Human-robot collaboration: optimizing stress and productivity based on game theory скачать в хорошем качестве

Human-robot collaboration: optimizing stress and productivity based on game theory

First, the video illustrates the HRC experimental framework where the proposed approach has been validated. This consists of a human operator, an ABB dual-arm cobot, a button (which is used by the human to inform the robot that he/she has finalized one workpiece) and a ProComp Infiniti encoder (which is used to acquire the ECG signal via disposable electrodes). The use-case consists in a synchronized collaborative task where the human operator and the cobot are required to assemble the components of a box. Then, the video shows the crucial phases of the proposed robot control strategy during the task: the human stress is evaluated based on the analysis of HRV indicators extracted by the ECG signal, while his/her cycle-time (performance) is evaluated through the button. Based on these data, the proposed game-theoreretic strategy allows to tune a suitable reward. This is used by a Learning Automaton to decide the most suitable next robot action to improve the current stress-performance state. In particular, at each step, the robot can increase the production rhythm to stimulate the human or synchronize with the pace dictated by the him/her.