Mercoledi' 1 aprile alle ore 14 presso la sede di Milano dell'IMATI, via Bassini 15, Sala Pentagonale al secondo piano, Valérie Bélanger dell'Universita' HEC Montréal terra' un seminario dal titolo
Analysis and development of real-time ambulance fleet management strategies
di cui segue il sunto.
Tutti gli interessati sono invitati a partecipare.
____________________________
Sunto:
Emergency medical services (EMS) evolve in highly random and dynamic contexts. Indeed, the arrival of emergency calls is uncertain and varies throughout the day, forcing decision makers to adapt their decisions in response to observed changes. Two types of decisions are generally considered in real-time: dispatching and relocation. Dispatching decisions consider the selection of which ambulance to send to an emergency call. Relocation decisions consist in relocating available ambulances throughout the day in response to changes in the system state. Given the situation faced by EMS, the analysis and development of real-time management strategies to ensure a good service to the population is clearly of interest.
In a first study, several ambulance relocation strategies are compared, and extensive simulation experiments are reported allowing to analyze the performance of these strategies when they face different level of system workload. Results of this study confirm the relevance of simulation methods to support decision-making in EMS stochastic environment. More importantly, they provide researchers and decisions makers with valuable insights on the use of relocation strategies, including the tradeoff between service level and personnel well-being. In a subsequent study, an innovative real-time ambulance fleet management strategy is proposed. New mathematical models and solution approaches are then developed to deal with both relocation and pre-assignment decisions. Preliminary results show that considering explicitly pre-assignment decisions, which are directly related to dispatching decisions, impact location decisions and can help better predict system performances.
Cordiali saluti.