Design of an Intelligent Control System for Safe Collaboration between Human and a Robotic Manipulator

Human Robot Collaboration (HRC) is a concept in which a human and a robot such as manipulators divide work tasks in a shared work area. In the environments where robots are employed, various safety systems such as fences and barriers are used to ensure the safety of common workspace. This paper presents a new intelligent control system for safe collaboration between human and a robotic manipulator. The designed control system has two main parts. The first part is object detection, where deep learning algorithm YOLOv3 has been implemented to detect various objects with different shapes. The images are partly captured by Intel® RealSense™ Depth Camera D435i and partly are taken from the Microsoft Common objects in context (COCO). The second part is a controller that enables the manipulator to move safely. To investigate the performance of the designed control system, a number of sample collaborative tasks have been simulated in the software Robot Operating System (ROS). The provided simulation results validate the proper performance of the object detection algorithm and the robot controller.