In the eld of multiagent systems, belonging to the more general area of Distributed Arti cial Intelligence (DAI), it is important to conceive a multiagent system as an unique machine, called agency. The conception of agency as an unitary machine, which derives from the work of Minsky [8], has the interesting role of focusing some of the critical problems involving the agency machine in general. The purpose of the paper is to illustrate a novel and recent approach to the design of the architecture of an agency, developed by the authors [5, 3, 4, 1] and called dynamic agency. After a brief introduction of the role of agency as an unitary machine and an illustration of the problem tackled by dynamic agencies, we present the architectural characteristic of a dynamic agency and the phases followed in its development. Finally, we shortly illustrate some applications involving dynamic agencies.
[1]
Francesco Amigoni,et al.
Dynamic Agencies and Multi-Robot Systems
,
1998,
DARS.
[2]
Francesco Amigoni,et al.
PROCESSING AND INTERACTION IN ROBOTICS
,
1999
.
[3]
Carlo Ghezzi,et al.
Analyzing Mobile Code Languages
,
1996,
Mobile Object Systems.
[4]
Francesco Amigoni,et al.
A theoretical framework for the conception of agency
,
1999
.
[5]
Michael Wooldridge,et al.
Agent technology: foundations, applications, and markets
,
1998
.
[6]
F. Amigoni,et al.
Learning by Exploration and Cooperation in Dynamic Agencies
,
1997
.
[7]
M. Minsky.
The Society of Mind
,
1986
.