Academics

Explaining patterns, one model at a time

A three-dimensional representation of the two-dimensional walker generated by Cusumano and his collaborators to investigate stability in walking. Credit: Soumyabrata Maiti. All Rights Reserved.

UNIVERSITY PARK, Pa. — A flock of birds moves in a way that seems random yet orderly. A ceiling fan rattles at a certain speed. Exploring the patterns of everyday phenomena is what piqued Joe Cusumano’s interest in dynamics — whether they are in nature or the living room.

Cusumano, professor of engineering science and mechanics, is largely focused on dynamical systems research. He uses mathematics to learn about the behavior of these dynamical systems, which evolve over time according to their current state. Cusumano specifically creates models, intended as simplified yet accurate representations of natural patterns, such as the undulating waves that appear along the edge of a rapidly spinning tire, to better understand these behaviors.

“What does it mean to understand something?” Cusumano asked. “For a researcher like me, understanding means you have a model. If you don’t have a good model, you don’t understand it.”

Cusumano is applying his methodology to better describe walking — and what factors contribute to falls or instability. He serves as co-investigator on two National Institute of Aging grants, totaling almost $2 million in funding. Working with principal investigator Jonathan Dingwell, professor of kinesiology, and former student Navendu Patil, currently a postdoctoral research associate in kinesiology at Penn State, Cusumano is developing physics-based models that mimic a person’s movement while walking.

In one study, Cusumano and his collaborators investigated the relationship between walking regulation strategies, such as maintaining a constant speed, and the likelihood of falling. The researchers simulated a two-dimensional walking model consisting of two straight legs joined at a circular "hip." The walker, though lacking knees, muscles and other structures of an anatomical human representation, pushes off with each step, switches legs in its gait and alternates standing on each leg as it moves — it walks like a human. Using simulations of the walker moving on a treadmill, the researchers found that the regulation of speed while walking promotes stability and can therefore help prevent disruption from a small disturbance, such as a random change in push-off force, or a large one like a trip or shove. The findings, according to Cusumano, can inform future research on minimizing fall risk — particularly in older adults, for whom falling can result in more severe injuries.

“I’m very interested in things at the human scale,” Cusumano said. “Billions of dollars per year are spent on medical costs for fall injuries. Our research is about comprehending the causes for these kinds of falls and developing models that can provide insight to researchers in other fields examining the same problem.”

Cusumano emphasized that his work — from understanding human experiences to theoretical explorations of the underlying concepts that give rise to them — would not be possible without collaboration across disciplines. 

“Starting from the math and models for phenomena as commonplace as the vibration of a door in a fast-moving car, there are various application domains to explore,” Cusumano said. “I was drawn to this field because I enjoy talking to researchers in other fields and learning about what work they’re doing to examine these kinds of problems.”

Last Updated February 11, 2021

Contact