The present invention is a servo control device that controls a servomotor based on a command position. The servo control device includes a correction unit that corrects the command position using a first neural network that performs processing based on parameters representing a network structure, and a servo amplifier that controls the servomotor based on a corrected command position outputted from the correction unit. The correction unit corrects the command position based on the corrected command position and the actual position of the servomotor.

A method for improving a cycle time of a process of a product is provided. The method includes: collecting process profile data from a plurality of tool groups running the process, and calculating values of a plurality of key-performance-indicators (KPIs) of each tool group including calculating a standard deviation of an output of a stage of a bottleneck tool group of the tool groups; feeding the values of the KPIs and a work-in-progress (WIP) of each tool group into a neural network model in order to output an impact on the WIP for each KPI of each tool group by the neural network model; selecting a set of major KPIs of each tool group from the KPIs according to the impact of each tool group; and controlling the tool groups according to the impact of the set of major KPIs of each tool group in order to reduce a total WIP.

Implementations utilize deep reinforcement learning to train a policy neural network that parameterizes a policy for determining a robotic action based on a current state. Some of those implementations collect experience data from multiple robots that operate simultaneously. Each robot generates instances of experience data during iterative performance of episodes that are each explorations of performing a task, and that are each guided based on the policy network and the current policy parameters for the policy network during the episode. The collected experience data is generated during the episodes and is used to train the policy network by iteratively updating policy parameters of the policy network based on a batch of collected experience data. Further, prior to performance of each of a plurality of episodes performed by the robots, the current updated policy parameters can be provided (or retrieved) for utilization in performance of the episode.

Software for controlling processes in a heterogeneous semiconductor manufacturing environment may include a wafer-centric database, a real-time scheduler using a neural network, and a graphical user interface displaying simulated operation of the system. These features may be employed alone or in combination to offer improved usability and computational efficiency for real time control and monitoring of a semiconductor manufacturing process. More generally, these techniques may be usefully employed in a variety of real time control systems, particularly systems requiring complex scheduling decisions or heterogeneous systems constructed of hardware from numerous independent vendors.

An apparatus for controlling a system including a plurality of sources of signals causing a plurality of events includes an input interface to receive signals from the sources of signals, a memory to store a neural network trained to diagnose a control state of the system, a processor to submit the signals into the neural network to produce the control state of the system, and a controller to execute a control action selected according to the control state of the system. The neural network includes a sequence of layers, each layer includes a set of nodes, each node of at least an input layer and a first hidden layer following the input layer corresponds to a source of signal in the system. A pair of nodes from neighboring layers corresponding to a pair of different sources of signals are connected in the neural network only when a probability of subsequent occurrence of the events in the pair of the different sources of signals is above a threshold, such that the neural network is a partially connected neural network.

Operating a substrate processing system includes receiving a plurality of sets of training data, storing a plurality of machine learning models, storing a plurality of physical process models, receiving a selection of a machine learning model from the plurality of machine learning models and a selection of a physical process model from the plurality of physical process models, generating an implemented machine learning model according o the selected machine learning model, calculating a characterizing value for each training spectrum in each set of training data thereby generating a plurality of training characterizing values with each training characterizing value associated with one of the plurality of training spectra, training the implemented machine learning model using the plurality of training characterizing values and plurality of training spectra to generate a trained machine learning model, and passing the trained machine learning model to a control system of the substrate processing system.

A method of operating a polishing system includes training a plurality of models using a machine learning algorithm to generate a plurality of trained models, each trained model configured to determine a characteristic value of a layer of a substrate based on a monitoring signal from an in-situ monitoring system of a semiconductor processing system, storing the plurality of trained models, receiving data indicating a characteristic of a substrate to be processed, selecting one of the plurality of trained models based on the data, and passing the selected trained model to the processing system.

A method for configuring a semiconductor manufacturing process, the method including: providing an initial prediction model including a plurality of model parameters to one or more remote locations; receiving at least one updated model parameter from the one or more remote locations, the at least one model parameter is updated by training the initial prediction model with local data at the one or more remote locations; determining aggregated model parameters based on the at least one updated model parameter received from the one or more remote locations; and adjusting the initial prediction model based on the aggregated model parameters, the adjusted prediction model being operable to configure the semiconductor manufacturing process.

Implementations utilize deep reinforcement learning to train a policy neural network that parameterizes a policy for determining a robotic action based on a current state. Some of those implementations collect experience data from multiple robots that operate simultaneously. Each robot generates instances of experience data during iterative performance of episodes that are each explorations of performing a task, and that are each guided based on the policy network and the current policy parameters for the policy network during the episode. The collected experience data is generated during the episodes and is used to train the policy network by iteratively updating policy parameters of the policy network based on a batch of collected experience data. Further, prior to performance of each of a plurality of episodes performed by the robots, the current updated policy parameters can be provided (or retrieved) for utilization in performance of the episode.

A method of generating training spectra for training of a neural network includes measuring a first plurality of training spectra from one or more sample substrates, measuring a characterizing value for each training spectra of the plurality of training spectra to generate a plurality of characterizing values with each training spectrum having an associated characterizing value, measuring a plurality of dummy spectra during processing of one or more dummy substrates, and generating a second plurality of training spectra by combining the first plurality of training spectra and the plurality of dummy spectra, there being a greater number of spectra in the second plurality of training spectra than in the first plurality of training spectra. Each training spectrum of the second plurality of training spectra having an associated characterizing value.