Prediction of transportation on granular level using survey and big data

The estimation of transportation demand and road use via survey data often lacks in timeliness and flexibility, in addition to high survey costs. Mobile Network Operator (MNO) data can provide timely mobility pattern estimates with a high spatial resolution. However, due to the different customer bases of the providers, estimates that are only extrapolated from MNO data suffer from different biases. The paper provides methods of jointly using survey and MNO data to deliver improved estimates for transportation demand and road use while utilizing the advantages of both data sources. The methods are demonstrated in a microsimulation, as a feasibility study with German data.