Science: New, Off-Road Robot Inspired by Nature

VIDEO: Demonstrations of the six-legged robotic device walking across a bed of dry, loose grains.
[Courtesy of Chen Li, Tingnan Zhang, Daniel Goldman]

Researchers can learn a lot from a lizard scampering across the hot desert sand or an insect crawling atop a pile of plant litter. Chen Li and colleagues from the Georgia Institute of Technology in Atlanta took cues from such creatures and designed a robot that uses six legs to traverse a bed of dry, loose grains.

The robotic design isn’t as effective as a lizard’s but it can move through sand at a reasonable pace without getting stuck, and it may help to boost the performance of roving and walking robots, such as the Mars rovers, the researchers said. They noted that previous studies of objects moving through air and water have led to improvements of industrial products such as aircraft wings and underwater robots.

“There’s only going to be an increasing number of robots running around our planet and others,” said Daniel Goldman, a co-author of the report that appears in the 22 March issue of Science. “We’d like to identify principles that allow devices to move effectively under diverse conditions.”

Insight gained from the model could also shed more light on the evolution of animal species.

“On the biological side of things, this robot and the associated ‘terradynamics’ help us to understand why animals have certain leg shapes and toes—both now and in the past,” said Goldman. “We now have a model that can generate hypotheses and predictions about that.”

0321_robot

Daniel Goldman and Chen Li. [Courtesy of Chen Li, Tingnan Zhang, Daniel Goldman]

Interactions with “flowable ground,” or surfaces like sand, soil, mud and grass, can often be more complex than interactions with fluids. So, Li and his colleagues built upon previous studies of insects and lizards to identify the optimal leg shapes and stride frequencies for traversing such deformable terrain. They considered various grain sizes and shapes—from small, round grains to larger, kidney-shaped grains—and used computer simulations to test different robotic designs.

“We wanted to see if we could develop a system to predict how a robot would move on different granular surfaces,” explained Goldman. “The resistive force theory, which was developed by some of the pioneers of fluid mechanics back in the ‘30s and ‘40s, works surprisingly well when applied to legs rotating in the sand—that is, when it’s supplemented with the appropriate drag and lift force relations for this granular media.”

Taking into account these forces and their varying influences on each robotic leg as the leg enters and exits the different granular surfaces, Chen Li and co-author Tingnan Zhang settled on a particular robotic design that optimized each step regardless of the grain shapes or sizes.

“We’re thinking of this study as robo-physics in addition to biological physics,” according to Goldman. “If you take a box, put a few legs on it and let it whirl around a bit, you may suddenly find dynamics that you can’t predict. The challenge is to understand these dynamics, which is a broader aspect of our study.”

Links

Read the abstract, “A Terradynamics of Legged Locomotion on Granular Media,” by Chen Li et al.

Listen to a related Science podcast with co-author Daniel Goldman