Preprocessing of GIS data for electric vehicle charging stations analysis and evaluation of the predictors significance

When a government or a municipality intends to incentivize electromobility a chicken-egg problem arises: shall citizens first buy electric vehicles or should the charging stations be installed first? Often, a charging infrastructure is deployed, even if the number of electric vehicles is small, in the hope to stimulate a larger interest in electric mobility. In such cases, charging station operators in collaboration with municipalities have to make appropriate decisions, including finding the placement of charging stations that meets the demands of electric vehicles (EVs) while expending the resources efficiently. In this paper, a data set coming from a large network of charging stations, located in one of the worlds electromobility leading countries the Netherlands, is analysed. Dataset comprises over one million charging transactions, more than 1700 charging stations and spans over four years. First, methodological tools to describe the urban context by data are introduced. Publicly available GIS data are collected and exploited to assess the geographical locations of charging stations. Data preprocessing procedures are described and their limitations are discussed. Finally, using correlation analysis the ability of the gathered GIS predictors to explain the utilization characteristics of EV charging stations is evaluated. (C) 2019 The Authors. Published by Elsevier B.V.