Slit valves play an important role in semiconductor manufacturing industry. They enable creation of a vacuum environment required for wafer processing. Due to the high volume of production in the modern semiconductor industry, slit valves experience severe degradation over their useful lifetime. If maintenance is not applied in due time, degraded valves may lead to defects in end products because of pressure loss and particle generation. In this thesis, we proposed methods for signal processing and feature extraction for analysis of slit valve vibration signatures. These methods would be used to demonstrate the ability of reliably, accurately and efficiently distinguish between each individual valve via a multi-class classification procedure. Such ability is a clear illustration of the feasibility of vibration based monitoring of the slit valve conditions. / text