Modelling an agent-based commercial vehicle transport system : a supply chain perspective

Abstract

Many state-of-practice commercial vehicle transport models are not representative of actual road transport movements, since they do not integrate supply chain elements. The objective of this research is to model stakeholders in a supply chain as agents in an agent-based commercial vehicle transport model. Furthermore, the objective is to model these agents’ decisions and interactions and to ensure that the model is sensitive to changes in the supply chain. To achieve this, various steps are followed. The literature on commercial vehicle modelling is reviewed and a distinction is made between three perspectives of commercial vehicle transport models: aggregate models; disaggregate, agent- and tour-based models; and behaviour-based models. A base case agent-based commercial vehicle model, that consists of both intra- and inter-provincial commercial vehicles, is developed using a complex network and GPS records. Utilising complex network metrics, supply chain stakeholders are identified and the most important nodes in the network are extracted. One of these important nodes, an organisation in the Fast Moving Consumer Goods (FMCG) industry, provides a dataset consisting of the details of distribution data over a 10-month period. This dataset is used in a case study to show how to model stakeholders in a supply chain. More specifically, the Carrier agent is introduced and the Carrier-Receiver interaction is modelled. Demand is generated from the dataset and the Carrier’s reaction to the demand is shown through its tour planning. The effect of different levels of traffic congestion as well as the order policy of customers on the Carrier’s tour planning is evaluated by showing the changes in distance travelled, tonne-kilometers moved, costs incurred, and travel time for different scenarios. The research is of value to both organisations that need to do fleet management and government who is responsible for infrastructure maintenance and development. Organisations can utilise these models to do fleet composition analyses and evaluate the impact of changes to their logistics decision making or the effect of government interventions on their operations. Government can benefit from these models by analysing the effect of infrastructure decision-making on tonne-kilometers moved and the impact on expected travel times.