We developed several demos show casing our new algorithms.

Learning from Demonstration

We are able to learn complex manipulation tasks from demonstrations and from an interactive process of correction.

Robot Programming from Demonstration and Feedback (

Robot Programming from Knowledge Transfer (

Learning and Adaptation in Cooperative Human-Robot Tasks

Efficient Behavior Learning in Human-Robot Collaboration

In several cases tasks need to be executed in close collaboration with a human user

Collaborative work with robots. Insertion Task with Comfort Optimization and Automatic Task Verification (

Preemptive Human-Robot Collaboration (

Collaborative assembly with phase estimation (ISRR 2015) (

Relational Activity Processes for Modelling Concurrent Cooperation (Learning and Planning) (

Gesture-Based Cooperative Assembly (

Anticipating the human motion by recognizing the phase of his/her movement

Transferring human demonstrated skills on scenarios with obstacles

Motor Control and Planning

Multi-Bound Tree Search for Logic-Geometric Programming in Cooperative Manipulation Domains

The tasks can then be executed using geometrical planning methods

Verification of actions Attach:demoverification.mp4

zum Seitenanfang