Can user data improve Bike Sharing Systems demand forecasting?

Nowadays, the reduction of carbon emissions requires the adoption of sustainable transport means such as Bike Sharing Systems. Many cities have already implemented these systems based on docks but new problems have arisen from their use, such as the lack or surplus of bicycles in each of the docks. To solve these problems predictive models have been proposed for forecasting the demand at each station. In this work, it is proposed to exploit the usage patterns of the users registered in these systems to build models able to forecast future demand. A case study is presented in the city of Salamanca where the methods based on dock information are evaluated against the proposed approach. The results show that the proposed method can be applied to this kind of problem shortly. Preliminary results showed that they can still be improved to obtain a better solution for the predictions of the Bike Sharing Systems.