User Behaviour Analysis of Public Charging Infrastructure for Electric Vehicles

Electromobility has been chosen by decision makers in industry and politics as the path to decarbonise transport. For owners of such electric vehicles, a key requirement is an adequate charging infrastructure. To best service their customers, charge point operators need an understanding of their customer groups’ requirements. This paper analyses a large dataset of customer information and clusters them into the three groups “resident”, “commuter”, and “opportunity user” based on their behaviour patterns. Clustering is done in an unsupervised manner and with the number of clusters being determined based on the data. The boundaries between customer groups are blurry as users may show multiple characteristics, but clustering results are stable across multiple iterations and clustering algorithms. Most customers use no more than three charging stations that are within 10 km to each other. Residents consume the highest amount of energy while opportunity users consume the highest amount of energy per connected time.