公开(公告)号：US20210182693A1

公开(公告)日：2021-06-17

申请号：US16716993

申请日：2019-12-17

申请人： Tignis, Inc.

发明人： Jonathan L. Herlocker ,  Matt McLaughlin ,  Alexander Fry

IPC分类号： G06N3/08 ,  G16Y40/10 ,  G16Y40/20 ,  G06F17/18

摘要： A method for detecting anomalies in a physical system generates from a set of physics rules and a process graph representing the system a set of candidate physics models that assign physics rules to portions of the process graph representing sensors. Candidate physics models are rejected if an error between the models and sensor data exceed a predetermined error tolerance. Supervised learning is used to train a machine learning model to predict an error between the physics models and the sensor data. The predicted error and predicted sensor measurements from the physics models are then used to detect anomalies using unsupervised learning on a distribution of error between the predicted sensor measurements and the sensor data.

公开(公告)号：US20220261393A1

公开(公告)日：2022-08-18

申请号：US17175688

申请日：2021-02-14

申请人： Tignis, Inc.

发明人： Jonathan L. Herlocker ,  Adam Ashenfelter ,  Steven Herchak

IPC分类号： G06F16/23 ,  G06N20/00

摘要： A method for analyzing time series sensor data of a physical system represented by a process graph retrieves sensor data streams from stored sensor time series data. Each of the sensor data streams comprises a sequence of time-value pairs and is associated with a sensor identifier, a time offset, and a sampling period. A metric data stream is produced from the retrieved sensor data streams in accordance with a stored physics model of the physical system. Producing the metric data stream includes i) synchronizing the sensor data streams by adjusting time offsets of the sensor data streams and adding interpolated values and times to the sensor data streams to produce synchronized streams with equal sampling periods; and ii) performing a point-wise computation over values of the sensor data streams in accordance with the physics model.

公开(公告)号：US20220415683A1

公开(公告)日：2022-12-29

申请号：US17359626

申请日：2021-06-27

申请人： Tignis, Inc.

发明人： Ryan J. Stoddard ,  Jonathan L. Herlocker ,  Matt McLaughlin

IPC分类号： H01L21/67 ,  G06N3/04 ,  G06N3/08

摘要： A method is disclosed for generating a corrective film pattern for reducing wafer bow in a semiconductor wafer fabrication process. The method inputs to a neural network a wafer bow signature for a predetermined semiconductor fabrication step. The neural network generates from the input a corrective film pattern corresponding to the wafer bow signature. The neural network is trained with a training dataset of wafer shape transformations and corresponding corrective film patterns.

公开(公告)号：US11630820B2

公开(公告)日：2023-04-18

申请号：US17175688

申请日：2021-02-14

申请人： Tignis, Inc.

发明人： Jonathan L. Herlocker ,  Adam Ashenfelter ,  Steven Herchak

IPC分类号： G06F16/23 ,  G06F11/30 ,  G06F11/34

摘要： A method for analyzing time series sensor data of a physical system represented by a process graph retrieves sensor data streams from stored sensor time series data. Each of the sensor data streams comprises a sequence of time-value pairs and is associated with a sensor identifier, a time offset, and a sampling period. A metric data stream is produced from the retrieved sensor data streams in accordance with a stored physics model of the physical system. Producing the metric data stream includes i) synchronizing the sensor data streams by adjusting time offsets of the sensor data streams and adding interpolated values and times to the sensor data streams to produce synchronized streams with equal sampling periods; and ii) performing a point-wise computation over values of the sensor data streams in accordance with the physics model.