AEEM 9074: Control of Networked Cyber-Physical Systems

From fleets of autonomous drones to distributed sensor networks and smart grids, the study of Networked Multiagent Systems forms the backbone of many modern technological innovations. This course provides a comprehensive introduction to the principles, theories, and applications of these systems, focusing on how multiple agents—each with limited local information—can collectively achieve global objectives. At the heart of multiagent systems lies the concept of multiagent control, which governs how individual agents interact and coordinate their actions. Each agent operates autonomously yet contributes to a shared goal, such as maintaining a formation, achieving consensus, or distributing tasks efficiently. This decentralized approach, where decision-making is distributed across agents, offers significant advantages in terms of scalability, fault tolerance, and adaptability to dynamic environments. An essential framework for understanding the interactions within multiagent systems is graph theory, which models the network structure connecting agents. In this context, nodes represent the agents, while edges signify communication links. Key concepts such as connectivity, graph Laplacians, and eigenvalues provide insights into how well agents can exchange information and collaborate. The course also covers cyber-physical systems (CPS), where multiagent systems interact with the physical world. CPS integrate computational algorithms with physical processes, creating systems that sense, compute, and actuate in real time. Within these systems, ensuring security, efficiency, and resilience is paramount, as disruptions—whether from cyber-attacks or hardware failures—can have significant consequences. The course highlights strategies to design robust systems capable of withstanding such challenges.