In this thesis, we investigate whether maximizing the coverage of taxis can be beneficial when the goal is to minimize the waiting times of the clients. When dispatching taxis, often only current requests are taken into account and not future ones. We examined how beneficial it c
...
In this thesis, we investigate whether maximizing the coverage of taxis can be beneficial when the goal is to minimize the waiting times of the clients. When dispatching taxis, often only current requests are taken into account and not future ones. We examined how beneficial it can be to take coverage into account. For taxi companies it is important to keep their customers satisfied, by serving them as quickly as possible. Often companies assign the taxis to the nearest requests. We investigate whether there are better dispatch methods. We developed three dispatch policies which use ILP models, and discussed what the best option is. The first policy does not consider the coverage and minimizes the waiting times of the clients among the requests that are taking place at that moment. The second one aims to maintain a good coverage and short waiting times when assigning taxis to requests. The third one also aims to maintain a good coverage and short waiting times when assigning taxis to requests and also relocates taxis to gain a better covered area. Those policies can be a contribution for taxi companies because they help to serve clients more quickly. To determine what the best way is of dispatching taxis, we run a simulation on the different policies on a real-data map and compared the results. The literature over taxi dispatch mostly conducts research on a small area like a city or a village. In this thesis, taxis dispatch takes place on a larger area where taxis commute between cities.