Besides simply being fascinating to watch, insect-inspired robots may someday find use as scouts in search-and-rescue operations. In order for them to function in such scenarios, however, they'll have to be able to move through fields of debris. While some scientists have looked at using sensors and algorithms that let the bots scan their surroundings and then plot paths around obstacles, researchers at UC Berkley have developed a much less complex but still effective approach – they've outfitted a robotic cockroach with a streamlined shell, that lets it just push its way through.
Led by postdoctoral fellow Chen Li, the scientists started by using high-speed cameras to observe the manner in which real cockroaches made their way through vertical grass-like obstacles. It turns out that the elongated disc shape of the insects' carapace (body shell) not only helped keep them from getting snagged, but it also naturally caused them to turn sideways, presenting a narrower profile that let them slip through gaps more easily.
In fact, when the cockroaches were outfitted with an artificial rectangular shell, it became much more difficult for them to get past those same obstacles. When an existing rectangular cockroach-inspired robot was initially tested in a similar course, it likewise didn't do very well.
To help that robot, the scientists encased it in an oval shell, inspired by the cockroach's carapace. When it was subsequently retested, the addition of the shell caused it to automatically execute the insect's "roll maneuver," turning on its side and sliding through spaces that had previously stopped it. No reprogramming or additional electronics were involved.
As an added bonus, the shell also helps protect the robot's electronics from water and dust. Li and his team are now studying other natural body shapes, to see if they could help robots not only overcome obstacles but also excel in tasks such as climbing.
A paper on the research was recently published in the journal Bioinspiration & Biomimetics. The cockroach and robot obstacle tests can be seen in the video below.
Sources: University of California, Berkely, IOP Publishing via IEEE Spectrum
