A data driven approach for fault detection in robotic actuation is disclosed. Here, a set of robotic tasks are received and analyzed by a Deep Learning (DL) analytics. The DL analytics includes a stateful (Long Short Term Memory) LSTM. Initially, the stateful LSTM is trained to match a set of activities associated with the robots based on a set of tasks gathered from the robots in a multi robot environment. Here, the stateful LSTM utilizes a master slave framework based load distribution technique and a probabilistic trellis approach to predict a next activity associated with the robot with minimum latency and increased accuracy. Further, the predicted next activity is compared with an actual activity of the robot to identify any faults associated robotic actuation.

A method and apparatus capable of monitoring performance of a process and of the condition of equipment units effecting such process is disclosed. A process model predicated upon mass and energy balancing is developed on the basis of a plurality of generally nonlinear models of the equipment units. At least one or more of such equipment models are characterized by one or more adjustable maintenance parameters. Data relating to mass and energy transfer within the process is collected and is reconciled with the mass and energy characteristics of the process predicted by the model. The condition of the equipment units and process performance may then be inferred by monitoring the values of the maintenance parameters over successive data reconciliation operations.

A plasma processing apparatus including: a monitor device which monitors a process quantity generated at plasma processing; a monitor value estimation unit which has monitor quantity variation models for storing change of a monitor value of the process quantity in accordance with the number of processed specimens and which estimates a monitor value for a process of a next specimen by referring to the monitor quantity variation models; and a control quantity calculation unit which stores a relation between a control quantity for controlling the process quantity of the vacuum processing device and a monitor value and which calculates the control quantity based on a deviation of the estimated monitor value from a target value to thereby control the process quantity for the process of the next specimen. Thus, stable processed results can be obtained even when variation occurs in processes.

A method of troubleshooting a process in an industrial plant is described in which a digital model of the industrial plant, which associates parameters with components of the industrial plant, and each parameter has a nominal range, has values selected for a specific parameter. The values extend below, within, and above the nominal range of the parameter. The digital model is executed using each of the values. Each of the values is associated with an output of the digital model, and the associations are saved in a database. Outputs associated with values within the nominal range of the parameter are distinguished from outputs associated with values extending below or above the nominal range of the parameter, and such correlations are saved in the database and called upon when troubleshooting errors in the plant.

A system and method for unsupervised prediction of machine failures. The method includes monitoring sensory inputs related to at least one machine; analyzing, via at least unsupervised machine learning, the monitored sensory inputs, wherein the output of the unsupervised machine learning includes at least one indicator; identifying, based on the at least one indicator, at least one pattern; and determining, based on the at least one pattern and the monitored sensory inputs, at least one machine failure prediction.

An on-line data analytics device can be installed in a process control system as a standalone device that operates in parallel with, but non-intrusively with respect to, the on-line control system to perform on-line analytics for a process without requiring the process control system to be reconfigured or recertified. The data analytics device includes a data analytics engine coupled to a logic engine that receives process data collected from the process control system in a non-intrusive manner. The logic engine operates to determine further process variable values not generated within the process control system and provides the collected process variable data and the further process variable values to the data analytics engine. The data analytics engine executes statistically based process models, such as batch models, stage models, and phase models, to produce a predicted process variable, such as an end of stage or end of batch quality variable for use in analyzing the operation of the on-line process.

This system-analyzing device has an analysis-model generation unit, and said analysis-model generation unit includes a data-point categorization unit, a many-body-correlation-model generation unit, and a model extraction unit. The data-point categorization unit categorizes a plurality of types of data points for a target system into one or more groups on the basis of how good a regression equation containing a given two of said data points is, and for each of said groups, the many-body-correlation-model generation unit selects a representative data point and generates a many-body-correlation model that includes at least the following: a regression equation containing the representative data point and one of two sets of data points from the group in question; and the allowable prediction-error range for said regression equation. The model extraction unit extracts one or more of the generated many-body-correlation models on the basis of how good each regression equation is.

A method includes identifying sets of sensor data associated with wafers processed via wafer processing equipment and identifying sets of metrology data associated with the wafers processed via the wafer processing equipment. The method further includes generating sets of aggregated sensor-metrology data, each of the sets of aggregated sensor-metrology data including a respective set of sensor data and a respective set of metrology data. The method further includes causing, based on the sets of aggregated sensor-metrology data, performance of a corrective action associated with the wafer processing equipment.

A method includes identifying sets of a first type of data associated with wafers processed via processing chambers of wafer processing equipment and identifying sets of a second type of data associated with the wafers processed via the processing chambers of the wafer processing equipment. The first type of data is different than the second type of data. The method further includes generating sets of aggregated data, where each of the sets of aggregated data includes a respective set of the first type of data and a respective set of the second type of data. The method further includes causing, based on the sets of aggregated data, performance of a corrective action associated with adjusting at least one operation associated with the wafer processing equipment.