Quasi Hands Free Interaction with a Robot for Online Task Correction

Abstract

New human machine interfaces are becoming more and more necessary in robotics applications where the robot operates in close cooperation with a human, and one near to each other. Many of these applications require intuitive interaction systems since they are used by non robotic experts, thus the use of natural language based interfaces (either voice or gestures) are highly convenient. This research tackles the problem of interpreting the user arm and hand gestures as orders to the system, while the user is working close to the robot. The proposed human-machine interface looks for a compromise between classical interfaces and the hands free operation requirements.