IP Library Granted Patent US 12,725,094
Granted Patent B2
US 12,725,094 · App. 17/956,117 · Granted Sep 1, 2026

Data processing for industrial machine learning

Inventors: Benjamin Kloepper (Mannheim, DE); Benedikt Schmidt (Heidelberg, DE); Ido Amihai (Heppenheim, DE); Moncef Chioua (Montreal, CA); Jan Christoph Schlake (Darmstadt, DE); Arzam Muzaffar Kotriwala (Ladenburg, DE); Martin Hollender (Dossenheim, DE); Dennis Janka (Heidelberg, DE); Felix Lenders (Darmstadt, DE); Hadil Abukwaik (Weinheim, DE)
Assignee: ABB Schweiz AG
G06N20/20G06N20/10
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,725,094
App. No.
17/956,117
Granted
Sep 1, 2026
Kind
B2
Abstract

A computer-implemented method for automating the development of industrial machine learning applications includes one or more sub-methods that, depending on the industrial machine learning problem, may be executed iteratively. These sub-methods include at least one of a method to automate the data cleaning in training and later application of machine learning models, a method to label time series (in particular signal data) with help of other timestamp records, feature engineering with the help of process mining, and automated hyper-parameter tuning for data segmentation and classification.

Claims (33)

1 . A computer-implemented method for developing a machine learning model for an industrial asset, the method comprising:

acquiring a first time series of data from a sensor of an industrial asset;

processing the first time series of data to obtain an event log, wherein processing the first time series of data comprises encoding the time series of data into discrete events associated with a change in a state of the industrial asset;

applying process mining to the event log to provide a bottleneck identification, the bottleneck identification comprising bottlenecks in batch processes and deviations from standard operating procedures of the industrial asset;

applying the process mining to the event log to provide conformity analysis relative to a learned normal operation of the industrial asset;

continuously determining a condition indicator of the industrial asset based on the conformity analysis and the bottleneck identification, wherein the condition indicator indicates a deviation of performance of the industrial asset from a baseline;

developing a first machine learning model to monitor the industrial asset with the continuously determined condition indicator being an input parameter of the first machine learning model; and

applying the first machine learning model to determine potential improvements, to perform condition-based monitoring, and/or to perform predictive maintenance of the industrial asset.

2 . The computer-implemented method of claim 1 , wherein applying the first machine learning model further includes to determine process deviations and/or to predict how a batch process will evolve.

3 . The computer-implemented method of claim 1 , wherein the processing of the first time series of data to obtain the event log comprises encoding the first time series of data by applying symbolic aggregate approximation or artificial intelligence techniques.

4 . The computer-implemented method of claim 3 , wherein the processing of the first time series of data to obtain the event log further comprises performing abstractions on the encoded first time series of data.

5 . The computer-implemented method of claim 4 , wherein the abstractions performed on the encoded first time series of data comprise data aggregations and/or noise suppression filters.

6 . The computer-implemented method of claim 1 , further comprising:

acquiring a second time series of data;

cleaning the second time series of data to obtain a third time series of data; and

training a data cleaning machine learning model using a plurality of first training samples;

wherein a first training sample comprises a clean data point from the third time series of data and a plurality of raw data points from the second time series of data.

7 . The computer-implemented method of claim 6 , wherein the cleaning of the second time series of data comprises handling missing values, removing noise, and/or removing outliers.

8 . The computer-implemented method of claim 1 , further comprising:

acquiring a fourth time series of data from the sensor or from a control system associated with the industrial asset; and

applying a data cleaning machine learning model to the fourth time series of data to obtain the first time series of data.

9 . The computer-implemented method of claim 1 , further comprising:

acquiring a first set of labels for training a machine learning model for automatic labelling;

acquiring one or more data sources;

extracting a first set of features from the one or more data sources; and

training the machine learning model for automatic labelling using a plurality of second training samples;

wherein a second training sample comprises a label from the first set of labels and one or more features from the first set of features.

10 . The computer-implemented method of claim 9 , wherein the one or more data sources comprise at least one of a shift book, an alarm list, an events list, and/or a data source from a computerized maintenance management system; and/or wherein the machine learning model for automatic labelling is a probabilistic model.

11 . The computer-implemented method of claim 9 , further comprising:

extracting a second set of features from the one or more data sources; and

applying the machine learning model for automatic labelling to features from the second set of features to obtain a second set of labels.

12 . The computer-implemented method of claim 9 , wherein the first machine learning model is trained using a plurality of third training samples; and wherein a third training sample comprises a label from the first or second sets of labels and/or the condition indicator of the industrial asset.

