Compatibility evaluation of a 4-DOF ergonomic exoskeleton for upper limb rehabilitation

https://doi.org/10.1016/j.mechmachtheory.2020.104146Get rights and content

Highlights

  • A novel ergonomic exoskeleton is presented for 4-DOF upper limb rehabilitation.

  • A corresponding prototype is developed based on the proposed exoskeleton.

  • The interaction forces and displacements at the physical connection were sampled.

  • The exoskeleton’s compatibility is quantitatively evaluated under two modes.

  • An ergonomically designed exoskeleton can markedly reduce the interaction loads.

Abstract

Due to their advantages of high durability, low labor intensiveness and high repeatability, upper limb exoskeletons have become promising tools in stroke rehabilitation. The act of decreasing the undesired interactional loads caused by exoskeleton incompatibility remains an enormous challenge in the design of ergonomic exoskeletons. In this article, a novel 4-DOF upper limb exoskeleton that is kinematically compatible with the upper limb is proposed. A prototype of the proposed exoskeleton was developed. Subsequently, the interaction forces, torques and displacements at the physical human-exoskeleton connection interfaces were detected under static and dynamic modes to quantitatively evaluate the compatibility of the exoskeleton. The results indicated that the proposed exoskeleton can significantly decrease the undesired interactional load at the connective interfaces, and its ergonomic design was found to be effective; thus, this exoskeleton may be used for the rehabilitation of human upper limbs.

Introduction

OVER the past two decades, a variety of upper extremity exoskeletons have been developed for the rehabilitation of patients suffering from stroke and dyskinesia [1], [2], [3]. Compared with traditional physiotherapeutic approaches in upper limb rehabilitation, exoskeletons have a plethora of advantages, such as long duration, low labor intensiveness, and high repeatability [4,5]. For the development of exoskeletons, configuration design is particularly important. Regarding the design of exoskeleton configurations, the compatibility of exoskeleton with the user is an urgent problem requiring resolution. If an exoskeleton is incompatible with the upper limb of a user, the connective interface of the exoskeleton will generate undesired interactional (UI) loads that are exceedingly detrimental to rehabilitation therapy [6,7]. These loads can appreciably increase the magnitude of constraints between the exoskeleton and upper limb, resulting in a decreased level of comfort for the user wearing the exoskeleton and, in extreme cases, a risk of injury to the user.

Due to factors such as the misalignment of the axes (centers) of the human-exoskeleton joints, the kinematics differences between the upper limb joints and exoskeleton joints, the initial wearable offsets and the slippage generated at the physical connection interfaces during the rehabilitation tasks [8], [9], [10], [11], [12], exoskeletons may be incompatible with the upper limb. Among these factors, axis (center) misalignment is the main cause of incompatibility. To resolve axis (center) misalignment, an effective approach is to introduce additional passive joints into the exoskeleton configuration, thus decreasing the UI loads [9], [10], [11], [12], [13], [14], [15], [16]. Based on the installed positions of the introduced joints in exoskeletons, exoskeletons are divided into two categories: self-tracing exoskeletons and self-adapting exoskeletons. For self-tracing exoskeletons, passive joints are located between the actuating joints and the base to realize axis (center) tracking of the joints. A typical example of a self-tracing exoskeleton is shown in Fig. 1(a), and passive joints (universal joint U and prismatic joint P) are located between the active joint RA3 and the base. This category of exoskeletons has obvious advantages. As shown in Fig. 1(a), when the exoskeleton is connected to the upper limb, the axes (centers) of the exoskeleton joints can track the posture trajectories of the corresponding axes (centers) of the upper limb joints during the rehabilitation task, thus reducing the UI loads. However, when universal joint U and prismatic joint P are introduced, the frictional loads of passive joints as well as the partial gravity of the exoskeleton are borne by the affected limbs during the rehabilitation task, which affects the recovery of the upper extremity. In similar investigations on self-tracing exoskeletons [17], [18], [19], [20], [21], two or three passive degrees of freedom (DOFs) were introduced to transform the closed chain into an even-actuation kinematic problem to improve the hyperstaticity of human-exoskeleton closed chain. For the self-adapting exoskeletons, additional passive joints are added to the connecting sub-chains between the upper limb and exoskeleton to decrease the UI loads [22], [23], [24], [25], [26]. Fig. 1(b) shows a classic example of self-adapting exoskeleton, where prismatic joint P and universal joint U are added in series at the connection between the exoskeleton and the upper extremity. When the exoskeleton is worn on the upper limb of the user, the exoskeleton may be compatible with the upper limb, even when the exoskeleton axes (centers) and upper limb joint axes (centers) are not aligned. Moreover, since the exoskeleton is connected to the base, its gravitational force as well as the frictional force of the prismatic joint and universal joint are borne by the base and the actuators, respectively, which can efficiently weaken the UI loads between the exoskeleton and upper limb. Nevertheless, problems exist with self-adapting exoskeletons that warrant additional discussion. To weaken the UI loads, prismatic joint P and universal joint U are included in the physical connection between the exoskeleton and the upper limb, which influences the force transmissibility of the active joints (RA1, RA2 and RA3). Moreover, the passive connective sub-chain (prismatic joint and universal joint in series) is too long, which causes a large offset between the relative motions of the upper extremity and exoskeleton during rehabilitation. In summary, if exoskeletons that have the advantages of both self-tracing exoskeletons and self-adapting exoskeletons can be developed, they will be extremely important for upper limb rehabilitation therapy.

