10th - 11th July, 2001

This will be the second meeting under AgentLink 2.

To participate in the ABSS sessions, you must submit the online registration application form which includes space for a statement of the contribution you intend to make.  The form does not ask for information previously given (apart from your name and institution, for obvious reasons).

The programme for the two days of the AgentLink meeting is available by clicking here.

As for all of the SIGs, participation is by invitation only.  Participants from AgentLink nodes can be funded in the usual parsimonious way (though only one participant from each node at each SIG).  If you want funding, that must be indicated on the online form.
 

The point of bringing these two branches of ABSS together is not entirely to develop new theory or methodology.  It is a part of a wider research programme to develop scalable multi agent systems with direct applications.  At the Amsterdam SIG meeting, a dominating subject of discussion was whether and how ABSS could provide guidance and insight into the achievement of multi agent systems for very large scale applications in complex environments.  A particular issue was whether the social simulation techniques developed to describe actual, large scale, complex social systems could provide the appropriate metaphors for large scale, complex software systems.  It was agreed at the Amsterdam meeting that one objective of the SIG under AgentLink II will be to identify and develop such synergies between the designing and engineering of agent based software systems on the one hand and, on the other hand, agent based social simulation.  The application of ABSS specifically to policy issues follows from the eager reception of ABSS by scientists concerned with policies in conditions of global climate change.  The current European project in this area is FIRMA which integrates social and physical models to analyse the impact of climate change on European water resource management.

Stimulated by the stress being placed on industrial collaboration and involvement in the SIGs under AgentLink 2 (but mainly because it is an appropriate departure), it was agreed in Amsterdam to set up a working group to identify and demonstrate the features of MAS that are scalable and where the properties of the software environment emerge from agent behaviour and interaction.  This work will be developed around a design problem described by André Meyer and Joost Reuzel from Philips Research in Eindhoven. The first session of the Prague meeting will be organised by André and Niik Wijngaards (Vrije Universiteit Amsterdam) to push forward the involvement of the SIG in the Philips design problem.

Although there was some discussion of policy issues relating to environmental issues and the impacts of human activity, no specific social issues were identified to provide the focus for a second working group.  The final session of the Prague meeting will consider the relevant issues.  An important gujiding principle in the establishment of a second working group will be to identify and, we hope, to exploit the synergies between the Philips application and social simulation applied to policy analysis.

A third session of the SIG meeting will focus on broader issues of European research in simulation and multi agent systems.  The SIG is required to produce a technological roadmap for ABSS -- a document identifying the current state of the field in Europe and suitable directions of development.  The current draft of the technological roadmap, inherited from AgentLink 1 and revised by Paul Davidson, is available here.  We anticipate that the issue of synergy between scalable MAS and the simulation modelling of very large scale social systems will be a central issue and will help to guide the development of the second working group on simulation for social policy.

There will be the usual funding for a limited number of participants, although unfunded participants will be welcome.  Only one person from each AgentLink node can be funded.