Toward a Simulation-Based Decision Support System to Predict and Cope with Mass Influx of Calls

The ever-increasing demand for emergency medical assistance contributes to the deterioration of prehospital care and affects the efficient use of resources initially designed to deal with urgent care. This phenomenon is associated with Mass Influx of Calls (MIC) in EMS, especially observed during emergencies. In an Emergency Medical Service (EMS) system that is unprepared or has limited real-time resource availability, such surges in demand can significantly delay response times and put patient lives at risk. In this chapter, we propose a Decision Support System (DSS) to rapidly anticipate MIC events and support the appropriate dimensioning of required resources. The system relies on a simulation-driven Online Reinforcement Learning (ORL) framework, triggered by a forecasting model based on Periodic Quantile Regression with Trend and eXogenous variables (PQRTX).