This paper discusses a symmetry transformation (Symmetric Wave Decomposition - SWD) that identifies the 'source' forcing functions that are contained in a time-series. Knowledge of the number of active elements (principal components) in a roller bearing is crucial to identifying how many elements may be defective and or if external forcing functions are dominant. With an understanding of these 'active components' the vibration data is decomposed into each active component wave. From this it can be estimated how many active components are associated with an event or events, if the damage is local or likely to be external and the extent of the damage. Vibration data from a large industrial machine containing large (4.2m diameter) slow speed (4 rpm) slew bearings has been processed with the SWD transform and the results are presented. The SWD algorithm is being implemented on an experimental test-rig that has been specifically built for the monitoring of slow speed (lrpm) bearings.