8.1.6 Hypercubed Ambulances

A student project on ambulance allocation in Boston in 1972 resulted in a Master's thesis [HOEY 73] in which the student used an early version of the hypercube queueing model to help determine the required number and locations of ambulances in Boston. The city, at that time, was planning a shift from a police-operated ambulance service to one run by the Department of Health and Hospitals. While the recommendations in the study were expressed in concrete terms (e.g., "have X ambulances available during these hours and put them in these locations"5), the student felt frustrated by the lack of any immediate action on his plans.

As it turned out, the actual decision process was a complicated multiyear process involving diverse citizen's groups, private ambulance firms, the police department, the fire department, and the Department of Health and Hospitals. Later, a state law was passed requiring upgrading of the emergency medical skills of all ambulance attendants. However, throughout this long, drawn-out process, the student's thesis (which was published) was used as the analytical blueprint for "what should be done, if we can figure out who should do it." Thus, it contributed to the analytical side of a highly political decision-making process.

In the spring of 1978, fully 6 years after the original thesis was written, the Department of Health and Hospitals contracted with a local not-for-profit firm to implement the hypercube queueing model on its own computer system. The model is now being used to deploy Boston's 14 new ambulances and to plan for system evolution over the next 10 to 15 years [BRAN 78]. 5 Special curb-side ambulance locations were allowed, thereby eliminating any constraint to locate them in prespecified ambulance garages.