Researchers are helping robots avoid expensive face plants

Sponsored Links

Researchers are helping robots avoid expensive face plants

Why are robot tumbles comedy gold? Unlike humans, they make no effort to protect themselves, resulting in flailing, extra hard falls (bottom). But the high-g impacts are hell on the extremely expensive, often one-of-a-kind machines, so researchers from Georgia tech developed algorithms to give them some sense of self-preservation. They made them copy exactly what we do instinctively -- stick out a limb to break the fall. "(That way), every time you make contact with the ground, some of the energy is dissipated," said Georgia Tech professor Karen Liu.

That algorithm helps robots to quickly contort their bodies to create the maximum number of contact points to spread the force of the fall. With tests on a small robot, the algorithm helped the robots land with considerably less force -- 5.5 g versus 8.04 g in the head area with a moderate push, for example. During a hard push, the robot was able to stick out a leg for a more graceful summersault (below).

Right now, robots like Big Dog and Atlas from Google-owned Boston Dynamics go limp when they detect a fall. However, as the robots become more advanced, they'll need better loss-of-balance tech to not just protect the robots, but the humans they could fall onto. After all, Boston Dynamic's Atlas stands 6 feet tall and 330 pounds, so it could cause considerable mayhem during a badly-timed spill.

All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission.
Popular on Engadget