On Car Sharing Usage Prediction with Open Socio-Demographic Data

Abstract

Free Floating Car Sharing (FFCS) services are a flexible alternative to car ownership. These transportation services show highly dynamic usage both over different hours of the day, and across different city areas. In this work, we study the problem of predicting FFCS demand patterns -- a problem of great importance to an adequate provisioning of the service. We tackle both the prediction of the demand i) over time and ii) over space. We rely on months of real FFCS rides in Vancouver, which constitute our ground truth. We enrich this data with detailed socio-demographic information obtained from large open-data repositories to predict usage patterns. Our aim is to offer a thorough comparison of several machine learning algorithms in terms of accuracy and easiness of training, and to assess the effectiveness of current state-of-art approaches to address the prediction problem. Our results show that it is possible to predict the future usage with relative errors down to 10%, and the spatial prediction can be estimated with relative errors of about 40%. Our study also uncovered the socio-demographic features that most strongly correlate with FFCS usage, providing interesting insights for providers opening service in new regions.

Forecasting the Carsharing Service Demand Using Uni and Multivariable Models

“…As we previously stated, most of the existing previous works focus on characterizing and modeling the characteristics and usage patterns of carsharing services [1, 5,2]. These works analyze Spatio-temporal service characteristics and clients' demands.…”

“…In the last few years, several studies have focused on the understanding of urban mobility and its applications [1, 2,3]. The growing attention to urban mobility occurs due to its relationship with -practically-all of the daily tasks, as people labor, education, security, and health.…”

“…This occurs since people can easily share goods (e.g., vehicles), mainly through online platforms [4]. In this way, carsharing is an urban mobility solution that is receiving increasing attention from the academic community and industry [2,1,5,6]. At a glance, the business models for carsharing are based on offering vehicles in a service area where users can book or use them at all times, without the inconveniences of owning a private vehicle [3].…”

“…The user behavior of a carsharing system has already been addressed. We are aware of several works that describe the usage and demand patterns of these types of services [1, 5,2]. In these recent works, researchers have shown results shown that the use of carsharing services follows a diuturnal pattern and, in one of the biggest challenges related to one-way and free-floating carsharing models is the fleet imbalance over the service area.…”

“…Note that most existing works in this area focus on a unique service model [2,12], using a specific forecasting technique. In this work, on the other hand, we use several time-series prediction techniques and their variations.…”