An agent-based architecture for urban transportation systems

Today, transport networks, because of increased density in urban population centers, are affected by more and more congested traffic which produces economic, social and political problems quantifiable in terms of fuel economy, pollution and increased level of stress due to bad service quality. Telematic infrastructures combined with automotive technology, by improving the operational efficiency of urban transport systems, can reduce the environmental and economic impacts. In this paper a Multi-Agent System (MAS) is proposed in order to manage a modern bus fleet equipped with sensors and communication devices. The multi-agent approach adapts well to describe complex systems characterized by geographical and functional distribution like transport networks where many subsystems with different dynamics and autonomy interact and for which real time control, high adaptivity and robustness are essential requirements. The proposed MAS represents a suitable decision support system for problem identification, diagnosis, action planning and prediction and can be used to plan, implement, and manage bus fleet operations. Moreover, the implemented MAS, by means of sensors and communication devices along the roads and on the bus fleet, can help to reduce traffic, gas emissions and to improve fuel economy, assuring at the same time a better service quality.