Technical faults cause 19% of all transport delays in the rail network. Malfunctions of automatic train doors account for 20% of these technical faults, i.e. for almost 4% of all delays and there is recent and increasing interest in the use of remote condition monitoring (RCM) of these doors. So far the RCM systems implemented monitor only operational electronic and time parameters and use unusual values of these as an indicator of likely developing faults. However, the root cause of unusual electronic conditions are usually developing mechanical wear faults of multiple causes which eventually cause excessive noise, overheating and electrical failure. Therefore the aim of this project is the development of an embedded low cost system for the direct monitoring of the health of the door mechanisms through the vibrational spectra captured by an accelerometer, with abnormal accelerometer signatures being used to provide early automated warnings before door failure occurs.