Agent Based Dynamic Service Synthesis in Large-Scale Open Environments: Experiences from the Agentcities Testbed Steven Willmott (Univ Politècnica de Catalunya *) David Bonnefoy (Motorola) Ion Constantinescu (Ecole Polytechnique Federale de Lausanne) Simon Thompson (BT Exact) Patricia Charlton (Motorola) Jonathan Dale (Fujitsu) Tianning Zhang (Agentscape) * contact author: Universitat Politècnica de Catalunya Modul C5, 211b, Campus Nord, UPC Barcelona, Spain, E-08034 steve@lsi.upc.es Abstract The notion of autonomous agents populating large- scale open environments, such as the public Internet, that are able to dynamically discover one another, interact and synthesise new software applications or results has become one of the key technology visions of the past few years. This Agentcities testbed represents one of the largest attempts to date to prototype such a vision: deploying current generations of agent and Semantic Web technologies to create a global test bed for dynamic service composition involving more than 100 participating organisations. The paper presents an overview of this initiative. 1 Context The past 2-3 years have seen unprecedented growth in a wide range of “network technologies” – from Peer-to-Peer systems, computational Grids, Web Services, e-Business technologies and Semantic Web systems. Unfortunately however, while such technologies are bringing us closer to common visions of future information systems visions ([BernersLee0] and [DeRoure02] for example) we are still faced with a wide range of challenging problems: - How do these technologies fit together to underpin a workable networked environment? Which choices are critical? Which are peripheral? Which tools should we be building and/or using? - Once technology is actually deployed how will the environment behave? Will dynamic service synthesis really be possible? How do we go from lab demonstrators to real systems? - How much practical progress can be made given the fundamentally intractable problems inherent in the endeavour (automated reasoning, machine processable semantics and so forth)? 2 Agentcities To help answer some of these questions the Agentcities testbed activity provides a live 24/7 experimental test environment which has been in use since late 2001. Based on the FIPA Agent standard, the testbed: - Supports a range of deployment projects (see Section 3) and over 150 deployed platforms worldwide in summer 2003. - Organisations involved in the project are able to deploy FIPA Agents on their own local machines and have them discover/interact with agents deployed by others elsewhere in the network. - A set of technology frameworks provide what may be one of the first relatively comprehensive architectural specifications for deployed open agent systems which has been widely deployed and used – providing a useful technical resource for future efforts. - One of the supporting European projects (Agentcities.RTD) also carried out a large- scale service composition demonstration in mid 2003 – involving nearly 200 agents interacting via 25+ service types and hosted on servers around the world. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. AAMAS'04, July 19-23, 2004, New York, New York, USA. Copyright 2004 ACM 1-58113-864-4/04/0007...$5.00