MAO-Multi-Agent Organisation (Lorentz Center)

Start Date: 2011-12-19

End Date: 2011-12-23

Location: Leiden

Country: The Netherlands


Multi-Agent Systems is a computational paradigm for modelling intelligent distributed systems. Open multi-agent systems are special kinds of multi-agent systems where individual agents may join and leave the system at run-time. Current trends in computer science support the importance of open multi-agent systems by addressing new challenges where typically complex software systems with large numbers of autonomous heterogeneous components are required. Industrial applications for open multi-agent systems are, for example, supply chain management, manufacturing control, and business process management.
A major challenge in the design and development of open multi-agent systems is the regulation and coordination of the behaviours of individual autonomous agents in order to guarantee the achievement of the overall objectives of open multi-agent systems. Recognized solutions for this challenge are the use of the so-called virtual normative organizations or electronic institutions. These solutions advocate the use of social and organizational concepts such as roles, interaction patterns, power and delegation structures, norms, sanctions, joint plans, and task allocations. For example, norms can be used to specify the standards of behaviour that agents ought to follow in order to achieve the overall objectives of open multi-agent systems. As individual agents are assumed to be autonomous, open multi-agent systems need to be organised in such a way that the specified standards of behaviour (norms) are actually followed by the agents. This can be realized by imposing the norms through sanctions and regimentation.
The goal of this workshop is to promote the discussion and exchange of ideas concerning computational models and programming frameworks for the design and development of normative multi-agent system organizations. In particular, we aim at bringing together researchers from different areas related to multi-agent systems, including software engineering, logic, and social sciences to discuss their computational theories, models, tools, and techniques that can be utilized in the development of normative multi-agent organisations. We also hope to set up collaborations between existing research groups that currently working on programming languages for normative multi-agent organisations in order to join their forces in the design and development of a powerful and unified normative programming language for multi-agent organisation. The collaboration can be set up by starting to define joint national or international research projects.

more information