Ontologies and agents for a bus fleet management system q M.V. Belmonte, J.L. Pe ´rez-de-la-Cruz * , F. Triguero Department of Languages and Computer Science, University of Ma ´ laga, Bulevar Louis Pasteur, N35, 29071 Ma ´ laga, Spain Abstract This paper reports on the design and implementation of a multi-agent decision support system for the bus fleet management domain. In particular, we show a complete description of the proposed multi-agent architecture and focus mainly on knowledge and software engineering features. The system has been developed in collaboration with the Ma ´laga urban bus company (EMT) and it is based on a faithful reproduction of the real operating conditions of three existing lines of this company. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Bus fleet management; Decision support; Multi-agent; Ontology 1. Introduction In recent years the progress of telematic infrastructure and technology has significantly increased the management possibilities of traffic control centers. In this sense, the use of new technologies such as Global Positioning Systems (GPS) in the case of public transport, and warning mes- sages via broadcast (RDS/TMC or cellular-phone-based services) or recommendations for drivers by means of VDS (Variable Direction Signs and text panels) in the case of private transport, provide real-time information to oper- ators in the control centers and to users of these kinds of transport respectively. As a consequence, complex infor- mation management tools need to be integrated into the traffic control centers, in order to benefit from these new infrastructures. The objective of these traffic management tools is to make use of this information infrastructure in order to interpret the different states of traffic flow, and to react appropriately in the face of disruptions. Obviously, this type of activity requires intelligent behavior of the system, and one way of achieving this is to endow the system with knowledge of the structure and the dynamics of the net- work. This knowledge is provided by the operators and is used to identify critical traffic situations, and to determine the most appropriate control actions. The application of such knowledge-based artificial intelligence systems to transport management has been an active area of research in recent years (Bielli, Ambrosino, & Marco, 1994), and these types of systems are usually known as Intelligent Transport Systems (ITS). The goal of these systems is to be able to reason about the traffic behavior in similar terms to an expert traffic oper- ator. The idea is not to replace traditional methods used with available traffic control technology (Hunt, Robertson, Bretherton, & Winton, 1981; Mauro & Di Taranto, 1989), but rather to complement them by extended capabilities and improved performance. ITS must provide operators with on-line support to cope with strategic management and intelligent supervision of traffic control. In general, traffic systems consist of many autonomous intelligent entities, which are distributed over a large area and interact with each other to achieve certain goals. These entities may be completely different: traffic lights, buses, trucks or even road users, but all of them are actively changing the traffic situation. In fact, these are some of the main characteristics assigned to ‘‘agents’’ (Wahle, Baz- zan, Klu ¨gl, & Schreckenberg, 2002). An agent, although 0957-4174/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2007.01.003 q Work supported by the Spanish Ministry of Science and Technology (MCyT) under grant TIC2000-1370-CC04. * Corresponding author. Tel.: +34 952 132801; fax: +34 952 131397. E-mail addresses: mavi@lcc.uma.es (M.V. Belmonte), perez@lcc. uma.es (J.L. Pe ´rez-de-la-Cruz), triguero@lcc.uma.es (F. Triguero). www.elsevier.com/locate/eswa Available online at www.sciencedirect.com Expert Systems with Applications 34 (2008) 1351–1365 Expert Systems with Applications