While the number of deliveries has been increasing rapidly, infrastructure such as parking and building configurations has changed less quickly, given limited space and funds. This may lead to an imbalance between supply and demand, preventing the current resources from meeting the future needs of urban freight activities.

This study aimed to discover the future delivery rates that would overflow the current delivery systems and find the optimal number of resources. To achieve this objective, we introduced a multi-objective, simulation-based optimization model to define the complex freight delivery cost relationships among delivery workers, building managers, and city planners, based on the real-world observations of the final 50 feet of urban freight activities at an office building in downtown Seattle, Washington, U.S.A.

Our discrete-event simulation model with increasing delivery arrival rates showed an inverse relationship in costs between delivery workers and building managers, while the cost of city planners decreased up to ten deliveries/h and then increased until 18 deliveries/h, at which point costs increased for all three parties and overflew the current building and parking resources. The optimal numbers of resources that would minimize the costs for all three parties were then explored by a non-dominated sorting genetic algorithm (NSGA-2) and a multi-objective, evolutionary algorithm based on decomposition (MOEA/D).

Our study sheds new light on a data-driven approach for determining the best combination of resources that would help the three entities work as a team to better prepare for the future demand for urban goods deliveries.