While driving a motorcycle at a race track knowing its position is desirable for various reasons. The position could be used as feedback to different vehicle systems or as a tool to analyze data after a track session. Positioning for vehicles are often done with global navigation satellite systems. However, modern motorcycles are usually equipped with many onboard sensors such as inertial measurement units and wheel speed sensors. When the motorcycle travel at a race track many of the signals recorded by these onboard sensors have a periodic behavior corresponding to a lap around the track.

This thesis work involves investigation of which of the signals recorded by the motorcycle's onboard sensors are suitable for positioning. It further includes development of methods to detect loops and perform localization based on features and hypothesis testing.

The methods developed are tested on recorded signals from motorcycles driving at race tracks and compared to recorded GPS positions. The developed localization algorithm shows promising results together with the developed loop detection algorithm. The estimated location does not drift over time but does lag behind the GPS. Further work should make it possible to increase the accuracyand robustness of the algorithms.