Causality-informed Anomaly Detection in Partially Observable Sensor Networks: Moving beyond Correlations

2025-07-13

AI Summary

In AI-driven manufacturing, sensor resource constraints lead to partial observability, hindering reliable real-time anomaly detection. Method: This paper proposes an end-to-end sensor placement optimization framework integrating causal inference with deep reinforcement learning. Contribution/Results: Its key innovation is the first explicit incorporation of causal structure into deep Q-network training under the no-intervention assumption—eliminating bias and risk from artificial anomaly injection. Causal-augmented state representation enhances policy interpretability and accelerates convergence, while yielding a tighter theoretical error bound. Experiments on diverse real-world data streams demonstrate significant reduction in average anomaly detection latency, achieving a favorable trade-off between deployment efficiency and detection performance. The method is scalable and suitable for large-scale industrial real-time monitoring systems.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Causal Learning
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Nowadays, as AI-driven manufacturing becomes increasingly popular, the volume of data streams requiring real-time monitoring continues to grow. However, due to limited resources, it is impractical to place sensors at every location to detect unexpected shifts. Therefore, it is necessary to develop an optimal sensor placement strategy that enables partial observability of the system while detecting anomalies as quickly as possible. Numerous approaches have been proposed to address this challenge; however, most existing methods consider only variable correlations and neglect a crucial factor: Causality. Moreover, although a few techniques incorporate causal analysis, they rely on interventions-artificially creating anomalies-to identify causal effects, which is impractical and might lead to catastrophic losses. In this paper, we introduce a causality-informed deep Q-network (Causal DQ) approach for partially observable sensor placement in anomaly detection. By integrating causal information at each stage of Q-network training, our method achieves faster convergence and tighter theoretical error bounds. Furthermore, the trained causal-informed Q-network significantly reduces the detection time for anomalies under various settings, demonstrating its effectiveness for sensor placement in large-scale, real-world data streams. Beyond the current implementation, our technique's fundamental insights can be applied to various reinforcement learning problems, opening up new possibilities for real-world causality-informed machine learning methods in engineering applications.

Problem

• Develop optimal sensor placement for partial observability
• Detect anomalies using causality beyond correlations
• Reduce detection time with causality-informed deep Q-network

Innovation

• Causality-informed deep Q-network for anomaly detection
• Optimal sensor placement with partial observability
• Integrates causal information in Q-network training