Specification Aware Multi-Agent Reinforcement Learning

@article{ ritzICAARTBook22,
    author      = "Fabian Ritz and Thomy Phan and Robert Müller and Thomas Gabor and Andreas Sedlmeier and Marc Zeller and Jan Wieghardt and Reiner Schmid and Horst Sauer and Cornel Klein and Claudia Linnhoff-Popien",
    title       = "Specification Aware Multi-Agent Reinforcement Learning",
    year        = "2022",
    
    
    abstract    = "Engineering intelligent industrial systems is challenging due to high complexity and uncertainty with respect to domain dynamics and multiple agents. If industrial systems act autonomously, their choices and results must be within specified bounds to satisfy these requirements. Reinforcement learning (RL) is promising to find solutions that outperform known or handcrafted heuristics. However in industrial scenarios, it also is crucial to prevent RL from inducing potentially undesired or even dangerous behavior. This paper considers specification alignment in industrial scenarios with multi-agent reinforcement learning (MARL). We propose to embed functional and non-functional requirements into the reward function, enabling the agents to learn to align with the specification. We evaluate our approach in a smart factory simulation representing an industrial lot-size-one production facility, where we train up to eight agents using DQN, VDN, and QMIX. Our results show that the proposed approach enables agents to satisfy a given set of requirements.",
    url         = "https://link.springer.com/chapter/10.1007/978-3-031-10161-8_1",
    eprint      = "https://arxiv.org/pdf/2012.07949.pdf",
    
    publisher   = "Springer-Verlag",
    journal     = "Agents and Artificial Intelligence: 13th International Conference",
    pages       = "3--21",
    keywords    = "reinforcement learning, multi-agent systems, specification compliance, AI safety",
    doi         = "https://doi.org/10.1007/978-3-031-10161-8_1"
}