Abstract
Lower-limb exoskeletons show great promise as devices for mobility assistance, rehabilitation, or augmentation. Indeed, much research has been on focused on their mechanical design and control methodology to optimize performance goals, such as improved walking speed, increased range of motion, or reduced metabolic cost. However, to truly realize these goals, it is crucial to consider not only the design and control of the device, but also the human user. Understanding how humans adapt to lower-limb exoskeletons and learn to use such devices is critical to creating solutions that can ultimately restore mobility and augment performance. This area is especially important when considering the patient populations that would most benefit from these devices, who may require different training paradigms to adapt to the devices.
Therefore, in this workshop we explore the human motor learning side of exoskeleton design and evaluation. Six experts in the field(s) will discuss how they incorporate motor learning principles into their research, and how this is or can be applied to enhance adaptation to devices and performance. These talks will be followed by a panel discussion. To culminate the workshop, attendees will be invited to rotate through information stations, in which speakers and their teams will provide demonstrations of their motor learning experimental techniques and/or analysis pipelines. With this workshop, we hope to bring together experts in human motor learning, adaptation to devices, and their intersection to work towards assistive technology that can maximally benefit the user.
Objectives
- Explores how humans learn to use lower-limb exoskeletons
- Discusses motor learning principles that are (or can be) applied to exoskeleton learning
- Reevaluates current research training paradigms for lower-limb assistive devices
- Examines the effects of different motor learning principles on training and performance
- Presents tangible ways to apply motor learning concepts in attendee research
- Encourages interaction between speakers and attendees
Program
9:00-9:05 Introduction
9:05-10:05 First Round of Invited Talks
- 9:05-9:25 Dr. Helen J. Huang
- 9:25-9:45 Dr. Maggie Wu
- 9:45-10:05 Dr. Kristen Jakubowski
- 10:05-10:25 Dr. Maura Eveld
10:25-11:00 Coffee Break
11:00-12:00 Second Round of Invited Talks + Panel Discussion
- 11:00-11:20 Dr. Laura Marchal-Crespo
- 11:20-11:40 Dr. Gelsy Torres-Oviedo
- 11:40-12:00 Panel Discussion
12:00-12:15 Break
12:15-13:00 Demonstrations
Location
Room HS10
Neue Universtät
Universitatsplätz 1
69117 Heidelberg
Speaker information
Helen J Huang, University of Central Florida
Title: Shifting locomotor strategies and brain dynamics using frequent discrete perturbations
Applying frequent discrete perturbations during a variety of locomotor tasks such as seated recumbent stepping, fixed speed treadmill walking, and self-paced treadmill walking can elicit locomotor adaptation and shift balance and gait strategies in young and older adults. Using mobile brain/body imaging (MoBI) with high-density electroencephalography (EEG), this motor adaptation and these shifts in gait strategies have been observed as changes in walking speed, stepping patterns, and electrocortical dynamics. As technologies for mobile brain/body imaging advance, new opportunities are emerging to uncover underlying neural principles of locomotor adaptation not fully observable with just biomechanical techniques.
About Dr. Huang: Helen J. Huang received her B.S. in materials science and engineering at the Massachusetts Institute of Technology, and her M.S. and Ph.D. in biomedical engineering at the University of Michigan, Ann Arbor. She worked at Michelin North America as a materials engineer prior to her graduate studies. She was a postdoctoral fellow on the University of Colorado NIH T32 Aging Grant, and an assistant research scientist in the Human Neuromechanics Laboratory directed by Daniel Ferris at the University of Michigan, prior to joining UCF in December 2015.
Huang directs the UCF Biomechanics, Rehabilitation, and Interdisciplinary Neuroscience Laboratory. Members of the BRaIN Lab team include students from biomedical engineering, mechanical engineering, electrical engineering, and biomedical sciences. The BRaIN team studies motor adaptation and neuromechanics of gait and locomotor tasks. Their research currently focuses on investigating brain dynamics underlying motor adaptation, gait, balance and interlimb coordination in young and older adults. The BRaIN team also works on developing robotic exercise devices for gait rehabilitation and fall interventions, and on developing new methods for recording and analyzing electroencephalography (EEG) and electromyography (EMG).
