Human robot Interaction - Object Detection and Distance Measurement Using Kinect V2

A collaborative robot is a specific type of robot that is aware of the environment, which it shares with its human operator. After the recent advancement of Industry 4.0, these robots are being rapidly adopted by the production and manufacturing industry. In comparison to a traditional industrial robot, collaborative robots offer enhanced security, process optimization, and versatility. This paper proposes a system that utilizes a time-of-flight vision camera, the Kinect V2, and an ultrasonic distance sensor to monitor the speed and separation of robots operating in shared workspaces with humans. The current issue with robots that lack additional sensors is that they stop automatically upon detecting collisions, resulting in unnecessary downtime and wear and tear on equipment. By continuously detecting the presence of human workers using the Kinect V2 and the Distance Ultrasonic sensor, this system can calculate the distances between all robot joints and humans based on joint angle information provided by robot control. Once the shortest distance, or critical distance time, is identified, the robot's velocity and acceleration values can be adjusted to safe values. This real-time detection and distance measurement system can improve the safety of human-robot interactions and prevent collisions with people or objects. The algorithm was tested in both virtual and actual environments using VREP software and the proposed setup. By observing human skeletons and object recognition, the robot's speed can be adjusted according to the distance at which people enter a shared workspace.