Mechanisms Underlying the Energy Optimization of Human Walking

For many people, walking is something done every day, without much thought and with relative ease. People can gracefully adapt their gait to changing terrains, task demands, and constraints on their body. Yet, how the nervous system navigates the expanse of all possible movements, to rapidly arrive at what appears to be an optimal coordination strategy, is largely unknown.

Here, we are interested in understanding how energy optimization is accomplished by the nervous system during locomotion. To do so, we use lightweight robotic exoskeletons to perturb the human walker—changing the energetically optimal way to move. We then combine behavioural data from these walking experiments with computational reinforcement learning models to inductively reason about how the nervous system optimizes coordination.

This work has a number of broad applications including: understanding gait adaptations that accompany aging or injury; designing rehabilitation programs that consider the nervous system’s internal objectives; designing assistive technologies that work in concert with the user to augment movement; and expediting locomotor skill learning.

Funding: National Sciences and Engineering Research Council, Discovery Grant (RGPIN-2019-05677); Canadian Foundation for Innovation, John R. Evans Leaders Fund.

Improving gait performance using energy incentivized movement therapy

Gait rehabilitation strategies often focus on restoring a desired ‘normal’ or ‘healthy’ gait, and for good reason—this is typically desired by those being rehabilitated. Traditional rehabilitation strategies often directly target the desired kinematically ‘normal’ gait, through repetitive practice under the guidance of a therapist. The expected outcome is that the desired gait will eventually be adopted and energetic costs will decrease. Here, we explore an alternative approach—directly targeting the energetic consequences of movement to make the desired gait energetically optimal. We expect the desired gait to then be naturally adopted by the individual, potentially leading to more effective and enduring rehabilitation.

This research requires an interdisciplinary team that combines expertise in fundamental human biomechanics, clinical rehabilitative medicine, and applied robotic control. Our approach, if effective, could be extended to individuals with a range of mobility impairments and has the potential to inform the next generation of assistive robotics and rehabilitation strategies.

Collaborators: Dr. Ryan Roemmich, Assistant Professor Physical Medicine and Rehabilitation John Hopkins University; Dr. Amy Wu, Assistant Professor Mechanical Engineering Queen’s University; Dr. Jennifer Tomasone, School of Kinesiology and Health Studies Queen’s University

Funding: New Frontiers in Research Fund – Exploration Grant (NFRFE-2018-02155); Canadian Foundation for Innovation, John R. Evans Leaders Fund.

Using large-scale wearable data to understand ecological gait

Understanding ecological human movement—and the myriad of factors that affect it—has been historically difficult. Yet, we are living through an unprecedented change in how human movement can be monitored, managed and understood. Mobile health technologies, operating on smartphones and other wearable devices, are being used by millions of people around the world to track their activity—be it steps, or heart rate, or other metrics—as they go about their daily lives.

Here, we are using large-scale data sets collected from wearable devices to understand real-world locomotion behavior and develop validated and reliable gait metrics.

Funding: Canadian Foundation for Innovation, John R. Evans Leaders Fund.