Optimal placement of electric scooter parkings : A Mixed-Integer Programming Approach to Align Facility Placement with User Demand

The usage of electric scooters is a rising trend in many cities, serving as a flexible alternative or complement to public transport. The rapid pace of development highlights the need to optimize the availability of these services. To ensure or- derly and unobstructed public streets, these vehicles would optimally be placed in assigned parking spaces around the city. This report aims to minimize the number of parking spaces required to meet user demand, while also reducing walking distances. This is accomplished through the use of mathematical op- timization models incorporating integer-valued variables, commonly referred to as Mixed Integer Programming (MIP). In this project, the set covering problem together with the facility location problem will be considered. The aim is to in- vestigate how these can be used to decide how the placement should look like in Stockholm, Sweden, with demand based on population distribution. The results indicate that between 85 − 414 parking spaces are needed for the set covering model, while 863 − 1039 spaces are needed for the facility location model. It is concluded that the most realistic model is the latter. The findings also sug- gest that the model can be refined by improving the modeling of demand and considering additional factors.