Moving in time and space - Location intelligence for carsharing decision support

In this paper we develop a spatial decision support system that assists free-floating carsharing providers in countering imbalances between vehicle supply and customer demand in existing business areas and reduces the risk of imbalance when expanding the carsharing business to a new city. For this purpose, we analyze rental data of a major carsharing provider in the city of Amsterdam in combination with points of interest (POIs). The spatio-temporal demand variations are used to develop pricing zones for existing business areas. We then apply the influence of POIs derived from carsharing usage in Amsterdam in order to predict carsharing demand in the city of Berlin. The results indicate that predicted and actual usage patterns are very similar. Hence, our approach can be used to define new business areas when expanding to new cities to include high demand areas and exclude low demand areas, thereby reducing the risk of supply-demand imbalance.