ABSTRACT
Different types of faults can happen to a railway system. If the fault is severe it is often more easy to understand what is wrong. But if the fault is less severe it can be more difficult to find. Often there is no need to stop the train because of the fault, the fault may only cause more wear and a less comfortable ride.
In this paper we have suggested an automatic fault detection system, which continuously measures the acceleration in the vehicle. From the measured accelerations “Estimated Gain Functions” are created, which purpose is to guide the workshop when searching for the fault(s). The accelerometer signals will be uploaded to the cloud. Giving the possibility to watch for trends. Making it possibility to repair the fault(s) until they gets too severe.
