Information Technology Reference
In-Depth Information
Agent-Based Container Terminal Optimisation
Michael Winikoff
1
, Hanno-Felix Wagner
2
, Thomas Young
3
, Stephen Cranefield
1
,
Roger Jarquin
4
, Guannan Li
1
, Brent Martin
3
, and Rainer Unland
2
1
University of Otago, Dunedin, New Zealand
2
Universitat Duisburg-Essen, Essen, Germany
3
University of Canterbury, Christchurch, New Zealand
4
Jade Software Corporation, Christchurch, New Zealand
Abstract.
Container terminals play a critical role in international shipping and
are under pressure to cope with increasing container traffic. The problem of man-
aging container terminals effectively has a number of characteristics which make
agents a suitable technology to consider applying. Container terminals involve
the operation of distributed entities (e.g. quay cranes, straddle carriers) which
coordinate to achieve competing goals in a dynamic environment. This paper de-
scribes a joint industry-university project which has explored the applicability of
agent technology to the domain of container terminal management. We describe
a simulation platform of a container terminal based on the JADE agent frame-
work, along with two optimisations that have been developed and integrated with
the simulator: allocating container moves to machines through negotiation, and
allocating containers to yard locations through an evolutionary algorithm.
1
Introduction
Container terminals play a crucial role in the process of shipping containerised goods.
Due to their critical role, and the growth in the amount of container traffic, container
terminals are under pressure to increase their operating capacity and efficiency.
The goal of the work described in this paper was to investigate how agent-based
solutions could be used to improve the efficiency of container terminals. A number of
characteristics of container terminals make them a natural candidate for agent-based
solutions. Firstly, they can be naturally described as a system of interacting entities
(e.g. cranes) which are distributed and autonomous, and which interact to solve a prob-
lem (e.g. loading and unloading ships) in an efficient way. Secondly, the environment is
dynamic: the situation is subject to change, and things can (and do) go wrong. Taken to-
gether, these three characteristics make it natural to investigate agent-based techniques
for container terminal management and optimisation.
This paper reports on a joint industry-university project. The industry partner was
Jade Software Corporation, whose portfolio of products includes Jade Master Terminal
(JMT), a comprehensive container terminal management solution. The project included
visits to a local container terminal port in order to obtain a detailed understanding of the
problem and its associated complexities. Additionally, real (but anonymised) data from
the port was used for evaluation purposes. This data included machine movements and
container arrival and departure information.
