Sensor metrology data integration
Inventors
Bhatia, Sidharth • Sin, Garrett H. • Pai, Heng-Cheng • Nambiar, Pramod • Balasubramanian, Ganesh • Jamil, Irfan
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12366853-B2
Publication Date
2025-07-22
Expiration Date
Abstract
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.
Core Innovation
The invention relates to receiving sensor data from sensors associated with processing chambers of wafer processing equipment and receiving metrology data from metrology equipment, wherein the sensor data and the metrology data are associated with wafers processed via the processing chambers. It identifies corresponding sets of sensor data and corresponding sets of metrology data, and then generates aggregated sensor-metrology data sets, where each aggregated set comprises a respective set of sensor data and a respective set of metrology data.
The invention trains a machine learning model based on the plurality of sets of aggregated sensor-metrology data to provide a trained machine learning model configured to generate one or more outputs. The outputs include determining predicted metrology data and/or determining predicted metrology/property data, and may include determining optimal design of manufacturing equipment or manufacturing processes.
The trained machine learning model is also used to generate outputs for performing a corrective action associated with at least one of processing subsequent wafers via the processing chambers of the wafer processing equipment. The corrective action is associated with at least one of determining predicted metrology data or determining optimal design of one or more manufacturing equipment or manufacturing processes, and may be based on historical property data of products where the metrology data correspond to the historical property data.
Claims Coverage
Independent claim coverage is provided by three independent claims, each centered on aggregating sensor data with metrology data for wafer processing, training a machine learning model on the aggregated data, and generating outputs for corrective action, predicted metrology, and/or optimal design. Across the independent claims, the core inventive coverage is the same, implemented respectively as a method, a non-transitory computer readable medium, and a system with memory and a processing device.
Aggregated sensor-metrology training for machine learning outputs
Identifying a plurality of sets of sensor data from sensors associated with processing chambers of wafer processing equipment; identifying a plurality of sets of metrology data from metrology equipment; generating a plurality of sets of aggregated sensor-metrology data, each set comprising a respective set of sensor data and a respective set of metrology data; training, based on the plurality of sets of aggregated sensor-metrology data, a machine learning model to provide a trained machine learning model configured to generate one or more outputs for performing a corrective action for processing subsequent wafers and/or determining predicted metrology data and/or determining optimal design of one or more of manufacturing equipment or manufacturing processes.
Non-transitory medium for aggregated sensor-metrology training
A non-transitory computer readable medium storing instructions that, when executed, cause operations comprising identifying a plurality of sets of sensor data associated with wafers processed via processing chambers; identifying a plurality of sets of metrology data associated with the wafers; generating a plurality of sets of aggregated sensor-metrology data comprising respective sets of sensor data and metrology data; and training, based on the plurality of sets of aggregated sensor-metrology data, a machine learning model to provide a trained machine learning model configured to generate outputs for performing a corrective action for processing subsequent wafers and/or determining predicted metrology data and/or determining optimal design of manufacturing equipment or manufacturing processes.
System for aggregated sensor-metrology training
A system comprising a memory and a processing device to identify a plurality of sets of sensor data from sensors associated with processing chambers of wafer processing equipment; identify a plurality of sets of metrology data from metrology equipment; generate a plurality of sets of aggregated sensor-metrology data, each comprising a respective set of sensor data and a respective set of metrology data; and train, based on the plurality of sets of aggregated sensor-metrology data, a machine learning model to provide a trained machine learning model configured to generate outputs for performing a corrective action for processing subsequent wafers and/or determining predicted metrology data and/or determining optimal design of one or more of manufacturing equipment or manufacturing processes.
Across the independent claims, the core claim coverage is the generation of aggregated sensor-metrology data from identified sensor and metrology datasets associated with wafers processed in processing chambers, followed by training a machine learning model on the aggregated data to produce outputs used for corrective action, predicted metrology data (including predicted metrology/property data), and/or optimal design of manufacturing equipment or manufacturing processes. The independent claim set implements this core concept as a method, as instructions on a non-transitory computer readable medium, and as a system including memory and a processing device.
Stated Advantages
Reducing errors compared to manual association and OCR-based identifier matching.
Reducing time and energy and processor usage and bandwidth and storage costs compared to manual association and OCR-based identifier matching.
Documented Applications
Generating predicted metrology data (virtual metrology) and using model outputs to perform corrective actions associated with processing subsequent wafers via processing chambers of wafer processing equipment.
Performing corrective actions that may include displaying an alert in a graphical user interface, interrupting wafer processing equipment operation, and/or updating manufacturing parameters of the wafer processing equipment.
Determining optimal design of one or more of manufacturing equipment or manufacturing processes based on outputs of the trained machine learning model.
Interested in licensing this patent?