In addition to the concerns described above, the evaluation of exoskeletons compatibility is another important problem that deserves attention [27,28]. Compatibility refers to that the exoskeleton configuration, on the basis of ensuring that it is basically consistent with the dimension characteristics of the upper limb, has similar movement posture when the upper limb is kinetostatic or dynamic. And the greater UI loads will not be generated on the connective interfaces during rehabilitation task, so as to guarantee the safety and comfort of user. Moreover, the mainly reason affecting compatibility is the axes (centers) misalignment of exoskeleton joints and upper limb joints, and the design parameters have little influence on it. Thus, the influence of dimension parameters on compatibility is ignored in this study. Through literatures retrieval and analysis, some compatibility evaluation methods based on interaction forces have been proposed, such as the quantification of interaction forces and evaluation of the response of the user to functional assistance [29], the prediction of the compatibility changes caused by the misalignment of joint axes (centers) [24], the detection of interaction forces at the connection interfaces [30], the quantification of the changes in the interaction forces between upper limb and exoskeleton and the comparison of the magnitude of the interaction forces with or without passive joints [31], and the evaluation of the influence of an ergonomic design on the characteristics of interaction forces [32]. As can be seen from the above studies [24,30], the researchers assessed the exoskeleton interaction performance by evaluating the reduction level of interaction forces at the connective interfaces. In our work, in addition to sampling the interaction forces and torques on the connective interfaces, the displacement of the passive prismatic joint was measured, so as to evaluate the compatibility of exoskeleton.

In this study, a 4-DOF ergonomic exoskeleton configuration for upper extremity rehabilitation that has the advantages of both self-tracing and self-adapting exoskeletons is presented. On the basis of this design, an exoskeleton prototype was developed. Subsequently, the interaction forces, torques and displacement between the exoskeleton and the upper limb were detected, and compatibility of exoskeleton was quantitatively evaluated. The remainder of the paper is organized as follows. In Section II, a mechanism for 4-DOF ergonomic exoskeleton is proposed, and the distribution characteristics of passive joints and configuration advantages are presented. Afterwards, a prototype of the exoskeleton is described in detail, and the interaction forces, torques and displacements under static and dynamic modes are recorded. The results are then presented in Section III. In Section IV, a discussion on the compatibility of the exoskeleton is provided on the basis of the quantitative experimental data. Finally, conclusions as well as potential directions for future work are given in Section V.

Section snippets

Description of the exoskeleton configuration

The kinematic characteristics of the human upper limb are highly complex, and it is extremely difficult to model and replicate upper extremity kinematics [33], [34], [35]. According to the anthropotomy studies, the upper extremity consists of both the shoulder and elbow complex, as shown in Fig. 2.

The shoulder complex mainly consists of the scapula, humerus, clavicle and several joints, which are the sterno- clavicular (SC) joint, acromio-clavicular (AC) joint, scapular- thoracic (ST) joint,

Results

The experimental data sampled in static mode and dynamic mode were used to calculate the defined performance indexes. It provided a quantification of the general residual, undesired and uncontrolled forces acting between the exoskeleton and the volunteer. It was important to note that all experiments were conducted in the passive mode of the exoskeleton; theoretically, the upper extremity should not bear any load at the connective interfaces. In the two experimental modes, we plotted the

Discussion

As far as the upper extremity exoskeleton is concerned, it is crucial to design an exoskeleton compatible with the upper limb. Otherwise, it will generate uncontrollable hyperstatic forces at the connective interfaces, which are extremely detrimental to rehabilitation training. Therefore, the act of decreasing the hyperstatic forces is an urgent problem that requires resolution. There are many factors that contribute to hyperstatic forces, among which the misalignment of the axes (centers) is

