
A Wearable Spring-guided Hand Exoskeleton for Continuous Passive Motion
To explore user experience (UX), the evaluation framework reflected system characteristics; I used a questionnaire to capture the subjects' opinions regarding usability and utilitarianism; Responses were scored using a 7-point Likert scale. Moreover, since the system allows users to engage with VR through intuitive control and feedback methods, satisfaction in terms of functional design was high. Negatives that were cited included the cumbersome and uncomfortable nature of the system and the feel of the force response not reflecting the 'feel' of real objects.
Open Issues
Zero Impedance Performance
Since zero impedance is not perfectly implemented using the control algorithm, we added an additional mechanism. In a simple braking system, zero impedance is easily realized by disconnecting the clutch, but forces cannot then be delivered. On the other hand, a current-controlled actuator accurately generates forces, but the zero-impedance performance is worse than that of a brake.
Therefore, by combining a brake and a current-controlled actuator, the actuator module can achieve high zero impedance and generate different forces, as shown in Fig.
Direction of the Force Feedback
Bae, "A wearable hand system for virtual reality," i IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2017. Bergamasco, "Mechanical design of a novel hand exoskeleton for accurate force displaying," i IEEE International Conference on Robotics and Automation (ICRA), 2009, s. Williamson, "Series elastic actuators," i IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 1995, pp.
Bae, “Design of a Wearable Hand Exoskeleton for Exercising Flexion/Extension of the Fingers,” in International Conference on Rehabilitation Robotics (ICORR), 2017.
Parameters of kinematic analysis
Parameters of the spring design
Orientation angle of the trapezium
Familiarity with VR devices
Results of user experience (UX) evaluation
Satisfaction scores of system design
The obtained ROM from the experiment
User information
The optimized design vector
Force distribution experiment
![Figure 1-4: The rehabilitation systems for the hand [9–19]](https://thumb-ap.123doks.com/thumbv2/123dokinfo/10484723.0/21.892.132.780.158.573/figure-1-4-rehabilitation-systems-hand-9-19.webp)
