Soft Hip Exosuit Optimization for Improving Mobility in Healthy Young Adults and Individuals with Parkinson’s Disease

Abstract

For decades, academia and industry have explored opportunities for rigid exoskeletons to improve mobility in a variety of populations. However, in the last 5-10 years the community has shifted focus and begun designing lightweight, and nonrestrictive exoskeletons or exosuits which target single-joint motions with a number of successful demonstrations of enhanced mobility. This dissertation details the design, control/algorithm development, simulation, and biomechanical evaluation of soft exosuits for both healthy young adults and individuals with Parkinson’s disease. For healthy young adults, this work explores a way to reduce the energetic cost of both walking and running with a single robotic device. Initially, the metabolic impact of assistance profiles during running was experimentally explored. A study with a tethered hip extension exosuit demonstrated that the assistance profile inspired by a musculoskeletal simulation can reduce the metabolic cost of running by 5.4% relative to walking without wearing a suit, the first reduction in running at the time of its publication. Building on this, a subsequent study developed an online walking and running detection algorithm to apply an assistance profile based on a wearer’s gait mode (i.e., walking vs. running). This was then applied to a portable hip extension exosuit and it was demonstrated for the first time that a single device could reduce the energetic cost of walking and running (9.3% and 4.0%, respectively) compared with locomotion without exosuit. Expanding the idea of a unique assistance profile per each gait mode, this dissertation explores the impact of individualized assistance profiles for each wearer to tackle inter-subject variability in response to wearable robotic assistance. To do so, this work leverages a human-in-the-loop (HIL) optimization technique, where metabolic cost is estimated in real-time and controller parameters updated until the metabolic cost is minimized. A first study on 2 subjects used a Bayesian optimization algorithm and a tethered multi-articular soft exosuit (active ankle plantarflexion and hip extension; passive hip flexion) and found metabolic reductions of up to 39% when comparing walking with optimal assistance to walking without wearing the suit. A second study on 8 subjects used Covariance Matrix Adaptation Evolution Strategy (CMA-ES) algorithm and a tethered hip flexion soft exosuit and demonstrated that the individualized assistance tailored to each participant can reduce the energetic cost of walking more than a fixed, generic assistance profile. Leveraging the knowledge gleaned from the experimental work of healthy exosuits, the final part of this dissertation explored the possibility of using a portable version of the hip flexion soft exosuit to ameliorate the gait deficits of individuals with Parkinson’s disease. Through a series of single-subject experiments, a proof-of-concept study demonstrated for the first time that wearable robotic assistance can prevent the onset of freezing of gait (FoG). Finally, another study on 2 subjects demonstrated that hip flexion assistance could reduce the number of short, shuffling steps but did not have an impact on walking speed. The foundational work presented herein demonstrates the potential of exosuit technology to reduce the energetic cost of walking and running for healthy individuals and improve one of the most challenging and distressing motor features in PD.