0
Research Papers

A Simulation Framework for Virtual Prototyping of Robotic Exoskeletons

[+] Author and Article Information
Priyanshu Agarwal

Mechanical Engineering Department,
The University of Texas at Austin,
Austin, TX 78712
e-mail: mail2priyanshu@utexas.edu

Richard R. Neptune

John T. MacGuire Professor
Mem. ASME
Mechanical Engineering Department,
The University of Texas at Austin,
Austin, TX 78712
e-mail: rneptune@mail.utexas.edu

Ashish D. Deshpande

Assistant Professor
Mem. ASME
Mechanical Engineering Department,
The University of Texas at Austin,
Austin, TX 78712
e-mail: ashish@austin.utexas.edu

1Corresponding author.

Manuscript received June 29, 2015; final manuscript received March 21, 2016; published online April 27, 2016. Assoc. Editor: Paul Rullkoetter.

J Biomech Eng 138(6), 061004 (Apr 27, 2016) (15 pages) Paper No: BIO-15-1318; doi: 10.1115/1.4033177 History: Received June 29, 2015; Revised March 21, 2016

A number of robotic exoskeletons are being developed to provide rehabilitation interventions for those with movement disabilities. We present a systematic framework that allows for virtual prototyping (i.e., design, control, and experimentation (i.e. design, control, and experimentation) of robotic exoskeletons. The framework merges computational musculoskeletal analyses with simulation-based design techniques which allows for exoskeleton design and control algorithm optimization. We introduce biomechanical, morphological, and controller measures to optimize the exoskeleton performance. A major advantage of the framework is that it provides a platform for carrying out hypothesis-driven virtual experiments to quantify device performance and rehabilitation progress. To illustrate the efficacy of the framework, we present a case study wherein the design and analysis of an index finger exoskeleton is carried out using the proposed framework.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 3

Musculoskeletal model of the human upper limb: (a) Stanford VA upper limb model and (b) isolated index finger model

Grahic Jump Location
Fig. 4

The fitting results for the index finger flexors at the three joints of the index finger. FDPI muscle at (a) MCP, (b) PIP, (c) DIP, and (d) FDSI muscle at MCP. LB and UB represent the upper and lower bound, respectively, on the fitted curve.

Grahic Jump Location
Fig. 5

A matlab–OpenSim interface for moment arm optimization. An OpenSim model was generated on-the-fly using matlab with the initial parameters. The model was then analyzed for the moment arms using the OpenSim API commands from within matlab. The resulting moment arms were used to evaluate the objective function, and the model was optimized iteratively until the termination criteria were met.

Grahic Jump Location
Fig. 2

Finger–exoskeleton coupled system: (a) a 3D model of the developed coupled index finger–exoskeleton system, (b) preliminary 3D printed prototype of the device, and (c) virtual prototype (musculoskeletal model) used for the simulations. The model has 6DOF consisting of index finger MCP flexion, PIP flexion, DIP flexion, and exoskeleton proximal, middle, and distal link rotation. The three exotendons, the distal flexion tendon (TFD), proximal flexion tendon (TFP), and extension tendon (TE) are also modeled as muscles. The four index finger muscles in the model are FDPI, FDSI, extensor digitorum communis (EDCI), and extensor indicis proprius (EIP).

Grahic Jump Location
Fig. 1

Virtual prototyping framework illustrating the virtual design and control of an index finger exoskeleton using biomechanical, morphological, and controller performance measures

Grahic Jump Location
Fig. 6

Optimized model moment arm comparison with the experimentally measured data for FDP muscle at (a) MCP joint, (b) PIP joint, (c) DIP joint, and (d) FDS muscle at MCP joint

Grahic Jump Location
Fig. 7

Comparison of model before (a) and after (b) muscle moment arm optimization. Alterations in both the origin-, via-, insertion-points and wrapping object dimensions resulted in optimized muscle moment arms.

Grahic Jump Location
Fig. 16

Muscle excitation variation during the finger flexion–extension task for the four finger muscles in a subject with (a) healthy control and (b) hypertonicity

Grahic Jump Location
Fig. 18

Muscle forces for the four finger muscles in a subject with (a) healthy FDSI muscle and (b) FDSI muscle contracture

Grahic Jump Location
Fig. 17

Finger joint angle tracking results for the finger flexion–extension task in a subject with (a) healthy control and (b) increased flexor excitation

Grahic Jump Location
Fig. 8

Exoskeleton joint angle metric mapping for various stiffness values as obtained through the parametric study using CMC analysis for a healthy subject model. The regions with no color indicate that the stiffness combination was found to be infeasible when analyzed using CMC analysis. The unit of the metric is degrees, and all stiffness values are expressed in Newton per meter. The stiffness values k3 and k4 are varied in steps of 50 N/m to limit the computational cost. (a) k1 = 250, k2 = 250; (b) k1 = 500, k2 = 250; (c) k1 = 750, k2 = 250; (d) k1 = 250, k2 = 500; (e) k1 = 500, k2 = 500; (f) k1 = 750, k2 = 500; (g) k1 = 250, k2 = 750; (h) k1 = 500, k2 = 750; and (i) k1 = 750, k2 = 750.