Conclusion and future work

In this paper, a 4-DOF ergonomic exoskeleton configuration kinematically compatible with the upper limb was presented, and a prototype was designed. For the purpose of evaluating the compatibility of the exoskeleton, numerous experiments were performed with the exoskeleton apparatus. Here, the interaction forces, torques and displacements of the connective interfaces under the static mode and the dynamic mode were recorded, respectively, and the compatibility of the exoskeleton was then

Declaration of Competing Interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported in part by the Beijing Natural Science Foundation, Grants No. 3171001, in part by the National Natural Science Foundation of China, grant No. 51675008 and 61903011, in part by the National Key R&D Program of China under Grant No. 2018YFB1307004 and 2020YFC2004200, in part by Beijing Natural Science Foundation under grant No. 3204036, and in part by the Natural Science Foundation of Beijing Education Committee under Grant No. KM202010005021.

Jianfeng Li received his Ph.D. degree in mechanical engineering from Beihang University, China, in 1999. After a post doctor experience from Tsinghua University, he works as a faculty member in the College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, China, since 2001.

He was elected as a full professor in 2008. His research interests include theory of parallel mechanism, wearable exoskeleton, external fixator and rehabilitation robotics.

References (44)

  • B.T. Volpe et al.

    Robotic devices as therapeutic and diagnostic tools for stroke recovery

    Arch. Neurol.

    (2009)
  • M. Gilliaux et al.

    A robotic device as a sensitive quantitative tool to assess upper limb impairments in stroke patients: a preliminary prospective cohort study

    J. Rehabil. Med.

    (2012)
  • H.S. Park et al.

    IntelliArm: an exoskeleton for diagnosis and treatment of patients with neurological impairments

  • N. Jarrassé et al.

    Robotic exoskeletons: a perspective for the rehabilitation of arm coordination in stroke patients

    J. Fro. Hum. Neu.

    (2014)
  • A. Schiele et al.

    Kinematic design to improve ergonomics in human machine interaction

    IEEE Trans. Neural Syst. Rehabil. Eng.

    (2006)
  • R.A.R.C. Gopura et al.

    SUEFUL-7: a 7DOF upper-limb exoskeleton robot with muscle-model-oriented EMG-based control

  • T. Nef et al.

    ARMin III-arm therapy exoskeleton with an ergonomic shoulder actuation

    Appl. Bionics Biomech.

    (2009)
  • S.J. Housman et al.

    Arm-training with T-WREX after chronic stroke: preliminary results of a randomized controlled trial

    IEEE Int. Conf. Rehabil. Rob. ICORR.

    (2007)
  • J.C. Perry et al.

    Upper-limb powered exoskeleton design

    IEEE ASME Trans. Mechatron.

    (2007)
  • H. Zhang et al.

    RUPERT closed loop control design

  • A. Frisoli et al.

    A New Force-Feedback Arm Exoskeleton for Haptic Interaction in Virtual Environments

  • C.J. Yang et al.

    A review of exoskeleton-type systems and their key technologies

    Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci.

    (2008)
  • Cited by (27)

    View all citing articles on Scopus

    Jianfeng Li received his Ph.D. degree in mechanical engineering from Beihang University, China, in 1999. After a post doctor experience from Tsinghua University, he works as a faculty member in the College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, China, since 2001.

    He was elected as a full professor in 2008. His research interests include theory of parallel mechanism, wearable exoskeleton, external fixator and rehabilitation robotics.

    Qiang Cao received the B.S. degree in mechanical design and automation and M.S. degree in mechanical engineering from LanZhou University of Technology, LanZhou, GanSu province, from 2010 to 2017. He is currently pursuing the Ph.D. degree in mechanical engineering at Beijing University of Technology.

    His research interests include the kinematic and kinetic analysis of the upper limb rehabilitation exoskeletons and the analysis and design of humanoid robots.

    Mingjie Dong received his B.E. from the School of Mechanical Engineering and Automation, Beihang University, China, in June 2012, and received his Ph.D. from the Robotics Institute of Beihang University in March 2018.

    Now he works as a faculty member in the College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, China. His current researches include compliance control of rehabilitation robots with active/passive rehabilitation training strategies.

    Chunzhao Zhang received the B.S. degree in mechanical engineering from Henan University of Technology, Jiaozuo, China, in 2012 and the M.E. degree in mechanical engineering from Beijing University of Technology, Beijing, China, in 2016. He is currently pursuing the Ph.D. degree in mechanical engineering from Beijing University of Technology.

    His current research includes the human skeleton kinematics and humanoid robot.

    View full text