A Framework for Explainable Root Cause Analysis in Manufacturing Systems - Combining Machine Learning, Explainable Artificial Intelligence and the Ishikawa Model for Industrial Manufacturing

Aug 6, 2026

This paper proposes a novel framework, TRACE (Transparent Reasoning in Artificial intelligence Cause Explanation), that integrates root cause analysis, explainable artificial intelligence, and machine learning in a worker-understandable way for industrial manufacturing. The framework targets transparency, interpretability, and explainability in AI-driven decision processes to improve acceptance of AI on the shop floor. The work motivates the need, details the design process, presents a preliminary mockup and software architecture, and outlines an evaluation and industrial integration plan.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Introduction

Root Cause Analysis

Theoretical Background, Related Work & Research Gap

Root Cause Analysis

Proceedings of the 58th Hawaii International Conference on System Sciences | 2025 A Framework for Explainable Root Cause Analysis in Manufacturing Systems – Combining Machine Learning, Explainable Artificial Intelligence and the Ishikawa Model for Industrial Manufacturing Daniel Kiefer Reutlingen University Daniel.Kiefer@ [Reutlingen-University.de](Reutlingen-University.de) Tim Straub Reutlingen University Tim. Straub@ [Reutlingen-University.de](Reutlingen-University.de) Günter Bitsch Reutlingen University Guenter.Bitsch@ [Reutlingen-University.de](Reutlingen-University.de) Clemens Van Dinther KIT [clemens.dinther2@kit.edu](clemens.dinther2@kit.edu) Abstract This paper proposes a novel framework –“Transparent Reasoning in Artificial intelligence Cause Explanation” (TRACE)– that combines root cause analysis, explainable artificial intelligence, and machine learning in an understandable way for the worker. The goal is to enhance transparency, interpretability, and explainability in AI-driven decision-making processes as well as to increase the acceptance of AI within an industrial manufacturing area. The paper outlines the need of such a framework, describes the design process, and shows a preliminary mockup, a possible underlying software architecture as well as an evaluation and integration plan in an industrial environment. Keywords: Explainable Artificial Intelligence, Root Cause Analysis, Ishikawa Model, Manufacturing Systems, TRACE Framework, Design Science. 1. Introduction Optimizing and understanding industrial processes is an important task for organizations (Ahmed et al., 2022) . There are two main approaches, that address this topic: The analytical approach built on domain knowledge, and the statistical approach including machine learning (ML) . Within the analytical field, a common strategy is the so-called“Root Cause Analysis” (RCA), which facilitates the identification of underlying causes of possibly occurring problems or faults (Ishikawa, 1990) . The analyses and explanations are key components in the context of manufacturing systems and serve as key factors in problem-solving and decision-making processes. In the area of statistical methods, the trend goes towards the use of Artificial Intelligence (AI) for predictions and the identification of relational dependencies. In many cases, AI is already transforming the operational landscape, for example in automating complex processes and unveiling new avenues for efficiency and productivity (Chakrabortyet al., 2017) . In comparison to traditional models, however, they are often hard to interpret or not interpretable at all, as slight parameter changes can have incomprehensible influence on the model’s outcome. For this reason, AI models are often referred to as black boxes, which has led to limited use of AI algorithms because stakeholders are unable to understand, trust, and effectively manage AI-driven processes (Doshi-Velez & Kim, 2017) . This resulted in an essential problem: In order to gain confidence in AI models or to derive further conclusions, one needs to understand the model (Ribeiro et al., 2016) . For this reason, the research field of explainable Artificial Intelligence (xAI) has emerged, which represents a family of methodologies that aim to enhance transparency and interpretability of AI algorithms, thereby building trust and promoting broader adoption of AI solutions (Samek et al., 2017) . In industrial areas, however, another problem arises: Workers are not necessarily used to dealing with results of xAI algorithms, as the influencing factors often are detached from their operational context. In contrast, traditional analytical RCA, such as Ishikawa (Ishikawa, 1990), are assumed to provide a better overview of the underlying processes and factors Hence, in order to maximize the understanding, the question comes up of how the explainable components of a ML model could be combined with familiar RCA tools. For this reason,

FAQs

What problem does the TRACE framework address in industrial manufacturing?

TRACE addresses the difficulty of using AI models in manufacturing when results are hard to interpret or not interpretable, limiting trust and adoption. It also targets the gap between xAI outputs and workers’ operational context.

How does TRACE combine machine learning with root cause analysis?

TRACE merges explainable AI methods with the Ishikawa model by placing the influence of ML parameters identified via xAI into the structured cause representations used in traditional RCA.

What outputs does the paper provide for validating TRACE?

The paper outlines the framework design, includes a preliminary mockup, proposes a possible software architecture, and describes an evaluation and integration plan for an industrial environment.

Aug 6, 2026

Upload by Emma Wilson