Scientists have developed a model to explain how humans continually adapt to complicated tasks such as walking while maintaining stability.
The findings were detailed in a recent issue article published in the magazine by Nidhi Seethapathi, assistant professor in MIT’s Department of Brain and Cognitive Sciences; Barrett C. Clark, robotics software engineer at Dazzling Minds Inc.; and Manoj Srinivasan, associate professor in the Department of Mechanical and Aerospace Engineering at The Ohio State University.
In episodic tasks, such as reaching for an object, errors made in one episode have no effect on the next. In tasks such as locomotion, errors can have a cascade of short- and long-term consequences on stability if left unchecked. This makes the challenge of adapting locomotion to a novel environment more complicated.
“Most of our previous theoretical understanding of adaptation was limited to episodic tasks, such as reaching for an object in a new environment,” Seethapathi says. “This new theoretical model captures adaptive phenomena in continuous, long-term tasks in multiple locomotor settings.”
To build the model, researchers identified general principles of locomotor adaptation in various task settings and developed a unified modular and hierarchical model of locomotor adaptation in which each component has its own unique mathematical structure.
The resulting model successfully reflects how people adapt to walking in novel conditions, such as on a split-belt treadmill, with each foot moving at a different speed, with asymmetric leg loading and an exoskeleton. The authors report that the model successfully reproduced human locomotor adaptation phenomena in novel conditions in 10 previous studies and correctly predicted the adaptive behaviors observed in two novel experiments conducted as part of the study.
The model has potential applications in sensorimotor learning, rehabilitation, and wearable robotics.
“Having a model that can predict how a person will adapt to a new environment has great utility in developing better rehabilitation and control paradigms for wearable robots,” Seethapathi says. “You can think of the wearable robot itself as a new environment that a person can move into, and our model can be used to predict how a person will adapt to different settings of the robot. Understanding such human-robot adaptation is currently an experimentally intensive process, and our model can accelerate this process by narrowing the search space.”