U-Park: A User-Centric Smart Parking Recommendation System for Electric Shared Micromobility Services

Abstract

At present, electric shared micromobility services (ESMS) have become an important part of the mobility as a service (MaaS) paradigm for sustainable transportation systems. However, current ESMS suffer from critical design issues such as a lack of integration, transparency and user-centric approaches, resulting in high operational costs and poor service quality. A key operational challenge in ESMS is related to parking, particularly how to ensure a shared vehicle has a parking space when the user is approaching the destination. For instance, our recent study illustrated that up to 12.9% of shared E-bike users in Dublin, Ireland cannot park their E-bikes properly due to a lack of planning and user-centric guidance [1]. To address this challenge, in this paper we propose U-Park, a user-centric smart parking recommendation system for ESMS aiming at making personalised recommendations to ESMS users by considering a given user's historical mobility data, current trip trajectory and parking space availability. We propose the system architecture, implement it, and evaluate its performance based on a realworld dataset from an Irish-based shared E-bike company, Moby bikes. Our results illustrate that U-Park can effectively predict a user's destination in a shared E-bike system with around 97.33% accuracy without direct user inputs, and such results can be subsequently used to recommend the best parking station for the user depending on the availability of predicted parking spaces. Finally, a blockchain-empowered module is blended to manage various transactions due to its advantage in safety and transparency in this process, including prepays and parking fines.