For researchers seeking links between biology and robotics, dogs are truly living up to the "man's best friend" axiom.
Andrew Spence, associate professor of bioengineering, served as primary investigator on the paper, which investigated how dogs alter their gait on various terrains. Dr. Spence collaborated with a team of six scientists from the University of Pennsylvania, Carnegie Mellon University, the University of Michigan, and the Royal Veterinary College in the United Kingdom.
The team is currently using robotic replicas to confirm whether the adjustments made from different speeds on different surfaces, like sand or uneven trails, make the animal more stable, and potentially more energy efficient.
Spence first encountered the intersection of biology and robotics at the University of California, Berkeley. After first studying physics, he changed fields after meeting Bob Full, a professor who traded in what Spence called "bio-inspiration."
"You meet Bob and he blows your mind after 20 minutes and you're like, I want to do this," Spence said.
Spence left Berkeley with his wife for the United Kingdom. He found kindred spirits at the Royal Veterinary College and began examining the pioneering studies by Mildred Hildebrand in the 1960s that first analyzed quadruped gaits at varied speeds.
While studying dogs at the Royal Veterinary College, he drew on his physics background to hone his observations with mathematics. Hildebrand created the overall graphs with broad strokes of observation, while Dr. Spence, Daniel Koditschek and Shai Revzen at the University of Pennsylvania, used the applied math developed by roboticists to make analyzing complicated quadrupedal gaits tractable.
According to Spence, Robotics hinges greatly on gait and balance, mainly so the robots can move around without falling over.
"You need to know the math you're putting in the robot is robust," said Spence. "So, we realized we could use some of these mathematical tools to try to get more insight into the gaits that animals use."
Visiting professor David Carrier from the University of Utah opened the group's eyes to the idea of using bumpy terrain to examine gait adaptation for balance. Think of how you sometimes speed up after your trail foot hits a tree root or uneven sidewalk to match your momentum and avoid falling.
As primary investigator, Spence sought the way to use the math to quantify the adjustments the dogs made in their gaits over the rough terrain. This gave the group a chance to update Hildenbrand's work with some new theoretical tools. Analyses showed that on this rough terrain, dogs will adapt their walking gait to be more trot-like, even though they continue to move at slow speeds where they would normally walk, increasing their stability.
"This paper is a story of a long collaboration with Shai and Dan using some of the math they use for robots to get some more insight into what animals do," Spence said.
"These finding could be used to improve the stability and perhaps even energetic economy of our legged robots."
His group showed off their findings through robots that use animal gaits and showcased them overseas.
Currently, he runs a lab at Temple University devoted to studying the neuromechanics of locomotion, both in animals and now in medical model rodents, mice and rats. Now he has engineers around him to work with the robots and test these theories. His team will work to advance the understanding of neuromechanics, one step—and paw—at a time.