Service robot learning flexible object manipulation planning from human demonstrations





Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Dec 22, 2011

www.dexmart.eu: Year 3 review - FZI video demo

This video teaser shows interim results and partial algorithm integration for learning autonomous manipulation planning of bi-manual, dexterous object handling in complex scenes from human demonstrations.
It was produced in the scope of the EU project Dexmart (FP7 ICT) for the 3rd project year review in April 2011 by FZI Karlsruhe.

The robot is able to apply learned task knowledge when dealing with new objects and scene configurations by acquisition of important qualitative aspects of manipulation tasks as motion planning constraint strategies.
Autonomous subtask (e.g. moving away obstacles) and motion planning leads to new, highly varying execution time trajectories, generated by online robot reasoning.

The aim of this research is you teaching your robot naturally like a child one day, with the robot being able to perform those tasks in situations it has not encountered before.

Stay tuned for full integration of more learning and planning layers which is currently under way.


When autoplay is enabled, a suggested video will automatically play next.

Up next

to add this to Watch Later

Add to

Loading playlists...