MULTI-AGENT MODELING FOR CITY LOGISTICS POLICY ANALYSIS: POTENTIALS AND CHALLENGES

Heterogeneous stakeholders with diverse interests and distributed decision making process add complexity and unpredictability in the large city logistics domain. This complexity demands well designed approach for policy analysis which cannot be achieved by traditional modeling techniques due to their restrictive hypothesis. Multi-agent modeling for city logistics policy analysis focuses on understanding urban freight movement not based either on buying-selling and supply-demand of commodities or aggregate totals of demand, but on importance of under what structures these forces play out, and how to measure these effects in order to determine effective policy measures for city logistics. Multi-agent system provides potential to meticulously replicate the urban freight movement by mapping complexity of domain, time and discipline simultaneously. Describing the domain naturally & flexibly and capturing its emergent behavior are other benefits of multi-agent modeling. Along with these potentials, some of the important challenges associated with developing multi-agent system for city logistics policy analysis such as describing different stakeholder roles, restricting the scale of the model, designing environment with all affecting factors, developing interaction mechanism and defining scope for the model are discussed