Our goal is to enable autonomous systems - such as the flying robot in the video - to 'learn' the way humans do: through practice.
We use the information of past experiments to improve the robot's performance. After each trial the commands that are send to the robot are updated.
The learning algorithm can handle changing environments, see last part of the video!
Work by Angela Schoellig, Fabian Mueller and Raffaello D'Andrea.
At the ETH Flying Machine Arena.
*More*
http://www.idsc.ethz.ch/Research_DAndrea/Iterative_Learning
http://www.FlyingMachineArena.org
this must be a bitch to program and calibrate .-.
Maxwilso 2 months ago
Did you use genetic algorithms?
PaXx 3 months ago
- YOU CAN WATCH LOTS OF HUMANOID ROBOTS ON MY CHANNEL -
luisbeck007 5 months ago
Unexpected air disturbance! *turns on the fan*
Boehoehuahoei 6 months ago