Learning Rewards, Not Labels: Adversarial Inverse Reinforcement Learning for Machinery Fault Detection

Dhiraj Neupane, Richard Dazeley, Mohamed Reda Bouadjenek, Sunil Aryal

Keywords: Adversarial Learning, Anomaly Detection, Machinery Fault Detection, Reinforcement Learning, Prediction

TL;DR: We propose an Adversarial Inverse Reinforcement Learning framework for machinery fault detection that learns anomaly scores directly from healthy sequences, enabling earlier and more reliable fault diagnosis across multiple run-to-failure datasets

Abstract: Reinforcement learning (RL) offers significant promise for machinery fault detection (MFD). However, most existing RL-based MFD approaches do not fully exploit RL's sequential decision-making strengths as they treat MFD as a simple guessing game. To bridge this gap, we formulate MFD as an offline inverse RL (IRL) problem, where the agent learns the reward dynamics directly from healthy operational sequences, thereby bypassing the need for manual reward engineering and fault labels. Our framework employs Adversarial Inverse Reinforcement Learning (AIRL) to train a discriminator that distinguishes between normal (expert) and policy-generated transitions. The discriminator’s learned reward serves as an anomaly score, indicating deviations from normal operating behaviour. When evaluated on three run-to-failure benchmark datasets, the model assigns anomaly scores to all sequences, with normal samples receiving low scores and faulty ones high scores, thereby enabling early fault detection. By aligning RL’s sequential reasoning with MFD’s temporal structure, this work opens a path toward RL-based diagnostics in data-driven industrial settings. The implementation code for this proposed model is found in https://anonymous.4open.science/r/AIRL-MFD-DN-1434/.

Area: Innovative Applications (IA)

Generative A I: I acknowledge that I have read and will follow this policy.

Submission Number: 393