Leonel Rozo


A robot learning to cooperatively transport an object while interacting with different compliance levels
March 9, 2016

In this video we show a human-robot collaborative task where the robot learns to handle both position and force constraints. At the beginning of the task, the two participants simultaneously reach for the object. Then, they start jointly transporting it along a given path to reach the target location. When the object gets to the final position, the human-human pair releases it and moves away from the object. The robot learns not only to adapt to changes of the task parameters (initial and final location of the object), but also to display different compliance levels according to the desired dynamics of the collaborative tasks.
 

A robot learning to assist a human in assembling an IKEA table while interacting with different compliance levels
March 9, 2016

This video shows how a robot learns to assist a human in assembling an IKEA wooden table. In the learning phase, a human kinesthetically teaches the robot its collaborative behavior from demonstrations. Using the training data, the robot is able to exploit the force information and the parameters of the task (position and orientation of the leg) to learn different impedance-based behaviors in order to successfully collaborate with its human partner.
 

A robot extracting task-frame authority from human demonstrations
January 13, 2017

We show in this video how a robot autonomously devises an appropriate control strategy from human demonstrations without a prior knowledge of the demonstrated task. The method is primarily based on observing the patterns and consistency in the observed dataset. This is obtained through a demonstration setting that uses a motion capture system, a force sensor and muscle activity measurements.
Video credits: Dr. Luka Peternel, Dr. Arash Ajoudani
 
Back
Top