Exploring the Cooperative Abilities Between Homogeneous Robotic Arms : An Explorative Study of Robotics and Reinforcement Learning

The field of robotics has witnessed significant advancements in recent years, with robotic arms playing a pivotal role in various industrial and research applications. In large-scale manufacturing, manual labour has been replaced with robots due to their efficiency in time and cost. However, in order to replace human labour, the robots need to collaborate in a way that humans do. This master's thesis, conducted at the Cyber-physical Systems Lab (CPS-Lab) at Uppsala University, delves into the intricacies of cooperative interactions between two homogenous robotic arms powered by machine learning algorithms, aiming to explore their collective capabilities. The project will focus on implementing a multi-agent cart-pole experiment that will challenge the two robotic arms' cooperative abilities. First, the problem is simulated, and afterwards implemented in real life. The experiment will be evaluated by the performance of various tested machine learning algorithms. In the end, The simulation yielded poor results due to the complexity of the problem and the lack of proper hyperparameter tuning. The real life experiment failed instantly, caused by the robotic arms not being designed for this application, a large simulation gap, and latency in the controller design. Overall, the results show that the experiment was challenging for the robotic arms, but that it might be possible under different circumstances.