Investigation of Shared-Bike Demand Using Data Analytics

Sustainable development commitments are of concern to the city's decision-makers as well as significantly impacting the existing transportation systems. The concept of the smart city and precisely the component of smart mobility centered on the user and soft transport comes to support this approach of transformation which aims at the low carbon city. Similarly, reducing carbon emissions is one of the main objectives of a smart city, thereby comes the focus on enhancing eco-friendly and active transportation means, for instance, the shared-bike system. The mode benefits from the technology implemented within the smart city concept. Seoul Government has implemented a shared-bike program “Ttareungyi” in 2015, within the big vision of “low carbon green growth”. However, the program struggles to achieve the targeted demand. Therefore, this study is using data analytics to help enlighten decision-makers about the shared-bike system and provide insights for future development. The research was conducted to investigate the influence of the built environment, including slope, land use mix, and centrality parameters, the influence of transport infrastructure, including bike and transit infrastructure, and the influence of the socio-economic characteristics, including population, retail number, car ownership, and job offers on bike demand. And to predict bike demand based on the mentioned variables using the ridge regression method. Results revealed that dock number, population density, and car ownership have a significant positive impact on biking demand, while slope has a significant negative impact. In contradiction to the research hypothesis, land use mix revealed a weak impact on biking demand using random forest, and a negative influence using ridge regression.