An optimised hybrid method for ambulance dispatch based on complex networks and Artificial Intelligence

Nowadays, monitoring people’s health and helping those in need of medical attention is one of the most important responsibilities of governments. Emergency Medical Services (EMS) are therefore setup to provide timely care to victims of sudden and life-threatening injuries or emergencies in order to prevent needless mortality or long-term morbidity. Providing rapid EMS with minimum response time will help improve survival rates. However, factors such as limited ambulance resources, large coverage area and high call-rates have caused delays in EMS, which have lead to lower survival rates. In this research the problem of ambulance dispatching with regards to lowering the response time is discussed. Response time is the most important factor in evaluating the performance of various EMS, which is directly proportional to mortality and survival rates. The most common method to reduce response time in ambulance dispatching is to choose the nearest available ambulance unit, i.e. the nearest neighbor (NN). Other methods use call-priority, first-in first-out (FIFO), preparedness and hybrid algorithms. One of the latest methods used in ambulance dispatching is based on complex networks. This method will apply a higher priority to the call that is more centrally located with respect to other calls. A hybrid algorithm has been proposed in this research which exceed the previous methods in terms of response time reduction.The proposed hybrid approach is based on complex network analysis and a search algorithm, which relies on the important parameters of dispatch, yet presents fewer constraints and overcomes some of the difficulties of the previous methods. In addition, prioritizing emergency calls is also considered. To evaluate these proposed algorithms, a combination of various parameters and operating environments has been simulated. The results of the simulations show improvements in response time reduction compared to previous methods.