Methods, systems, and apparatuses for developing linear process models to improve performance of components that make up operations in a plant are described herein. In some arrangements, a system may leverage one or more sensors and/or measurement devices to identify rates and compositions of feed and yield. The system may use one or more stoichiometric matrices and/or differential equations to identify molar and mass solutions for each feed component and predict the yield for reaction rates on a component-by-component basis. The system may further adjust the reaction rate coefficients to minimize the deviation between the yield results and the yield identified by system sensors and/or measuring devices. The resulting linear process models may be utilized to optimize plant processes in order to minimize reaction waste and maximize reaction yield.

A method includes obtaining one or more models associated with a model-based controller in an industrial process having multiple actuator arrays and performing spatial tuning of the controller. The spatial tuning includes identifying weighting matrices that suppress one or more frequency components in actuator profiles of the actuator arrays. The spatial tuning could also include finding a worst-case cutoff frequency over all output channels for each process input, designing the weighting matrices to penalize higher-frequency actuator variability based on the model(s) and the cutoff frequencies, and finding a multiplier for a spatial frequency weighted actuator variability term in a function that guarantees robust spatial stability. The controller could be configured to use a function during control of the industrial process, where a change to one or more terms of the function alters operation of the controller and the industrial process and at least one term is based on the weighting matrices.

A method and a corresponding system and computer program product are provided. A model of an object to be manufactured is obtained. Information about one or more available machine tools and one or more available cutting tools is obtained. A geometric feature to be machined as part of manufacturing the object is identified. A database including strategies for machining different geometric features is accessed. The database includes a plurality of strategies defining different ways of machining the identified geometric feature. One or more strategies are selected from the plurality of strategies based on the obtained information. A computer simulation is performed for the one or more selected strategies, and user instructions responsive to the computer simulation are received. Instructions for causing one or more machine tools to manufacture the object via subtractive manufacturing are provided based on the user instructions and a strategy of the one or more selected strategies.

The invention relates to a method of monitoring a machine in which a mechanical system is moved by a motor, wherein the mechanical system comprises more than two components coupled to one another, wherein one of the components comprises the motor, wherein at least the two components move differently when the motor drives the mechanical system, and wherein at least one movement factor of one of the components of the mechanical system is determined repeatedly; at least one dynamic factor of one of the components of the mechanical system is determined repeatedly; the movement factors of the remaining components of the mechanical system are calculated by means of a model of the mechanical system; and separate mechanical parameters for the components of the mechanical system are determined from the determined movement factor, from the determined dynamic factor, and from the calculated movement factors.

Embodiments provide methods, apparatuses, and systems for cutting wood workpieces, such as logs and cants, into desired products. In various embodiments, after a log is chipped into a cant, the cant may be scanned and re-optimized based on the new scan data and information about the source log, such as simulated orientation parameters, a 3D model, and/or potential cut solutions. In other embodiments, data from multiple sensor types may be used in combination to detect splits in logs, cants, or both. Optionally, re-optimization and split detection techniques may be used in combination to improve wood volume recovery, value, and/or throughput speed.

In certain embodiments, a control/optimization system includes an instantiated model object stored in memory on a model server. The model object includes a model of a plant or process being controlled. The model object comprises an interface that precludes the transmission of proprietary information via the interface. The control/optimization system also includes a decision engine software module stored in memory on a decision support server. The decision engine software module is configured to request information from the model object through a communication network via a communication protocol that precludes the transmission of proprietary information, and to receive the requested information from the model object through the communication network via the communication protocol.

An automated welding system includes a mounting platform, a welding tool, an imaging device configured to acquire data associated with an object, and a controller. The controller is configured to receive the acquired data, determine an area to be welded in the acquired data, retrieve stored master model data associated with the object, and compare the acquired data to the stored master model data to identify a master model area in the acquired data. The controller is also configured to mask the master model area in the acquired data, such that the master model area is excluded from the area to be welded, and generate control instructions for controlling at least one of the mounting platform and the welding tool to weld the area to be welded.

Systems and methods for controlling device performance variability during manufacturing of a device on wafers are disclosed. The system includes a process platform, on-board metrology (OBM) tools, and a first server that stores a machine-learning based process control model. The first server combines virtual metrology (VM) data and OBM data to predict a spatial distribution of one or more dimensions of interest on a wafer. The system further comprises an in-line metrology tool, such as SEM, to measure the one or more dimensions of interest on a subset of wafers sampled from each lot. A second server having a machine-learning engine receives from the first server the predicted spatial distribution of the one or more dimensions of interest based on VM and OBM, and also receives SEM metrology data, and updates the process control model periodically (e.g., to account for chamber-to-chamber variability) using machine learning techniques.

Provided is a numerical controller capable of easily controlling a pressure without a pressure sensor. The numerical controller estimates a pressure based on at least one of a command value and a feedback value. A machine learning device for learning the pressure corresponding to the at least one of the command value and the feedback value is included. The machine learning device includes a state observation unit for observing the at least one of the command value and the feedback value as a state variable, a label data acquisition unit for acquiring label data indicating the pressure, and a learning unit for associating and learning the state variable with the label data.

Techniques are described for receiving manufacturing data and events over real time and non-real time interfaces and associating the data with one another. In one example, real time data is received, the real time data associated with a counter value assigned by a precision counter. The received real time data is stored in a storage buffer, and non-real time data is received, where the non-real time data associated with a counter value assigned by the precision counter. In response to receiving the non-real time data, the buffer is searched for real time data having a matching counter value and, in response to identifying stored real time data associated with a matching counter value, the non-real time data is associated with the real time data based on the match. Data packages are generated including related real time and non-real time data associated with matching counter values.