Abstract: A computer-implemented method for determining defect regions of products in a manufacturing process, is disclosed. The computer-implemented method includes steps of: obtaining experimental data from a machine; (b) obtaining first geometry data associated with historical products; (c) computing first geometrical parameters, based on the first geometry data associated with the historical products, by a geometry model; (d) computing second geometrical parameters, based on second geometry data associated with new products, by the geometry model; and (e) determining the defect regions in the new and historical products, based on the computed statistical features associated with defect types and locations, and the first and second geometrical parameters, by a machine learning model. The machine learning model is configured to determine the optimized recipe parameter to reduce the defect in at least one of: the new products and the historical products.

Abstract: A computer-implemented method for recommending a recipe to produce a product in a manufacturing process is disclosed. The computer-implemented method includes steps of: obtaining experimental data from a machine; generating a physics-based-simulation model based on the experimental data obtained from the machine; generating synthetic data for a first plurality of recipes using the physics-based-simulation model; determining an optimized physical range from physical ranges of each parameter by analyzing the experimental data and the synthetic data using a trained AI model; generating a second plurality of recipes when the optimized physical range of each parameter creating the second plurality of recipes is valid; validating the second plurality of recipes to extract an optimized recipe using the physics-based-simulation model; and recommending the optimized recipe for producing the product in the machine.

Abstract: A cross domain generalization system for industrial artificial intelligence (AI) applications is disclosed. A target encoder subsystem obtains target data from a target machine product and generates lower dimensional data for obtained target data using a target artificial intelligence (AI) model. The generated lower dimensional data are corresponding to a plurality of target embeddings data. The target encoder subsystem further applies the plurality of target embeddings data into a source classifier AI model. A source classifier subsystem predicts a quality of the target machine product by generating class labels for each of the plurality of target embeddings data based on a result of the classifier AI model. The goal of the present invention is to learn features or representations such that the correlation with a label space is similar both in source and target domains while being invariant of data distributions.