Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainability for Fault Detection System in Chemical Processes

Georgios Gravanis, Dimitrios Kyriakou, Spyros Voutetakis, Simira Papadopoulou, Konstantinos Diamantaras

TL;DR

The paper tackles explainability for fault detection in chemical processes by comparing two post hoc XAI methods, IG and SHAP, applied to a high-accuracy LSTM classifier trained on the Tennessee Eastman Process. It evaluates whether explanations converge on the same influential features and whether the explanations align with physical plant understanding, using case studies such as IDV 11 and IDV 8. The findings show that IG and SHAP largely agree on important variables, though SHAP can offer more informative insights for certain faults, and the approach is model-agnostic and transferable. This work advances trust in DL-based fault detection in chemical processes by delivering interpretable explanations that support root-cause analysis and operational decision-making.

Abstract

In this work, we apply and compare two state-of-the-art eXplainability Artificial Intelligence (XAI) methods, the Integrated Gradients (IG) and the SHapley Additive exPlanations (SHAP), that explain the fault diagnosis decisions of a highly accurate Long Short-Time Memory (LSTM) classifier. The classifier is trained to detect faults in a benchmark non-linear chemical process, the Tennessee Eastman Process (TEP). It is highlighted how XAI methods can help identify the subsystem of the process where the fault occurred. Using our knowledge of the process, we note that in most cases the same features are indicated as the most important for the decision, while insome cases the SHAP method seems to be more informative and closer to the root cause of the fault. Finally, since the used XAI methods are model-agnostic, the proposed approach is not limited to the specific process and can also be used in similar problems.

Explainability for Fault Detection System in Chemical Processes

TL;DR

The paper tackles explainability for fault detection in chemical processes by comparing two post hoc XAI methods, IG and SHAP, applied to a high-accuracy LSTM classifier trained on the Tennessee Eastman Process. It evaluates whether explanations converge on the same influential features and whether the explanations align with physical plant understanding, using case studies such as IDV 11 and IDV 8. The findings show that IG and SHAP largely agree on important variables, though SHAP can offer more informative insights for certain faults, and the approach is model-agnostic and transferable. This work advances trust in DL-based fault detection in chemical processes by delivering interpretable explanations that support root-cause analysis and operational decision-making.

Abstract

In this work, we apply and compare two state-of-the-art eXplainability Artificial Intelligence (XAI) methods, the Integrated Gradients (IG) and the SHapley Additive exPlanations (SHAP), that explain the fault diagnosis decisions of a highly accurate Long Short-Time Memory (LSTM) classifier. The classifier is trained to detect faults in a benchmark non-linear chemical process, the Tennessee Eastman Process (TEP). It is highlighted how XAI methods can help identify the subsystem of the process where the fault occurred. Using our knowledge of the process, we note that in most cases the same features are indicated as the most important for the decision, while insome cases the SHAP method seems to be more informative and closer to the root cause of the fault. Finally, since the used XAI methods are model-agnostic, the proposed approach is not limited to the specific process and can also be used in similar problems.
Paper Structure (10 sections, 2 equations, 8 figures, 6 tables)

This paper contains 10 sections, 2 equations, 8 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Explainability methods
  3. Integrated Gradients
  4. SHapley Additive exPlanations
  5. TEP: A base case study
  6. Explainability experiments and results
  7. Result evaluation
  8. IDV 11 example case
  9. IDV 8 example case
  10. Conclusions

Figures (8)

  • Figure 1: Tennessee Eastman Process Piping and Instrumentation Diagram.
  • Figure 2: Heatmap with the most contributing features. The darker the color, the higher the contribution. A columns for IG , B columns for SHAP
  • Figure 3: IG and SHAP most important features for IDV 11.
  • Figure 4: IG attributions for IDV 11. Left: most important features based on IG score, right: less important features based on IG score.
  • Figure 5: Most and less contributing variable behavior plots for Fault 11.
  • ...and 3 more figures