Scientists at the Laboratory of Intelligent Systems in Lausanne, Switzerland, have been hard at work inventing the world’s first flying robot that can crash, pick itself up, and fly off again.
“We build robots that are better adapted to the real world, not just the lab environment,” says Adam Klaptocz, a doctoral student working on the project.
So what’s the advantage of a crashing robot other than the fun of watching it fly into things?
The applications are many, according to Klaptocz, including exploration of dangerous, hard-to-reach collapsed mines, dark caves, and irradiated nuclear plants.
The latest incarnation, the AirBurr, uses optic-flow based algorithms resembling those used by insects. The robot’s moving parts are protected in a carbon cage and it has carbon-fiber “legs” so that when it bumps into an obstacle, it has an active recovery system that allows the robot to bounce back up again, no matter how bad the crash. Klaptocz and his team accomplished this by fiddling with the dimensions to give it the right force to stand up again without changing the center of gravity.
While the basic engineering is the same, the robot comes in many iterations including the “Samurai,” allowing it to return to an upright position and take off after a collision, no matter how it lands on the ground; “Sticky,” with gecko-inspired dry adhesive to attach to smooth surfaces; “Bumpy,” which has a sense of touch; and “HoverMouse,” which has a gravity-based self-recovery system to allow it to return to takeoff position after collisions.
Next up for the lab: Fying robots with folding wings that can roll on the ground, using audio and sonar like bats for detecting other airborne robots. After that? Flying robots that can break into pieces, fit back together again, then resume their flight path.