In mobility-on-demand services, fleet sizes are usually dictated by rush-hour demand, leaving many vehicles idle during off-peak times. But what if those vehicles could take on other tasks instead of sitting unused?

Using agent-based modeling, we explored the impact of vehicles switching between passenger rides and food deliveries in Cambridge (MA, USA) and San Sebastian (Gipuzkoa, Spain).

The results show that multifunctional fleets can reduce the total number of vehicles needed—though the effectiveness depends on the city context. We also tested a strategic dispatching algorithm that helps cut wait times and reduce the total distance traveled by the vehicles.

By making better use of available resources, these approaches could help cities develop more efficient and sustainable mobility systems.