Location optimization for new car-sharing stations

The objective of this research study is two-fold – one, to analyze the performance of Auto Bleue service across all stations and estimate the key drivers of demand for the service, and secondly, to use these drivers to identify future station locations such that the overall system performance is maximized.

The objective of the EPFL study for Veolia’s Auto Bleue car-sharing service is two-fold – one, to analyze the performance of Auto Bleue service across all stations and estimate the key drivers of demand for the service, and secondly, to use these drivers to identify future station locations, such that the overall system performance is maximized.

Our mathematical model establishes that performance of Auto Bleue stations can largely be explained by variables such as share of high income / education population in the target locality, public transport ridership, population density and presence of mobility attractors such as hotels and commercial centers. Model also establishes that there is a clear negative influence, due to cannibalization, of other Auto Bleue stations in the vicinity of one and uses this idea to optimize the locations to balance the presence of stations in the high-potential center, which is quickly reaching saturation levels, versus low-potential outskirts that virtually still remains largely untapped.

Principal investigator Prof. Michel Bierlaire
Project managers Prem Kumar Viswanathan, Michaël Thémans
Sponsor Veolia Transport Auto Bleue
Period 2012
Laboratory TRANSP-OR
Collaboration TRACE