Abstract: A method for controlling a system including a plurality of subsystems, includes receiving operational data from the plurality of subsystems of the system (S21). A future condition of each of the plurality of subsystems is estimated from the received operational data (S22). A control strategy for delaying a need for system maintenance is generated based on the estimated future condition of each of the plurality of subsystems (S23). An operation of the system is controlled based on the generated control strategy (S24).

Abstract: A prognostics analysis software module is embedded in a programmable logic controller (PLC) software platform. During cycling of the PLC real-time operating program, data is read from sensors and written to a buffer only when the prognostics analysis software module is idle. The prognostics analysis software module is then activated by a system function block of the PLC software platform. Before determining any prognostic information, prediction models within the prognostics analysis software module are automatically trained using features extracted from the sensor data.

Abstract: A method for estimating a remaining useful life of a system includes monitoring sensor data from sensors deployed within a system. A plurality of features are extracted from the sensor data. Tree graphs are generated including mathematical operators and features as nodes and a advanced feature is produced from each of the tree graphs by transforming the tree graphs into equations. A recursive operation including analyzing a fitness of each of the advanced features, performing crossover/mutation on the tree graphs, producing advanced features from the altered tree graphs, and analyzing the fitness of the altered tree graphs to produce at least one final advanced feature is performed. A remaining useful life of the system is calculated based on the final advanced feature.

Abstract: A machine tool system is diagnosed by identifying a fault class to which an input measurement vector belongs. The fault class corresponds to a group of weight vectors in a code book of a self organized map that describes the machine tool system based on training data. Probabilities that the input measurement vector belongs to a given class are estimated based on the posterior probability of the weight vectors of the code book corresponding to the given class given the input measurement vector. Training data to create the code book may be collected under a first operating condition while the input measurement vector is collected under a second operating condition.

Abstract: A method for detecting an anomaly in a machine under test includes monitoring operational data from a control unit of the machine under test. An operational state of the machine under test is identified based on the monitored operational data. Sensor data is monitored from one or more sensors installed within or near to the machine under test. A model corresponding to the identified operational state of the machine under test is consulted to identify one or more key parameters and corresponding normal operating ranges for each determined key parameter. It is determined when a key parameter of the one or more key parameters is not within its corresponding normal operating range based on the monitored sensor data.

Abstract: A method of prognosing a mechanical system to predict when a failure may occur is disclosed. Measurement data corresponding to the mechanical system is used to extract one or more features by decomposing the measurement data into a feature space. A prediction model is then selected from a plurality of prediction models for the one or more features based at least on part on a degradation status of the mechanical system and a reinforcement learning model. A predicted feature space is generated by applying the selective prediction model to the feature space as well as a confidence value by comparing the predicted feature space with a normal baseline distribution, a faulty baseline distribution, or a combination thereof. A status of mechanical system based at least in part on the confidence value is then provided.

Abstract: A method of prognosing a mechanical system to predict when a failure may occur is disclosed. Measurement data corresponding to the mechanical system is used to extract one or more features by decomposing the measurement data into a feature space. A prediction model is then selected from a plurality of prediction models for the one or more features based at least on part on a degradation status of the mechanical system and a reinforcement learning model. A predicted feature space is generated by applying the selective prediction model to the feature space as well as a confidence value by comparing the predicted feature space with a normal baseline distribution, a faulty baseline distribution, or a combination thereof. A status of mechanical system based at least in part on the confidence value is then provided.