This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
Where should ambulances be placed to minimize the time required to reach calls for help? Emergency Medical Services (EMS) providers increasingly rely on sophisticated operations research (OR) models for such ambulance-deployment questions. These models depend on reliable forecasts of call-arrival rates over different parts of a city and at different times. Unfortunately, current forecasting methods are not very effective, and result in poor inputs to the OR models. This can lead to poor deployment decisions and thereby inflated response times for calls. The first part of this research will investigate advanced statistical methods for better estimating these arrival rates. The OR models also rely on accurate representations of travel times on road networks. Travel-time estimation is also in its infancy, and still does not effectively use the plethora of GPS data that ambulances accumulate. The second part of this research will investigate advanced statistical methods for better estimating travel times on road networks.
The key issue is that there is a lack of guiding theory to help software developers and EMS providers determine how to best use their available data to make quality decisions. If successful, this research will improve this lamentable situation, resulting in more effective representations of arrival rates and travel times that are the key determinant in ambulance-deployment decisions. The outcome for society is that ambulance deployment decisions will be more effective, leading to reduced response times for calls, with the associated medical benefits. In addition, we will advance the state of the art in statistical methods for complex engineering problems, and raise the profile of advanced statistical methodology in the operations research arena, as well as the profile of operations research problems within the statistical community.
