Researchers from the University of Lorraine have developed the D-Reflex system, according which uses the TALOS humanoid robot to lean against a wall when one of its legs is broken, like a person who has just lost their balance. The neural network-based system uses its experience (in this case, 882,000 training simulations) to quickly find the point on the wall that is most likely to provide stability.

As a result, instead of falling to the ground, the robot is pressed against the wall like a person who has just sprained an ankle. It’s not very graceful, but it’s effective in three out of four tests.

D-Reflex does not prevent falls, it cannot account for all possible positions or surfaces. It also does not help work to recover.

However, the researchers hope to create a system that will be useful in motion, and envision robots that can grab chairs and other complex objects when a fall is imminent. This could save the cost of replacing worker robots that would otherwise fall and crash, and could lead to more “natural” robots that learn to use their environment to their advantage.