Integrating urban last-mile package deliveries into an agent-based travel demand model

With the expected increase of e-commerce activity, we can expect the share of delivery vehicles in cities to rise as well. On the one hand, this puts great pressure on cities and surrounding areas as emissions rise and space becomes scarce. On the other hand, people are adjusting their travel behaviour such that the increase in e-commerce affects not only last-mile delivery but also private passenger traffic. This paper presents an integrated approach of modelling last-mile deliveries using an agent-based travel demand model. It is intended to account for reciprocal effects between online shopping behaviour and last-mile deliveries. The package orders are generated by agents in the study area and distributed among the package centres. For each package centre, the tour for each delivery agent is created. The presented model allows for the simultaneous simulation of private trips and last-mile deliveries and thus realistic delivery conditions: the model can detect e.g. if an agent or another household member is at home to receive their order. We have applied the model to the city of Karlsruhe, Germany, and describe first results of that simulation. ... mehrApplication of the model allows for a detailed analysis e.g. of delivery success rates both in terms of time and space. The presented modelling framework provides insight into effects of last-mile deliveries on a transportation system and can be availed to analyse policy measures or alternative delivery strategies.