Disclosed herein are methods for quality control of contact lenses. An example method comprises receiving an input image indicative of a subject contact lens. The example method comprises outputting, based on analysis of the input image and using a first learning model, a first foreign matter metric and rejection data based on the analysis of the input image. The example method comprises outputting, using an artifact model and based on at least the rejection data, an artifact metric. The artifact model may be implemented based on one or more artifact attributes. The example method comprises outputting, using a second learning model and based at least on the first foreign matter metric and the artifact metric, a second foreign matter metric. The second learning model may be scale dependent. The second foreign matter metric is indicative of at least an accept or reject condition of the subject contact lens.

A system for determining a processing procedure including a plurality of processes for controlling an object, the system includes a learning unit for performing a learning process for determining a processing condition of each of a plurality of processes, and the learning unit acquires a physical quantity correlating with a state of the object on which a process has been performed under a predetermined processing condition, from a device for controlling the object on the basis of the processing procedure, calculates a pseudo state corresponding to the state of the object on the basis of the physical quantity, performs a learning process using a value function, and determines a processing condition of each of the plurality of processes to achieve a target state of the object.

A system including a deep learning processor obtains response data of at least two data types from a set of process stations performing operations as part of a manufacturing process. The system analyzes factory operation and control data to generate expected behavioral pattern data. Further, the system uses the response data to generate actual behavior pattern data for the process stations. Based on an analysis of the actual behavior pattern data in relation to the expected behavioral pattern data, the system determines whether anomalous activity has occurred as a result of the manufacturing process. If it is determined that anomalous activity has occurred, the system provides an indication of this anomalous activity.

A motor position compensation method for an ophthalmic surgical laser system employs a deep artificial neural network to characterize motor following errors of the motors of the system. The artificial neural network is trained using a large number of commanded motor positions and corresponding measured actual motor positions (measured by encoders associated with the motors) as training data, to obtain a trained artificial neural network that can predict the actual motor position for any commanded motor position. Before executing a treatment scan, the original commanded motor positions calculated from the intended scan pattern are inputted to the trained artificial neural network to predict the actual motor positions, and the predicted actual motor positions are used to adjust the original commanded motor positions. The adjusted commanded motor positions are then used to perform the treatment scan, which produces an actual scan pattern that more closely match the intended scan pattern.

Controlling product production in multi-stage manufacturing process automatically generates by machine learning causal relationships between the processing conditions and the product quality based on product genealogy data and product quality data. Real time sensor data from sensors coupled to processing units in a manufacturing facility implementing the multi-stage manufacturing process are received, and control rules are instantiated based on the real time sensor data. An instantiated control rule firing causes an actuator to automatically set a processing variable to a set point specified in the control rule.

Robust machine learning predictions. Temporal dependencies of process targets for different machine learning models can be captured and evaluated for the impact on process performance for target. The most robust of these different models is selected for deployment based on minimizing variance for the desired performance characteristic.

Setting of parameters that determine filter characteristics is facilitated. Machine learning of optimizing the coefficients of a filter provided in a motor control device that controls rotation of a motor for a machine tool, a robot, or an industrial machine is performed on the basis of measurement information of an external measuring instrument provided outside the motor control device and a control command input to the motor control device.

A robot system including at least one robot arm and a control unit which is designed such that it can pre-set at least one pre-defined operation carried out by the robot system. The robot system also includes a display device and at least one input device applied to the robot arm, which is designed such that the sequence of operations of the robot system is set and/or the pre-defined operations of the robot system is parameterized by means of the input device, and which is also designed such that it allows the user to control, on a graphic user interface, represented by the control unit on the display device, the setting of the pre-defined operations of the robot system, the setting of the sequence of operations and/or the parameterization of the pre-defined operations for the robot system.

When a product is produced at the facility, a product ID of the product to be produced and setting values of a plurality of control parameters are received from a console terminal, and a transformation model corresponding to the combination of the product ID of the product to be produced and setting values of the plurality of the control parameters is read from a transformation model storage part. Then, in accordance with the read transformation model, the reference learning data stored in a reference learning data storage part is data-transformed and individual learning data corresponding to the product to be produced is generated, and with use of the individual learning data, whether measurement data output from sensors in the facility is abnormal is discriminated.

The disclosure provides a method and system of improvement in the real-time operation of a terminal station in an industrial facility using a machine learning-based self-adaptive system comprising obtaining real-time operations data and historical data stored in a local database or at a cloud-storage. The data relates to input parameters of the terminal station. The method includes inputting the input parameter to a machine learning configurable module of the machine learning-based self-adaptive system and analyzing the input parameter using dynamic machine learning models and algorithms to identify patterns to each of the input parameters. The method further includes evaluating the identified pattern against the real-time operations data obtained from the terminal station and predicting at least one output parameter based on the input parameter and the identified pattern against the real-time operations. Based on the prediction, adjusting the output parameter of the real-time operations data.