Power transmissions are one of the most important parts of any mechanical system. In order to achieve the reliable operation of these systems, effective maintenance strategies must be used. Predictive health monitoring (PHM) and condition based maintenance (CBM) strategies are currently gaining in popularity due to their effectiveness in reducing maintenance costs; however, these require reliable monitoring techniques such as vibration analysis, acoustic emission, oil debris analysis and thermal analysis. This paper different monitoring features have been studied in order to develop an online monitoring tool that able to track the condition of an operating transmission system, classifying faults, and detect the onset of failure. The study presents an online PHM system utilising autoregressive (AR) parametric method algorithms based on vibration data. The online monitoring algorithm can support CBM and PHM of automotive multistage manual transmissions. The design, operation and validation of the online system are described and demonstrated. The results of
the experimental test prove the system's capability and support the recent trend of using CBM and PHM strategies.