RampScope

Short-term rentals of shared mobility devices (SMDs) including bikes, e-bikes, and e-scooters are gaining significant popularity across different countries. These services equip their SMDs with GPS receivers which allows the riders the flexibility to park their SMDs anywhere, and the next user simply finds the nearest parked SMDs. However, GPS accuracy decreases significantly in urban areas and causes real-world problems (e.g., users and chargers not being able to locate the SMDs). To overcome this problem, we propose Ramp-Scope that utilizes physical characteristics of ramps on the sidewalks - which are prevalent in urban areas - to correct for GPS error. As the user rides over a sidewalk ramp, the SMD equipped with a gyroscope captures the motion signal to uniquely identify the ramp and localizes the SMD to the nearest driven ramp. As a proof-of-concept, we present a preliminary evaluation of RampScope with real-world experiments by driving three different SMD types over 800 m to demonstrate an average ramp prediction accuracy of 98.1%.