Semantic Knowledge Models for Industrial Data Analytics

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In industrial plants, especially in the steel industry, the complexity of production processes leads to errors due to concatenation of slight deviations in production conditions and unreliable measurements, making it difficult for human operators and standard statistical analysis to detect multivariate disturbances, resulting in higher costs and inefficient data analytics applications.

Innovation Solution

A method and system utilizing semantic knowledge models to instantiate industrial plant processes, enabling seamless integration and reuse of pre-processed data analytics results by describing plant structures, measurement data, storage locations, and transformation paths, allowing for automated data preprocessing and integration across different scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data analytics approaches are applied to analyze huge amounts of process data, then the ability to extract dependencies and identify multivariate disturbances is improved, but the complexity of data preprocessing and integration increases significantly

Engineering Contradiction:
Improvedetection accuracy of multivariate disturbancesVSAvoiddata preprocessing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the complex data preprocessing task into distinct modules: data collection from multiple sources, data transformation to standardized formats, data aggregation by product identity, and storage in a structured database. Each module handles a specific aspect of the preprocessing pipeline, making the overall system more manageable and reusable.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary data warehouse that stores pre-processed and standardized process data. This intermediary layer acts as a mediator between raw data sources and analytics applications, eliminating the need for each application to perform redundant preprocessing and enabling seamless data retrieval.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If manual preprocessing of data sources is performed by experts or special software algorithms, then data analytics applications can be realized, but the time consumption and effort increase

Engineering Contradiction:
Improvedata analytics capabilityVSAvoidpreprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by pre-processing and standardizing data from multiple sources before analytics applications are executed. Data is collected, transformed to unified formats, aggregated by product identity, and stored in advance in a data warehouse, so that when analytics applications run, they can directly access ready-to-analyze data without time-consuming preprocessing.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If different measurement intervals and precision are used for different process steps, then accurate process monitoring is achieved, but data integration and mapping become cumbersome

Engineering Contradiction:
Improveprocess data accuracyVSAvoiddata integration complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies local quality by allowing each data source to maintain its own measurement characteristics (intervals, precision, formats) while introducing a transformation layer that adapts data to a unified standard. Each process step's data is transformed according to its specific needs while contributing to a standardized overall structure, enabling both local optimization and global integration.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements parameter changes through data transformation processes that convert data from various measurement intervals and precision levels into a standardized format. The system changes parameters such as sampling intervals, data formats, and precision levels during the transformation phase, enabling seamless integration of heterogeneous data sources while preserving the original measurement qualities when needed.

Inventive Principle:
Principle #35Parameter changes

4Adaptability or versatility

If products consist of different parts of intermediate products or change direction during production, then production flexibility is improved, but data mapping and rededication become more difficult

Engineering Contradiction:
Improveproduction flexibilityVSAvoiddata mapping complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements dynamics by introducing a flexible data aggregation mechanism that dynamically tracks product identity through the production process. The system can adapt to changing product compositions and directions by dynamically updating product identifiers and mapping data to the correct product instances, regardless of how the product path changes during manufacturing.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3001266B1Method and system for providing data analytics results
Publication Date: 2020.01.08 SIEMENS AG
  • EP3001266B1 patent drawingFigure 1~2
  • EP3001266B1 patent drawingFigure 3
  • EP3001266B1 patent drawingFigure 4

AI summary

A system for providing data analytics results for a process performed in an industrial plant, said system (1) comprising: a knowledge model repository (2) configured to store semantic knowledge models, said semantic knowledge models comprising at least one semantic plant model of an industrial plant which describes semantically a configuration of the respective industrial plant and storage locations of process data provided by data sources of the industrial plant when performing at least one process therein, at least one semantic process model of a process which describes semantically the respective process steps of the process performed within an industrial plant; a selection unit adapted to select at least one analytics application which describes semantically at least one process step and at least one parameter required for accomplishing an analytics task; a processing unit (6) adapted to process the selected analytics application and selected instantiated semantic knowledge models to infer at least one storage location of at least one data source of the industrial plant, and an execution engine (7) adapted to execute the selected data analytics application using accessed process data provided by the inferred data sources of the industrial plant to generate the data analytics results.