Optimizing Shared Micro-mobility Services: Edge-Enabled Rebalance for Dock-based System
Conférence : Communications avec actes dans un congrès international
The user experience is an important aspect of micromobility fleet operations, and placing micro-vehicles in a suitable
and optimized manner is a key element to enhancing user
service. This paper aims to establish an effective methodology for
optimizing shared micro-mobility rebalance operations through
spatio-temporal prediction of user demand in dock-based sys tems. It is based on forecasting the occupancy levels of each
station to avoid completely empty and jammed stations in the
future, ensuring service availability throughout the day. Since the
processed data is local, we employ edge computing, yielding a
scalable solution that minimizes latency and enhances reliability,
making it suitable for urban environments with fluctuating
conditions. The results demonstrate significant improvements in
service availability, validating the efficiency of our edge-adapted
prediction model for dock-based micro-mobility fleets.
Index Terms—Shared micro-mobility, machine learning, im balance problem, edge computing system.