Multiagent Systems: Planning and Coordination

This lecture introduces multi-agent systems, focusing on the reasons for their use, such as the need for multiple agents to solve tasks, security against manipulation, and modeling real-life scenarios. The instructor discusses the benefits of cooperation among agents, different architectures for multi-agent systems, and the challenges of self-interest. The lecture covers various planning and coordination strategies for reactive and deliberative agents, emphasizing the importance of joint plans and mechanisms to motivate agents. Additionally, it explores the concept of task assignment using contract nets and market-based approaches, highlighting the role of communication and ontologies in coordinating heterogeneous agents.