Pneumatic actuators exhibit highly nonlinear characteristics due to air compressibility, significant friction presence and the nonlinearities of control valves. Monitoring by acquiring the system's transfer function accurately can be difficult for nonlinear systems. This paper outlines the use of a statistical multivariate technique called Principal Component Analysis (PCA) and Projection to Latent Structure Discriminant Analysis (PLS-DA) and applies them to the monitoring of electro-pneumatic system. This approach is used to integrate system sensory using data (Pressure, displacement and vibration) from a printing machine, an electro-pneumatic system. Experiments were conducted under three conditions, normal, throttled and leaked system. The models were tested under different conditions with the experimental results, showing that the proposed technique can be used for electro-pneumatic system monitoring in a real-time environment and could successfully differentiate between new process conditions.