公开(公告)号：US20230384752A1

公开(公告)日：2023-11-30

申请号：US18448523

申请日：2023-08-11

Applicant: ABB Schweiz AG

Inventor： Pablo Rodriguez ,  Jens Doppelhamer ,  Benjamin Kloepper ,  Reuben Borrison ,  Marcel Dix ,  Benedikt Schmidt ,  Hadil Abukwaik ,  Arzam Muzaffar Kotriwala ,  Sylvia Maczey ,  Dawid Ziobro ,  Simon Hallstadius Linge ,  Marco Gaertler ,  Divyasheel Sharma ,  Chandrika K R ,  Gayathri Gopalakrishnan ,  Matthias Berning ,  Roland Braun

IPC: G05B19/05

CPC classification number: G05B19/056 ,  G05B2219/1204

Abstract: A method includes acquiring state variables that characterize an operational state of an industrial plant; acquiring interaction events of a plant operator interacting with the distributed control system via a human-machine interface; determining based on the interaction events, and with state variables as input data, whether one or more interaction events are indicative of the plant operator executing a task that is not sufficiently covered by engineering of the distributed control system. When this determination is positive, mapping the input data to an amendment and/or augmentation for the engineering tool that has generated the application code.

公开(公告)号：US20230221684A1

公开(公告)日：2023-07-13

申请号：US18184043

申请日：2023-03-15

Applicant: ABB Schweiz AG

Inventor： Benjamin Kloepper ,  Arzam Muzaffar Kotriwala ,  Marcel Dix

IPC: G05B13/02 ,  G05B13/04

CPC classification number: G05B13/0265 ,  G05B13/045

Abstract: An explainer system includes a system-monitor machine learning model trained to predict states of a monitored system, a perturbator applying predetermined perturbations to original sample data collected from the monitored system to produce perturbed sample data. The system is configured to input the perturbed sample data to the prediction system. The explainer comprises a tester that receives model output from the prediction system, the model output comprising original model output produced by the system-monitor machine learning model based on the original sample data and deviated model output produced by the system-monitor machine learning model based on the perturbed sample data, the deviated model output comprising deviations from the original model output, the deviations resulting from the applied perturbations. An extractor receives data defining the perturbations and the resulting deviations and extracts therefrom important features for explaining the model output.

公开(公告)号：US20230214724A1

公开(公告)日：2023-07-06

申请号：US18184279

申请日：2023-03-15

Applicant: ABB Schweiz AG

Inventor： Arzam Kotriwala ,  Andreas Potschka ,  Benjamin Kloepper ,  Marcel Dix

IPC: G06N20/00

CPC classification number: G06N20/00

Abstract: A method and system for removing undesirable inferences from a machine learning model include a search component configured to receive a rejected explanation of model output provided by the machine learning model, identify data samples to unlearn by selecting training samples from training data that were used to train the machine learning model, the selected training samples being associated with explanations that are similar to the rejected explanation according to a calculated similarity measure, and pass the data samples to unlearn to a machine unlearning unit.

公开(公告)号：US20230094914A1

公开(公告)日：2023-03-30

申请号：US17956097

申请日：2022-09-29

Applicant: ABB Schweiz AG

Inventor： Benedikt Schmidt ,  Ido Amihai ,  Arzam Muzaffar Kotriwala ,  Moncef Chioua ,  Felix Lenders ,  Dennis Janka ,  Martin Hollender ,  Jan Christoph Schlake ,  Hadil Abukwaik ,  Benjamin Kloepper

IPC: G06N20/00

Abstract: A computer-implemented method of generating a training data set for training an artificial intelligence module includes providing first and second data sets, the first data set including first data elements indicative of a first operational condition, the second data set including second data elements indicative of a second operational condition that matches the first operational condition. The method further comprises determining a data transformation for transforming the first data elements into the second data elements; applying the data transformation to the first data elements and/or to further data elements of further data sets, thereby generating a transformed data set; and generating a training data set for training the AI module based on at least a part of the transformed data set.

公开(公告)号：US20230050321A1

公开(公告)日：2023-02-16

申请号：US17977355

申请日：2022-10-31

Applicant: ABB Schweiz AG

Inventor： Benedikt Schmidt ,  Marcel Dix ,  Martin Hollender ,  Andrew Cohen ,  Arzam Muzaffar Kotriwala ,  Marco Gaertler ,  Sylvia Maczey ,  Benjamin Kloepper

IPC: G05B19/042

Abstract: A method for generating a process model modeling a manual mode procedure instance of a plant process includes providing log events of operational actions; selecting related sequences of manual mode operational actions from the log events; filtering the related sequences according to an individual plant section; identifying a sequential order from the filtered related sequences; determining statistical properties of values of related process variables and/or statistical properties of values of related set point changes to each sequential ordered manual mode operational action from the filtered related sequences; generating the process model of the manual mode procedure instance by arranging related manual mode operational actions with the sequential order of each operational action assigned with the statistical properties of the values of related process variables and/or assigned with the statistical properties of the values of the related set point changes.

