Collision-Risk Assessment Model for Teleoperation Robots Considering Acceleration

Owing to the development of robotics and the emergence of Industry 4.0, robotics applications are continuously expanding in various fields, with teleoperated robots widely used in space exploration, medical care, and other fields owing to their strong operability and high precision. Studies have shown that robots are susceptible to collision with obstacles in complex environments, resulting in equipment damage, reduced production efficiency, and operator safety issues. This study focuses on predicting robot collisions and their intervention. Based on the characteristics of teleoperated robot motion, an acceleration discrimination factor is introduced in addition to factors such as distance and speed to provide a collision-risk detection model for single obstacles. First, utilizing a risk matrix for screening and categorizing indirect risks generated by robots during motion, a teleoperated robot collision-risk detection model based on acceleration is established, with a prediction cycle set to ensure the safety of robot motion. Second, based on the actual conditions of the experimental site, severity levels and weights are allocated to indirect factors to calculate the safety and dangerous collision-risk thresholds. Third, experiments are conducted using the UR5e six-degree-of-freedom robot and the Omega.7 high-precision manipulator to validate model effectiveness. Finally, the magnitudes of acceleration and deceleration movements are adjusted based on the different requirements of the robot tasks, thus significantly enhancing robot efficiency while ensuring safety. Results indicate that the proposed acceleration collision-risk model outperforms conventional models in terms of risk-detection accuracy, motion efficiency, motion type, and integration with collision factors.