A method and system for monitoring and predicting the health of electro-mechanical systems and components includes collecting data for a fixed pattern of actuation of such system or component. This data is used to build statistical models that correspond to a normal state of the system or component. New measurements are compared to this model in order to monitor the health of the system or component. The comparison can be made using a distance calculation. The combination of new measurements with historical data provides the prediction for future health states of the system or component.