Impact Force Minimization Algorithm for Collaborative Robots Using Impact Force Prediction Model

Recently, the demand for collaborative robots is increasing in the industrial field. However, as the collaborative robots share the same workspace with human workers, there is a high possibility of collision between the robot and the worker. A possible method to ensure the safety of a human worker is to restrict the impact force that the robot exerts on the worker during a collision. That is, if the impact force can be predicted, the robot motion that causes excessive impact force can be detected and handled properly before the actual robot motion. To this end, an algorithm for predicting the impact force generated by a collision is proposed, and a method for ensuring the human safety, by modifying the trajectory of the robot when the excessive impact is predicted with current motion, is investigated. To establish the impact force prediction model, collision experiments were performed with a 6-DOF collaborative robot and a dummy. Moreover, an algorithm for minimizing the impact force, by reducing the end-effector velocity of the robot when excessive impact is predicted from the established model, is proposed to ensure the human safety. The performance of the algorithm was verified through various experiments.