Forecasting the Usage of Bike-Sharing Systems through Machine Learning Techniques to Foster Sustainable Urban Mobility

Aug 5, 2026

Upload by Oliver

Bike-sharing systems can support sustainable urban mobility, yet planning and operation face persistent challenges due to spatial and temporal imbalance in bike availability. Such imbalance forces operators to perform costly repositioning moves to rebalance scarce and accumulating stations, and prediction errors propagate directly into suboptimal rebalancing decisions. The paper applies machine-learning regression models using historical usage and meteorological inputs, comparing random forest, gradient boosting, and artificial neural networks on a New York City case study.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Introduction and Background

Bike-sharing and mobility-as-a-service

Operational challenges and repositioning (rebalancing)

Need for accurate usage forecasting

sustainability Article Forecasting the Usage of Bike-Sharing Systems through Machine Learning Techniques to Foster Sustainable Urban Mobility Jaume Torres, Enrique Jiménez-Meroño and Francesc Soriguera * Citation: Torres, J.; Jiménez-Meroño, E.; Soriguera, F. Forecasting the Usage of Bike-Sharing Systems through Machine Learning Techniques to Foster Sustainable Urban Mobility. Sustainability 2024, 16, 6910. [https://](https://)[ ](https://)[doi.org/10.3390/su16166910](doi.org/10.3390/su16166910) Academic Editors: Giovanni Leonardi, Yusheng Ci, Lina Wu and Ming Wei Received: 14 May 2024 Revised: 29 July 2024 Accepted: 8 August 2024 Published: 12 August 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . BIT—Barcelona Innovative Transportation, Universitat Politècnica de Catalunya-BarcelonaTech, 08034 Barcelona, Spain * Correspondence: [francesc.soriguera@upc.edu](francesc.soriguera@upc.edu) Abstract: Bike-sharing systems can definitely contribute to the achievement of sustainable urban mobility. In spite of this potential, their planning and operation are not free of difficulties. The main operational problem of bike-sharing systems is the unbalanced distribution of bicycles over the service region, resulting in zones where bicycles are scarce and zones where bicycles accumulate. In order to provide an acceptable level of service, the operator needs to carry out repositioning movements, which are costly. Bike-sharing repositioning optimization solutions have been developed that rely on the estimation of the expected number of requests and returns at each location. Errors in this prediction are directly transferred to suboptimal repositioning solutions. For this reason, the development of methodologies able to accurately forecast bike-sharing usage is an issue of great concern. This paper deals with this problem using machine learning regression methods, which yield usage predictions from inputs such as historical usage and meteorological data. Three different machine learning regression techniques have been analyzed (i.e., random forest, gradient boosting, and artificial neural networks) and applied to a case study based on the New York City bike-sharing system. This paper describes the variables of the models and their calibration processes. Results are analyzed and compared in order to determine which one of the three techniques and under what conditions is the most adequate. Comparisons are not only made in terms of accuracy but also with respect to the applicability of the algorithms. Results indicate that, given the similar accuracy of all methods, the simpler calibration process of the random forest technique makes it advisable formost applications. Keywords: sustainable urban mobility; sustainable transportation; shared-mobility; bike-sharing; station-based; demand prediction; machine learning regression 1. Introduction and Background The concept of bike-sharing consists of providing a fleet of bicycles for users to use to make trips without needing to own them. This shifts the focus to a mobility-as-a-service model. From the user’s perspective, bike-sharing, as with other vehicle-sharing initiatives (e.g., car-sharing, motorbike-sharing, scooter-sharing), offers the advantages of low-cost on-demand transportation (i.e., flexibility, traveling when and where needed) in front of the traditional public transportation alternatives with predefined routes and schedules [1] . Politically and societally, bike-sharing systems represent a significant step towards sustainable urban mobility by balancing ecological, economic, and social interests [2] . Bicycling is therefore promoted for its b

FAQs

Why is forecasting bike-sharing usage important for sustainable operations?

Bike availability is often imbalanced across space and time, leading operators to reposition bikes to keep service levels acceptable. Accurate forecasts reduce prediction errors that would otherwise cause suboptimal rebalancing plans.

Which input data are used to predict bike-sharing usage in this study?

The models generate usage predictions from historical usage patterns and meteorological data, enabling demand estimation under varying conditions.

Which machine learning regression techniques are compared, and what conclusion is drawn?

The study analyzes random forest, gradient boosting, and artificial neural networks on New York City data. With similar accuracy across methods, the random forest approach is favored for most applications due to its simpler calibration process.

Aug 5, 2026

Upload by Oliver