Collision Driven Multi Scenario Approach for Human Collaboration with Industrial Robot

Abstract

This study focuses on the problem of robot interaction with a dynamically changing environment. Particular attention is paid to the problem of human collaboration with industrial robot in a shared common workspace. The paper identifies collisions and provides algorithms for different scenarios of obstacle avoidance, considering the nature of interaction and contact point location. The developed mathematical framework is based on the neural network classification and finite state machine, followed by appropriate collision reaction/avoidance algorithms. The advantages of the developed approach were demonstrated by an experimental study dealing with Kuka LBR IIWA 14 robot interaction with a human and dynamic environment.