Gelsy Torres-Oviedo, University of Pittsburg
Title: Generalization of Motor Learning and Its Impact on Walking
Quality of life as people age is heavily dependent on maintaining mobility. This research evaluates how walking patterns learned in one situation can generalize to another situation and how this ability changes with age. This is a significant issue for people with brain or body injuries that affect walking because rehabilitation techniques typically include training on treadmills or exoskeletons. The investigator will assess how walking patterns learned on a treadmill generalize to walking on the ground and whether that ability is modifiable, changes with age, and interacts with attentional demands. One broader impact of the work lies in its potential to improve rehabilitation tools, enhance mobility and reduce the risk of falls in the elderly.
This research identifies factors regulating the generalization of walking patterns in older adults, based on the hypothesis that aging increases generalization of motor learning due to reduced sensitivity to contextual cues and greater reliance on attentional resources. Researchers will compare electromyographic (EMG) changes and generalization of newly learned walking patterns during unassisted walking overground between young and older adults. The study employs innovative methods, such as motorized shoes and split-belt treadmills, to manipulate sensory and attentional demands. The findings are expected to reveal distinct age-specific contributions of subcortical vs. conscious strategies for generalization, advancing theoretical models of motor learning generalization and informing the design of age-specific rehabilitation interventions.
About Dr. Torres-Oviedo: Dr. Gelsy Torres-Oviedo graduated in physics from The University of Texas at Austin in 2001, earned her Ph.D. in Biomedical Engineering at Emory University and Georgia Tech in 2007, and completed a post-doc in neuroscience at Johns Hopkins University in 2011. Currently a faculty member in the bioengineering department at The University of Pittsburgh, her work has significantly advanced our understanding of human locomotor adaptation.
Dr. Torres-Oviedo's research focuses on how healthy aging and brain lesions from stroke impact locomotor learning, making significant contributions to neurorehabilitation. Her notable achievements include elucidating the mechanisms by which humans adapt their walking patterns and the generalization of gait patterns across distinct contexts.
Beyond her research, Dr. Torres-Oviedo collaborative spirit and mentoring passion led her to direct two training programs between Carnegie Mellon University and The University of Pittsburgh: The Program For Neural Computation and the MS2PhD BRIDGE. Her work continues to shape motor learning and control, driving innovative rehabilitation strategies and improving mobility for aging populations and individuals with brain lesions.
Kristen Jakubowski, Emory University/Georgia Institute of Technology
Title: Tool for quantifying the sensorimotor and cortical responses to using and learning an exoskeleton
Ideally, we want devices that are easy to use and operate, ensuring that individuals can effectively use the device while maintaining sufficient cortical resources to move safely during daily life. Yet few studies concurrently examine biomechanical performance as well as the cortical resources required to learn and operate devices. Here, I discuss different tools that we can use to assess the cortical resources required to operate devices as well as the motor adaptation to devices. Integrating these measures with common biomechanical measures can aid in the development of exoskeleton controllers that symbiotically integrate with the wearer’s neuromuscular system.
About Dr. Jakubowski: Kristen Jakubowski is an ASEE eFellows Postdoctoral Fellow in the Department of Biomedical Engineering at Emory University and the Georgia Institute of Technology. Kristen works with Dr. Lena Ting in the Neuromechanics Laboratory and Dr. Gregory Sawicki in the Physiology of Wearable Robotics (PoWeR) Laboratory. Kristen’s research seeks to understand principles that govern human neuromuscular control, and how these principles are altered by pathologies. Through this work, she aims to restore, maintain, and augment neuromuscular health and mobility through target rehabilitation and personalized wearable robotic devices. Prior to Emory and Georgia Tech, Kristen received a B.S. and M. Eng in Biomedical Engineering from Rensselaer Polytechnic Institute. She completed a PhD in the Department of Biomedical Engineering at Northwestern University supported by an NIH NIA NRSA F31 fellowship, investigating how the nervous system varies the mechanics of the ankle across different tasks and how that control is impacted by healthy aging.
Laura Marchal-Crespo, TU Delft
Title: Facilitating Learning of Wearable Lower-limb Exoskeletons with Immersive Virtual Reality
Recent advancements in wearable lower-limb exoskeletons have prompted research into integrating feedback systems to facilitate their use. For users to successfully use this technology, they first need to undergo a long and tedious learning process on how to control the devices, e.g., by shifting the weight from one leg to the other and balancing. This is especially challenging for individuals with sensorimotor disorders who might also suffer from sensory loss.
