Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Automated Knowledge Graph Learning in Industrial Processes

Lolitta Ammann, Jorge Martinez-Gil, Michael Mayr, Georgios C. Chasparis

TL;DR

This paper addresses the challenge of extracting meaningful structure from large industrial time-series by introducing an automated KG learning framework that converts sensor data into RDF-based knowledge graphs. The approach jointly leverages contemporaneous correlations and Granger-causality through a $VAR(p)$-based analysis, with automated time-lag selection and statistical testing to reveal causal influences. The framework is implemented as a web application supporting data preprocessing, correlation/causality analysis, and SPARQL-enabled KG querying, and is demonstrated on a real-world electrostatic particle transfer case to show actionable insights and improved interpretability. The work highlights the practical impact of structured industrial knowledge representations for decision support, root-cause analysis, and predictive modeling, while outlining future directions for scalability and integration with KG infrastructures.

Abstract

Industrial processes generate vast amounts of time series data, yet extracting meaningful relationships and insights remains challenging. This paper introduces a framework for automated knowledge graph learning from time series data, specifically tailored for industrial applications. Our framework addresses the complexities inherent in industrial datasets, transforming them into knowledge graphs that improve decision-making, process optimization, and knowledge discovery. Additionally, it employs Granger causality to identify key attributes that can inform the design of predictive models. To illustrate the practical utility of our approach, we also present a motivating use case demonstrating the benefits of our framework in a real-world industrial scenario. Further, we demonstrate how the automated conversion of time series data into knowledge graphs can identify causal influences or dependencies between important process parameters.

Automated Knowledge Graph Learning in Industrial Processes

TL;DR

This paper addresses the challenge of extracting meaningful structure from large industrial time-series by introducing an automated KG learning framework that converts sensor data into RDF-based knowledge graphs. The approach jointly leverages contemporaneous correlations and Granger-causality through a -based analysis, with automated time-lag selection and statistical testing to reveal causal influences. The framework is implemented as a web application supporting data preprocessing, correlation/causality analysis, and SPARQL-enabled KG querying, and is demonstrated on a real-world electrostatic particle transfer case to show actionable insights and improved interpretability. The work highlights the practical impact of structured industrial knowledge representations for decision support, root-cause analysis, and predictive modeling, while outlining future directions for scalability and integration with KG infrastructures.

Abstract

Industrial processes generate vast amounts of time series data, yet extracting meaningful relationships and insights remains challenging. This paper introduces a framework for automated knowledge graph learning from time series data, specifically tailored for industrial applications. Our framework addresses the complexities inherent in industrial datasets, transforming them into knowledge graphs that improve decision-making, process optimization, and knowledge discovery. Additionally, it employs Granger causality to identify key attributes that can inform the design of predictive models. To illustrate the practical utility of our approach, we also present a motivating use case demonstrating the benefits of our framework in a real-world industrial scenario. Further, we demonstrate how the automated conversion of time series data into knowledge graphs can identify causal influences or dependencies between important process parameters.
Paper Structure (12 sections, 1 theorem, 3 equations, 6 figures)

This paper contains 12 sections, 1 theorem, 3 equations, 6 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Contributions
  4. Motivating Example
  5. Methods
  6. Proposed Framework
  7. Pre-processing
  8. Correlation analysis
  9. Granger Causality Analysis
  10. Query capabilities
  11. Demonstrations
  12. Conclusions and Future Work

Key Result

Proposition 2.1

If $z_t$ is a stable VAR($p$) processThe stability condition of a VAR($p$) process can be verified by the eigenvalues of its coefficients matrix, cf. lutkepohl_new_2005. Stability also implies stationarity. with a nonsingular white noise covariance matrix $\sigma_u$, then

Figures (6)

  • Figure 1: Sketch of a real-world industrial process of electrostatic particle transfer.
  • Figure 2: A general overview of the proposed framework for automated KG discovery.
  • Figure 3: Combination of correlation analysis and Granger causality for search of contemporary and lagged relations between parameters. Green arrows depict inspected connections using Granger causality on VAR($p$) model, where $p$ corresponds to the time-lag over which causality is tested. The yellow arrow depicts contemporary correlation analysis.
  • Figure 4: Example of automated KG generation based on correlation and causality analysis in the electrostatic particle transfer industrial process of Section \ref{['sec:MotivatingExample']}.
  • Figure 5: Correlation analysis example in a synthetic data set.
  • ...and 1 more figures

Theorems & Definitions (1)

  • Proposition 2.1: Granger Noncausality