You are here: Home / Research / projects / ongoing / MobiSense






Instantaneous, reliable and area-covering information on road surface and infrastructure state is of great interest not only to individuals, but also to infrastructure owners, as it would reduce pre-emptive maintenance costs and avoid damage claims. Current monitoring state-of-the-art involves expensive equipment and manpower, and does not provide timely data.

MobiSense addresses this need, by providing a robust and cost-effective set of on-board hardware and software modules for vehicle-based collection of geospatial information on road infrastructure and the environment. MobiSense combines data from different sources (dedicated low-cost sensor box with microphone, tri-axial accelerometer, GPS and CAN-bus access, and smartphones) in a secure and privacy-respecting way, and applies advanced machine learning algorithms to transform vehicle-bound data into location-bound data. Good spatio-temporal coverage of cities can already be obtained when only a very small number of vehicles are equipped with the hardware, resulting in a system that has only a fraction of the cost of state-of-the-art techniques and that provides more up-to-date information.


For more information on the project, you can consult:

Time span: 2017-10-01 - 2019-09-30
Assigned by: imec - ICON
Promoter: Peggy VALCKE
Staff: Stijn STORMS
Partner(s): imec - WAVES - UGent => Coordinator
imec - IDLab MOSAIC - UAntwerpen
imec - CiTiP - KU Leuven