Researchers have unveiled a groundbreaking model that sheds light on how humans continuously adapt during complex tasks, such as walking, while maintaining stability. This innovative research is detailed in a recent Nature Communications paper, co-authored by Nidhi Seethapathi, an assistant professor at MIT’s Department of Brain and Cognitive Sciences, Barrett C. Clark, a robotics software engineer from Bright Minds Inc., and Manoj Srinivasan, an associate professor in the Department of Mechanical and Aerospace Engineering at Ohio State University.
In episodic tasks, such as reaching for an object, mistakes made during one instance don’t impact the next. However, in locomotion, errors can lead to a sequence of short- and long-term stability challenges, making adaptation in a new environment particularly complex.
“Traditionally, our understanding of adaptation has focused on episodic tasks, like reaching for an item in a new space,” says Seethapathi. “This new theoretical framework captures the complexities of adaptation in continuous, long-horizon tasks across various locomotor scenarios.”
To create the model, the researchers pinpointed fundamental principles of locomotor adaptation across diverse task settings and developed a unified, modular, and hierarchical approach, with each component structured mathematically in a unique way.
This comprehensive model effectively illustrates how humans adjust their walking in unfamiliar environments, such as on a split-belt treadmill with differing speeds for each foot, while wearing uneven leg weights, or during the use of exoskeletons. Notably, the model accurately reproduced human locomotor adaptation behaviors observed in ten prior studies and successfully predicted adaptations in two new experiments carried out as part of this research.
The implications of this model extend to fields like sensorimotor learning, rehabilitation, and wearable robotics.
“A predictive model of how individuals adjust to new environments is invaluable for enhancing rehabilitation programs and controlling wearable robots,” remarks Seethapathi. “One could view a wearable robot as a new environment itself, and our model offers insights into how individuals will adapt to various robotic settings. Understanding human-robot adaptation often requires extensive experimentation, and our approach could streamline this process by refining the areas to be explored.”
Photo credit & article inspired by: Massachusetts Institute of Technology
