Related Experiment Video
Updated: Aug 27, 2026

A Structured Rehabilitation Protocol for Improved Multifunctional Prosthetic Control: A Case Study
Published on: November 6, 2015
The variable stiffness orthosis: customizable mechanics for assistance and rehabilitation
Nikko Van Crey1,2, David Lam3,4, Emily Bywater3,4
1Robotics Department, University of Michigan, Hayward St, Ann Arbor, 48109, MI, USA. nikkovc@umich.edu.
Background:
Challenges with community mobility are among the most prevalent disabilities worldwide, yet the passive orthoses available for daily use have remained largely unchanged for centuries. In contrast, the powered orthoses developed by researchers have advanced rapidly, but without a matching rate of user impact. To date, there are no powered ankle-foot orthoses available for daily use, stemming from challenges with reliability, safety, weight, noise, and cost. This work aims to explore the untapped potential of passive mechanisms, as a pathway that could more rapidly translate improved functionality to the people who rely on orthoses. To this end, we developed an orthosis with passive and quasi-passive mechanisms that render capabilities typically attributed to powered devices, blurring the line between powered and passive functionality.
Methods:
In this work, we present the Variable Stiffness Orthosis, an ankle-foot orthosis that strikes a balance between powered and passive orthoses in terms of functionality and daily-use practicality. The Variable Stiffness Orthosis has torque-angle relationships that can be customized with a cam-based transmission, interchanged between gait phases with a cam-switching mechanism, and softened or stiffened between activities with a motorized spring support. These mechanisms enable precise control over the torque-angle relationship, including continuously variable stiffness, decoupled energy storage and return, step-to-step adjustment of stiffness magnitude, exchanging energy between gait phases, changing equilibrium angle between gait phases, negative stiffness, and extreme stiffness. These capabilities were validated on a rotary dynamometer and pilot tested on participants with and without sciatic nerve injury.
Results:
Dynamometer testing verified the capabilities of the VSO. In pilot testing, the participants had activity-dependent stiffness preferences spanning a large range. Using the VSO, the participant with sciatic nerve injury had reduced foot drop, increased total ankle moments, reduced biological ankle moments, reduced toe striking, and reduced steppage on their AFO side compared to walking with and without their daily-use device.
Conclusions:
This work demonstrated that the passive and quasi-passive mechanisms within the VSO can replicate many powered functions, extending the capabilities of unpowered devices. The VSO also showed promise as a daily-use device, a clinical tool for orthotic prescription, and a research tool for investigating unexplored passive mechanics.