13 . The computer-implemented method of claim 1 wherein acquiring the first time series of data further includes acquiring data from a control system associated with the industrial asset.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 13, 2023
From: KLOEPPER, BENJAMIN; SCHMIDT, BENEDIKT; AMIHAI, IDO; CHIOUA, MONCEF; SCHLAKE, JAN CHRISTOPH; KOTRIWALA, ARZAM MUZAFFAR; HOLLENDER, MARTIN; JANKA, DENNIS; LENDERS, FELIX; ABUKWAIK, HADIL
To: ABB SCHWEIZ AG
Reel/Frame 063927/0641 →
Continuity (3)
Continuation PCTEP2021056093 · Mar 10, 2021
Continuation PCTEP2020059135 · Mar 31, 2020
Related Publication 20230019404A1 · Jan 19, 2023
References Cited (49)
US 7391821B2 · Hu · 2008 [cited by examiner]
US 10474956B2 · Li · 2019 [cited by examiner]
US 10769909B1 · Modestine · 2020 [cited by examiner]
US 11113048B1 · Tanniru · 2021 [cited by examiner]
US 11258825B1 · Yang · 2022 [cited by examiner]
US 11314561B2 · Copier · 2022 [cited by examiner]
US 11551103B2 · Cook · 2023 [cited by examiner]
US 11768996B2 · Saripalli · 2023 [cited by examiner]
US 20150227838A1 · Wang · 2015 [cited by examiner]
US 20160104093A1 · Fletcher · 2016 [cited by examiner]
US 20170006914A1 · Birchler · 2017 [cited by examiner]
US 20180275667A1 · Liu · 2018 [cited by examiner]
US 20180307713A1 · Shin et al. · 2018 [cited by applicant]
US 20190065053A1 · Eads · 2019 [cited by examiner]
US 20200184278A1 · Zadeh · 2020 [cited by examiner]
US 20200201950A1 · Wang · 2020 [cited by examiner]
US 20210038163A1 · Agrawal · 2021 [cited by examiner]
US 20210264332A1 · Pingali · 2021 [cited by examiner]
US 20220206878A1 · Copier · 2022 [cited by examiner]
US 20250225368A1 · Nguyen · 2025 [cited by examiner]
CA 3055187A1 · 2018 [cited by examiner]
CN 105956077A · 2016 [cited by applicant]
CN 107666410A · 2018 [cited by examiner]
CN 109492772A · 2019 [cited by applicant]
CN 109784249A · 2019 [cited by examiner]
EP 3798911A1 · 2021 [cited by examiner]
EP 3902992B1 · 2024 [cited by examiner]
JP 8202444A · 1996 [cited by applicant]
JP 2011145846A · 2011 [cited by applicant]
JP 2013041448A · 2013 [cited by applicant]
JP 2014096050A · 2014 [cited by applicant]
JP 2020027424A · 2020 [cited by applicant]
KR 20210115991A · 2021 [cited by examiner]
WO WO2014043623A1 · 2014 [cited by applicant]
WO WO2020059099A1 · 2020 [cited by applicant]
Atzmueller et al., “Explanation-Aware Feature Selection using Symbolic Time Series Abstraction: Approaches and Experiences in a Petro-Chemical Production Context,” [cited by applicant]
Badakhshan et al., “The Action Engine—Turning Process Insights into Action,” [cited by applicant]
Berti, “Process Mining on Event Graphs: a Framework to Extensively Support Projects,” [cited by applicant]
Osman et al., “When Industry 4.0 meets Process Mining,” [cited by applicant]
Tax et al., “Event Abstraction for Process Mining using Supervised Learning Techniques,” [cited by applicant]
Veit et al., “The Proactive Insights Engine: Process Mining meets Machine Learning and Artificial Intelligence,” [cited by applicant]
European Patent Office, International Search Report in International Patent Application No. PCT/EP2021/056093, 4 pp. (Jun. 8, 2021). [cited by applicant]
European Patent Office, Written Opinion in International Patent Application No. PCT/EP2021/056093, 5 pp. (Jun. 8, 2021). [cited by applicant]
Rozinat, “How to Perform a Bottleneck Analysis With Process Mining,” webpage, downloaded from the Internet on Apr. 4, 2025, at: https://fluxicon.com/blog/2017/01/how-to-perform-a-bottleneck-analysis-with-process-mining/… [cited by applicant]
Canadian Intellectual Property Office, Office Action in Canadian Patent Application No. 3,173,398, 6 pp. (Mar. 17, 2025). [cited by applicant]
Cai et al., “A Real-Time Trace-Level Root-Cause Diagnosis System in Alibaba Datacenters,” [cited by applicant]
Wei et al., “Anomaly Prediction Approach in Business Process Based on Machine Learning,” [cited by applicant]
China National Intellectual Property Administration, Office Action in Chinese Patent Application No. 202180026316.6, 11 pp. (Aug. 28, 2025). [cited by applicant]
European Patent Office, Office Action in European Patent Application No. 21710006.4, 10 pp. (Oct. 16, 2025). [cited by applicant]