ARMAR IIIb Humanoid Robot Learns to Wipe a Table

Visual feedback is used to learn a basic wiping trajectory. After it is learned, feedback is switched to force-feedback. The height of wiping is adjusted. The height is continuously learned and after a few periods finally learned. The robot thus know how to maintain contact with the table. The work was performed at KIT.

 

Leave a Reply

Your email address will not be published. Required fields are marked *