Coordinating a robot swarm in a warehouse by aggregate programming: A case study

Multi-robot systems (MRS) are becoming increasingly relevant within diverse application domains, including warehouse logistics, where robots are in charge of moving and storing crates, by navigating through the warehouse shelves. Engineering such robust multi-robot systems is a significant challenge, especially if the goal is to build a fully decentralized system, without single points of failure, or potential bottlenecks. Aggregate Programming (AP) is an established approach to engineering resilient distributed systems, which has already been applied successfully for building prototypes in Industrial Internet of Things and robotics scenarios. In this paper we present an in depth description of how we have exploited AP for designing and simulating a MRS to address a non-trivial farm warehouse logistics case study proposed by a company operating in the Autonomous Mobile Robots (AMR) market, which cooperated in this research. Importantly, to better validate the implemented prototype, we have relied on two different levels of simulation, one scalable and high-level, and another one small-scale but highly realistic.