In this talk, I will present our efforts in developing a new system based on Immersive Virtual Reality using commercial head-mounted displays for training to control wearable lower-limb exoskeletons for people with sensorimotor disorders. The system simulates a virtual walking task of an avatar resembling the sub-tasks needed to trigger steps with an exoskeleton. I will present the results from an experiment with forty healthy participants that investigated the effects of first- vs. third-person perspective and the provision (or not) of concurrent visual feedback of participants' movements on the walking performance -- namely number of steps, trunk inclination, and stride length --, as well as the effects on embodiment, usability, cybersickness, and perceived workload.
About Dr. Marchal-Crespo: Laura Marchal-Crespo is an Associate Professor at the Department of Cognitive Robotics, Faculty ME. She is also associated with Erasmus MC (Rotterdam) and the Faculty of Medicine at the University of Bern (Switzerland). Her research focuses on the general areas of human-machine interaction and biological learning and, in particular, the use of robotic devices and immersive virtual reality for the assessment and rehabilitation of patients with acquired brain injuries such as stroke. Webpage: www.mlnlab.nl
Maura Eveld, University of Twente
Title: Adaptation to Ankle Exoskeleton Balance Assistance
While the prevalence of human adaptation to ankle exoskeleton support has been demonstrated for general walking assistance, the extent to which humans adapt and/or learn to use ankle exoskeleton assistance aimed at restoring balance has yet to be explored. In this work, 12 healthy participants experienced a training protocol of 180 forward pushes while walking on a treadmill and receiving a constant ankle plantarflexion assistance profile for each push. COM kinematics, ankle mechanics, ankle muscle activity, as well as spatiotemporal metrics were analyzed as a function of exposure (number of pushes) to quantify adaptation to the balance assistance. Furthermore, responses were compared to pre- and post-training responses without balance assistance to determine potential after-effects. This work will provide the first investigation of human adaptation to balance assistance, which can inform training protocols, device controller design, and rehabilitation programs.
About Dr. Eveld: Maura is a Biomechatronics and Rehabilitation Robotics postdoctoral researcher in the University of Twente’s biomechanical engineering department (NL; September 2022-Present). She received her PhD in Mechanical Engineering from Vanderbilt University at the Center for Rehabilitation Engineering & Assistive Technology (US; 2017-2022) and her B.S. in Mechanical Engineering with a concentration in Bioengineering from the University of Notre Dame (US; 2013-2017). Her research encompasses (1) studying human movement control fof healthy and mobility-impaired populations, (2) designing and testing wearable robotics (e.g., powered exoskeletons) for balance assistance, and (3) exploring how humans learn to use such assistive devices.
Maggie Wu, University of Michigan
Title: Insights from Human-Exoskeleton Interactions to Support Cognitive Fit
A key factor impacting human-exoskeleton interaction is cognitive fit, which describes how wearing an exoskeleton can affect sensation, attention, and motor strategies. Cognitive fit can be altered by the exoskeleton’s control strategy and reliability and may cause users to modulate their behavior in response to exoskeleton performance. In this talk, we highlight two studies related to cognitive fit with powered assistive exoskeletons. The first study explores how users modulate their motor strategies when walking with an ankle exoskeleton and pseudo-random perturbations (loss of exoskeleton plantarflexion torque). The second study examines the effect of visual and haptic biofeedback on motor learning with an upper-limb exoskeleton. The results of these studies may inform future exoskeleton controller design to support human-exoskeleton collaboration and cognitive fit.
About Maggie Wu: Man I (Maggie) Wu received her M.S. and Ph.D. in Robotics at the University of Michigan, Ann Arbor (2024) and her B.S. in Biomedical Engineering and Mechanical Engineering at Boston University (2020). Her doctoral research in the Stirling Research Group involved characterizing human-exoskeleton fluency to develop co-adaptive algorithms for assistive lower-limb exoskeletons. Maggie currently works as a software engineer at Medtronic in surgical robotics.
Organisers
Maura Eveld
Katie Poggensee
Sasha Voloshina
Edwin van Asseldonk