Formal Modelling for Multi-Robot Systems Under Uncertainty

Abstract Purpose of Review To effectively synthesise and analyse multi-robot behaviour, we require formal task-level models which accurately capture execution. In this paper, review modelling formalisms for systems under uncertainty discuss how they can be used planning, reinforcement learning, model checking, simulation. Recent Findings work has investigated more execution by considering different forms uncertainty, such as temporal partial observability, the effects robot interactions on action Other strands have presented approaches reducing size to admit efficient solution methods. This achieved decoupling robots independence assumptions or reasoning over higher-level macro actions. Summary Existing demonstrate a trade-off between capturing dependencies being small enough tractably solve real-world problems. Therefore, future research should exploit realistic behaviour develop smaller retain accurate representations interactions; structure problems, factored state spaces, scalable