Efficiency and fairness in ambulance planning

In emergency situations where every second counts, the timely presence of an ambulance can be a matter of life or death. Thereto, we've developed several optimization models for the logistics of Emergency Medical Services (EMS).

In this talk we highlight two aspects:

1) Facility location models: efficiency versus fairness. Rather than simply maximizing the number of people served, we consider the distribution over the different areas where people live. To that end, we view ambulance optimization models from a social welfare perspective. We analyze existing ambulance planning models and show that they tend to maximize either the number of people served (utilitarian social welfare) or maximize the service to the person who is worst off (egalitarian social welfare).  We propose a third option: the so-called Bernoulli-Nash social welfare. We introduce a new facility location model that maximizes this Bernoulli-Nash social welfare, and show that it offers an attractive compromise between the utilitarian and egalitarian optimum.

2) Improving response times in operational planning: relocating idle vehicles to increase coverage. We introduce an efficient heuristic to solve this complex problem. The practical relevance of this heuristic was demonstrated by the implementation in a decision support tool used by the EMS region Flevoland.