Estimating and forecasting urban freight origin–destination flows

This chapter reviews models used for the simulation of urban goods mobility. The focus is on models that provide origin-destination (O–D) matrices of goods vehicle flows. These are essential inputs for subsequent impact analysis. Tours, rather than direct trips, characterise urban goods mobility and make the modelling task difficult. Conventional classification includes trip-based, commodity-based and delivery-based models, according to the unit considered. Trip-based models are able to represent the different urban goods mobility segments, whereas commodity-based models and delivery-based models are able to represent only restocking of shops and of food-and-drink outlets and e-commerce flows to end-consumers. The latter two classes of models use, typically, a sequence of discrete choice random utility sub-models to represent the dependence of flows on both the economic and spatial characteristics of demand and the supply of transport infrastructure and services. Most recently, a shift towards integrated modelling of restocking, e-commerce deliveries and passenger shopping mobility has been observed.