An equity-based static relocation and battery swapping optimisation model for free-floating micromobility sharing systems

Micromobility completely changes the way we move within cities. It allows to cover smallmedium distances and promote an environmentally friendly way of moving, reducing pollution and traffic. However, in micromobility sharing systems, despite these advantages, the location of vehicles can generate an imbalance between supply and demand by making some areas better served than others and an unequal level of service among system users. To solve these issues, that may jeopardise the implementation of a sharing system, an appropriate relocation planning can be proposed. To the best of our knowledge in literature, for free-floating micromobility electric vehicle sharing systems, no studies deal with relocation considering battery swapping operations and the evaluation of the system equity during the day. For this reason, we propose an integer linear programming optimisation model that aims at providing minimum cost night balancing tours, with related pick-up, drop-off and battery swapping operations, that also optimise demand satisfaction and equity as a measure of a fair distribution of electric vehicles among the zones of the system area. In this study, to evaluate equity, we propose a modified Gini index which considers the available number of electric vehicles and their state of charge. The model pursues equity by defining, for each zone, a target fair number of electric vehicles to be available at the end of the rebalancing tours. These parameters are inferred by means of data-driven microsimulation and data analytics techniques. The proposed data-driven microsimulator, that simulates the functioning of a free-floating micromobility system, is also used to assess the results of the optimisation model. The methodology has been applied to a real case study considering an escooter-sharing system: the optimisation model, fed by the parameters learned from data-driven simulation, was able to compute efficient rebalancing tours with different trade-offs between demand satisfaction and equity. Simulation confirmed that optimal balancing tours improve equity while preserving operational efficiency and demand satisfaction, although intermediate (dynamic) rebalancing may be beneficial from an equity standpoint after the midday peak hour. Overall, the proposed methodology provides an interesting tool to support equity-aware electric micromobilty systems' operations.