THE PROJECT

Our client operates a large number of unmanned weighbridges, where truck configuration must be compared with the measured gross vehicle mass to ensure that public roads are not being damaged.

We undertook the calculation of truck axle group configurations by automatically analysing videos taken at the weighbridge. A machine learning model detected wheels on the trucks and then derived distance to the next axle for identifying the axle groupings.

Various novel techniques were used to speed up processing, such as stitching together a region of interest clipped from each frame. Our algorithm has gone on to form the basis of a subsequent deployment for the client.

Contact us for a more comprehensive version of this case study.