Application of Multi Agent Reinforcement Learning to Robotic Pick and Place System

Recent advances in deep reinforcement learning are enabling the creation and use of powerful multi-agent systems in complex areas such as multi-robot coordination. These show great promise to help solve many of the difficult challenges of rapidly growing domains such as smart manufacturing. In this paper we present a novel simulator for a multi-robot pick and place system, leveraging the OpenGym framework. We further evaluate the performance of three distinct reinforcement learning algorithms, name as Qmix, VDN, and IQL, employing the Epymarl framework with our simulator. Our primary objective is to show the effectiveness of these algorithms within a manufacturing context, with a specific focus on their impact on the gripping rate-a vital metric for assessing performance and efficiency.