Information Technology Reference
In-Depth Information
Designing and Evaluating an Adaptive Trading Agent
for Supply Chain Management
Minghua He, Alex Rogers, Esther David, and Nicholas R. Jennings
School of Electronics and Computer Science, University of Southampton,
Southampton, SO17 1BJ, United Kingdom
{
mh, acr, ed, nrj
}
@ecs.soton.ac.uk
Abstract.
This paper describes the design and evaluation of SouthamptonSCM,
a finalist in the 2004 International Trading Agent Supply Chain Management
Competition (TAC SCM). In particular, we focus on the way in which our agent
sets its prices according to the prevailing market situation and its own inventory
level (because this adaptivity and flexibility are the key components of its suc-
cess). Specifically, we analyse our pricing model's performance both in the actual
competition and in controlled experiments. Through this evaluation, we show that
SouthamptonSCM performs well across a broad range of environments.
1
Introduction
Internet technologies have contributed significantly to e-commerce by increasing the
mutual visibility of consumers and suppliers, and by raising the possibility that some
of their trading processes may be automated. However, despite these advances, most
procurement activities within supply chains are still based on static long-term con-
tracts and relationships. Now, in many cases, such contracts are detrimental because
they fail to handle the dynamic nature of these environments, where new suppliers and
consumers may enter the market at anytime and where trading partners may fail to ful-
fill their commitments. To rectify this, we believe agent-based solutions are needed.
To date, however, the use of agents within e-commerce has generally focused on sim-
ple auctions [4]. Whereas, the supply chain domain typically requires handling a more
complex
setting
in
the
presence
of
much
greater
degrees
of
uncertainty
and
dynamism [6].
To this end, the International Trading Agents Competition for Supply Chain Man-
agement (
http://www.sics.se/tac
) (TAC SCM [1]) represents an ideal envi-
ronment in which to test the autonomous agents that we develop. Such multi-agent
research competitions present well-defined problems in which alternative solutions can
be tested, compared and evaluated. In the TAC SCM scenario, agents are competing
as computer manufacturers in a virtual business world to handle three basic subtasks:
acquiring components, managing manufacturing process, and selling assembled com-
puters to customers.
Against this background, we present our work in developing an adaptive agent that
was a finalist in the 2004 TAC SCM competition (6 out of 29 participants reached the
finals). The key contribution of this work is the techniques that we develop to enable the
