23.01.2014 Views

7 - Indira Gandhi Centre for Atomic Research

7 - Indira Gandhi Centre for Atomic Research

7 - Indira Gandhi Centre for Atomic Research

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

suit open applications, where a number of possibly a-priori unknown and autonomous<br />

entities have to cooperate. In this case, focusing on dependencies between the components,<br />

as a control-driven model would do, would somehow clash with the autonomy of the<br />

components and the dynamics of the open environment. Focusing on data preserve<br />

autonomy and dynamics of autonomous components, which are usually designed to<br />

acquire in<strong>for</strong>mation rather than control, as in the case of software agents.<br />

4. Conclusion<br />

Multi-agent concepts and methodologies are finding increasing applications in controlling<br />

complex, unpredictable systems in real time. Agent-like approaches routinely accept<br />

uncertainty and distribution, leading to control schemes where decision-making and<br />

responsibility are distributed much more widely than in conventional engineering practice.<br />

The area of intelligent manufacturing systems aims to use multi-agent systems both as a<br />

modeling tool and a control software system. The challenge <strong>for</strong> the researchers is to<br />

identify and focus on the general issues of coordination and control, which are important<br />

<strong>for</strong> such applications, and to use the applications to stimulate and test advances in research<br />

on:<br />

• How and where to represent in<strong>for</strong>mation about the systems and the agent groups that<br />

operate on them<br />

• Mechanisms <strong>for</strong> delivery of adequate responses to time-critical demands<br />

• Coordination methods that are computationally economical enough <strong>for</strong> real-time use<br />

• Decentralized management of limited common resources<br />

• Adaptation of results from machine learning to collective learning in multi-agent<br />

systems<br />

• New schemes of decentralized control to take account of the “real time” dimension<br />

than in conventional engineering practice<br />

References<br />

1. Marilyn, Deegan, & Simon, Tanner. (2002). Digital futures: strategies <strong>for</strong> the in<strong>for</strong>mation<br />

age. London: Library Association.<br />

2. David V. Pynadath and Milind Tambe. (2002). Team Coordination among Distributed<br />

Agents:Analyzing Key Teamwork Theories and Models. Chicago,American Association <strong>for</strong><br />

Artificial Intelligence.<br />

3. Michael Wooldridge. (1999). Multi-Agent systems: A modern approach to distributed<br />

artificial intelligence. Massachusetts,MIT Press.<br />

4. Rosaria Conte. (2002). Agent-based modeling <strong>for</strong> understanding social intelligence. PNAS,<br />

May 14.<br />

5. T. G. Dietterich. (1998) Machine-learning research:Four current directions. The AI<br />

Magazine, 18(4): 97 Œ136).<br />

6. D. Frey, T. Stockheim, P.O. Woelk and R. Zimmermann. (2003). Integrated multi-agentbased<br />

supply chain management. In: Proceedings of the 12th IEEE International<br />

Workshops on Enabling Technologies: Infrastructure <strong>for</strong> Collaborative Enterprises<br />

(WETICE-2003) IEEE, p.24.<br />

7. Y. Guo, J. Muller, and B. Bauer. (2004). A multi-agent approach <strong>for</strong> logistics per<strong>for</strong>mance<br />

prediction using historical and context in<strong>for</strong>mation. In: Proceedings of The Third<br />

International Joint Conference on Autonomous Agents and Multi-agent Systems<br />

(AAMAS'04), New York, IEEE Press.<br />

214

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!