Grahic Jump Location
Fig. 9

Finger joint angle error metric mapping for various stiffness values as obtained through the parametric study using CMC analysis for a healthy subject model. The regions with no color indicate that the stiffness combination was found to be infeasible when analyzed using CMC analysis. The unit of the metric is degrees, and all the stiffness values are expressed in Newton per meter. (a) k1 = 250, k2 = 250; (b) k1 = 500, k2 = 250; (c) k1 = 750, k2 = 250; (d) k1 = 250, k2 = 500; (e) k1 = 500, k2 = 500; (f) k1 = 750, k2 = 500; (g) k1 = 250, k2 = 750; (h) k1 = 500, k2 = 750; and (i) k1 = 750, k2 = 750.

Grahic Jump Location
Fig. 10

Joint angle tracking performance for various joints in the system with different exoskeleton stiffnesses as obtained from the CMC analysis for a healthy subject model. (a) Low stiffness (k1 = 250, k2 = 275, k3 = 500, and k4 = 600) N/m and (b) high stiffness (k1 = 800, k2 = 900, k3 = 1000, and k4 = 1100) N/m. The exoskeleton links show large fluctuations at lower exoskeleton stiffness. Also, there is an appropriate relationship of stiffness among different links of the exoskeleton for kinematic and dynamic compatibility of the exoskeleton with the finger. Increasing exoskeleton stiffness resulted in improved tracking performance. “Ref” refers to the reference trajectory for tracking for the respective joint angle.

Grahic Jump Location
Fig. 11

Finger joint reaction force metric mapping for various stiffness values as obtained through the parametric study using CMC analysis for a healthy subject model. The regions with no color indicate that the stiffness combination was found to be infeasible when analyzed using CMC analysis. The unit of the metric is Newton, and all the stiffness values are expressed in Newton per meter. (a) k1 = 250, k2 = 250; (b) k1 = 500, k2 = 250; (c) k1 = 750, k2 = 250; (d) k1 = 250, k2 = 500; (e) k1 = 500, k2 = 500; (f) k1 = 750, k2 = 500; (g) k1 = 250, k2 = 750; (h) k1 = 500, k2 = 750; and (i) k1 = 750, k2 = 750.

Grahic Jump Location
Fig. 12

Joint reactions induced at the various index finger joints with different exoskeleton stiffnesses as obtained from the CMC analysis for a healthy subject model. (a) Low stiffness (k1 = 250, k2 = 275, k3 = 500, and k4 = 600) N/m and (b) high stiffness (k1 = 800, k2 = 900, k3 = 1000, and k4 = 1100) N/m.

Grahic Jump Location
Fig. 13

Actuator (muscle/exotendon) forces needed for tracking with different exoskeleton stiffnesses as obtained from the CMC analysis for a healthy subject model. (a) Low stiffness (k1 = 250, k2 = 275, k3 = 500, and k4 = 600) N/m and (b) high stiffness (k1 = 800, k2 = 900, k3 = 1000, and k4 = 1100) N/m.

Grahic Jump Location
Fig. 14

Performance metric mappings for various stiffness values as obtained through the parametric study using CMC analysis for a healthy subject model with k1 = 275 N/m and k2 = 325 N/m: (a) exoskeleton joint angle metric, (b) finger joint angle metric, and (c) finger joint reaction force metric. The regions with no color indicate that the stiffness combination was found to be infeasible when analyzed using CMC analysis.

Grahic Jump Location
Fig. 15

Muscle force variation during the finger flexion–extension task in a subject with (a) healthy muscles and (b) weak FDSI muscle due to 50% reduced isometric strength

Grahic Jump Location
Fig. 19

Active muscle fiber forces in two simulations representing the improving rehabilitation scenario. (a) 15% recovery and (b) 85% recovery from complete motor disability. Different rehabilitation scenarios are simulated by changing the upper bound on the allowable excitation of the index finger muscles. With increased recovery, more forces are applied by the finger muscles (FDSI muscle carries relatively higher load than the TFD exotendon under 85% recovery)

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In