Mobility Solutions for BMW Group: Classification of Car Usage Patterns

Problem Statement

Sensor data from vehicles are recorded, so that longitudinal information on car use is collected over a longer period of time. However, this data does not include information for trip purposes, which is essential for the description of usage patterns.


With CUMILE (Car Use Model Integrating Long Distance Events) the Institute for Transport Studies has developed a model that describes the car use of a representative car fleet in Germany over a period of one year. The model is based on the German Mobility Panel (travel survey data of one week, car use data) and allows the classification of car use patterns. The aim of the project is to bring the sensor data into a CUMILE-comparable form in order to use and evaluate the data for different tasks. The results will be used to gain insights for the development of passenger cars and innovative mobility solutions.


The task of the Institute for Transport Studies is to evaluate the CUMILE model with regard to relevant aspects for the identification of trip purposes. In addition to the duration and distance of the trips, the time of day and the standing time at the destination describe relevant characteristics.