Wheel alignment method feasibility study

Sammanfattning: Throughout the history of motor vehicles, the tyres have always been consideredas one of the most important components of the vehicle due to their interactionwith the road. One important aspect is the wheel alignment, with the purposeto adjust the static wheel angles that are essential for many reasons, such assafety and fuel consumption for instance. Despite the numerous methods forwheel angle measurements, there seems to be no existing technical solutionbased on computer vision, that is suitable for residential use, regarding bothcost and size of the equipment. The study aims to investigate the feasibility ofsuch a system.The proposed system is based on planar fiducial markers called ArUco.From images or video frames of the marker, the pose of the marker can beestimated. Thus, by placing such markers on the ground, on the wheel andon the vehicle, the estimated pose of the markers can be used to measure andcalculate the wheel alignment parameters. Only toe and camber angles aremeasured within the scope of this thesis, even if the system has the potential tomeasure other wheel alignment parameters as well.After camera calibration, simplified ArUco marker tests were done by measuringthe known displacement and inclination of a marker with respect to areference marker. The mean absolute error was 030400 and 0:024mm for theinclination angle and displacement, respectively. Furthermore, the toe and camberangles of a vehicle were measured and compared to reference measurementsperformed with a commercial wheel alignment system, giving mean absoluteerrors of 0520 and 0280 for the camber and toe angles, respectively. Despitethe relatively large errors for the toe and camber angle measurements, theresults from the initial inclination and displacement tests show the potential ofthe system. In addition, several error sources and suggestions for improvementcan be identified.As a conclusion, the proposed system can be considered a working firstprototype, which after improvement and optimisation has the potential tobecome a feasible alternative, especially for residential use and for mobileworkshops due to the low cost, size and usability of the system.