The course aims at presenting a family of distributed artificial intelligence systems, called multiagent systems, composed of autonomous entities – the agents – that interact together is purposeful ways. The course focuses on cooperative interaction between the agents. Autonomous agents are a paradigm that can be instantiated in several concrete entities: robots, software programs, web services, nodes of a sensor network, humans, ....
The course is about conceptual, methodological, and algorithmic aspects of the cooperative interaction between multiple agents that will be exemplified by some experimental implementations.
1) Introduction to autonomous agents and multiagent systems.
2) Introduction to multiagent organizations.
3) Introduction to agent communication.
4) Basics of multiagent interactions: task allocation, coordination.
5) Multiagent planning: social laws, state-space and plan-space techniques, decision-theoretic multiagent planning.
6) Distributed constraint handling and optimization: distributed constraint satisfaction problems, distributed constraint optimization problems (complete and approximated algorithms).
7) Multiagent path finding: problems and algorithms.
8) Other issues: cooperative search, emerging behaviors.
9) Programming multiagent systems: agent programming languages, development frameworks and environments.
Students will learn the main models and algorithms to represent complex systems as collections of autonomous agents that cooperatively interact together. In particular, students will learn about several advanced types of interaction mechanisms between multiple cooperative autonomous agents and about some of the relevant algorithms available to implement these interaction mechanisms and their properties (range of application, optimality guarantees, ...).
Moreover, students, while developing a software project, will learn how to build agent-based systems that are based on the identification of individual decision-making entities (autonomous agents) and on the identification of the cooperative interaction mechanisms between them.