公开(公告)号：US11556111B2

公开(公告)日：2023-01-17

申请号：US16921957

申请日：2020-07-07

Applicant: ABB Schweiz AG

Inventor： Subanatarajan Subbiah ,  Benjamin Kloepper

IPC: G05B19/4155 ,  G06N20/00

Abstract: A method for controlling an industrial process includes: determining, by a process controller, based at least in part on a set of current values and/or past values of state variables of the industrial process, a set of control outputs to be applied to at least one actor and/or lower-level controller configured to cause a performing of at least one physical action on the process; querying, based on at least a subset of the set of current values and/or past values of state variables and on at least a subset of the set of control outputs, a trained machine-learning model configured to output a classification value, and/or a regression value, that is indicative of a propensity of a watching human operator to at least partially override the control outputs delivered by the process controller; and determining that the classification value, the regression value, and/or the propensity, meets a predetermined criterion.

公开(公告)号：US20220342382A1

公开(公告)日：2022-10-27

申请号：US17725490

申请日：2022-04-20

Applicant: ABB Schweiz AG

Inventor： Hadil Abukwaik ,  Jens Doppelhamer ,  Marcel Dix ,  Benjamin Kloepper ,  Pablo Rodriguez

IPC: G05B19/4065

Abstract: A system and method provides an impact list of affecting equipment elements that affect an industrial sub-process. The method comprises the steps of selecting, in a topology model, the sub-process, wherein the sub-process is an equipment element that is a part of an industrial plant or process, and wherein the topology model is a graph, whose nodes represent equipment elements and whose edges represent interconnections between the equipment elements; traversing the nodes of the topology model, wherein the traversing starts from the selected sub-process and uses a traversing strategy; and for each of the at least one equipment elements, if the equipment element affects the industrial sub-process by an affecting degree greater than a first predefined affecting degree, adding the equipment element to the impact list of affecting equipment elements.

公开(公告)号：US11238371B2

公开(公告)日：2022-02-01

申请号：US16441028

申请日：2019-06-14

Applicant: ABB Schweiz AG

Inventor： Benjamin Kloepper ,  Benedikt Schmidt ,  Mohamed-Zied Ouertani

IPC: G06N20/00 ,  H04W4/38 ,  G05B23/02

Abstract: A computer system can be configured to: receive, in a low-precision mode, first status data generated by one or more sensors, the first status data reflecting technical parameters of a technical system, the first status data exhibiting a first precision level; apply a low-precision machine learning model to analyze the first status data for one or more indicators of an abnormal technical status, the machine learning model having been trained with data exhibiting the first precision level; send, based on an abnormal technical status being indicated, instructions for the one or more sensors to generate second status data exhibiting a second precision level, the second precision level being associated with greater accuracy than the first precision level; receive the second status data exhibiting the second precision level based on the sent instructions; providing the second status data to a data analyzer.

公开(公告)号：US10969774B2

公开(公告)日：2021-04-06

申请号：US16576832

申请日：2019-09-20

Applicant: ABB Schweiz AG

Inventor： Martin Hollender ,  Benjamin Kloepper ,  Michael Lundh ,  Moncef Chioua

IPC: G06G7/00 ,  G05B23/02 ,  G05B13/04 ,  G06N20/10

Abstract: An anomaly detection module is configured to apply a plurality of machine learning models to received technical status data to detect one or more indicators of an abnormal technical status prevailing in the industrial process system. The plurality of machine learning models are trained on historic raw or pre-processed sensor data and the anomaly detection module configured to generate the anomaly alert based on the one or more indicators. The received technical status data is assigned to signal groups and the generated anomaly alert is a vector with each vector element representing a group anomaly indicator for the respective signal group. Each vector element is determined by applying a respective group specific machine learning model.

公开(公告)号：US20250117537A1

公开(公告)日：2025-04-10

申请号：US18929853

申请日：2024-10-29

Applicant: ABB Schweiz AG

Inventor： Joakim Astrom ,  Divyasheel Sharma ,  Yemao Man ,  Gayathri Gopalakrishnan ,  Benjamin Kloepper ,  Dawid Ziobro ,  Benedikt Schmidt ,  Arzam Muzaffar Kotriwala ,  Marcel Dix

IPC: G06F30/18 ,  G06F30/27 ,  G06F113/14

Abstract: A method for interactive explanations in industrial artificial intelligence systems includes providing a machine learning model and a set of test data, a set of training data and a set of historical data simulating a piping and process equipment; predicting a result for the piping and process equipment based on the machine learning model using the set of test data and the set of training data, wherein the set of historical data is used by the machine learning model to predict at least one parameter of the piping and process equipment; and presenting the predicted at least one parameter on a piping and instrumentation diagram of the piping and process equipment.