Retrieval-Augmented Generation with Covariate Time Series

2026-03-05

AI Summary

This work addresses the challenge of distinguishing similar operational states in industrial time series settings characterized by data scarcity, short sequences, and tightly coupled covariates—such as predictive maintenance for Pressure Regulating and Shut-Off Valves (PRSOVs)—where existing retrieval-augmented generation (RAG) methods fall short. The authors propose RAG4CTS, a novel training-free, condition-aware RAG framework that preserves historical operating conditions losslessly via a hierarchical, time-series-native knowledge base. It aligns dynamic trends through a dual-weighted retrieval mechanism combining point-wise and multivariate similarity, and employs an agent-driven, self-supervised context optimization strategy to enable physics-informed adaptive fusion. By overcoming limitations of static embeddings and learnable augmenters, RAG4CTS significantly outperforms current approaches on PRSOV tasks, successfully identifying one real fault with zero false alarms within two months of deployment at China Southern Airlines.

Technology Category

• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Time-Series/Data Streams

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services

Abstract

While RAG has greatly enhanced LLMs, extending this paradigm to Time-Series Foundation Models (TSFMs) remains a challenge. This is exemplified in the Predictive Maintenance of the Pressure Regulating and Shut-Off Valve (PRSOV), a high-stakes industrial scenario characterized by (1) data scarcity, (2) short transient sequences, and (3) covariate coupled dynamics. Unfortunately, existing time-series RAG approaches predominantly rely on generated static vector embeddings and learnable context augmenters, which may fail to distinguish similar regimes in such scarce, transient, and covariate coupled scenarios. To address these limitations, we propose RAG4CTS, a regime-aware, training-free RAG framework for Covariate Time-Series. Specifically, we construct a hierarchal time-series native knowledge base to enable lossless storage and physics-informed retrieval of raw historical regimes. We design a two-stage bi-weighted retrieval mechanism that aligns historical trends through point-wise and multivariate similarities. For context augmentation, we introduce an agent-driven strategy to dynamically optimize context in a self-supervised manner. Extensive experiments on PRSOV demonstrate that our framework significantly outperforms state-of-the-art baselines in prediction accuracy. The proposed system is deployed in Apache IoTDB within China Southern Airlines. Since deployment, our method has successfully identified one PRSOV fault in two months with zero false alarm.

Problem

• Retrieval-Augmented Generation
• Time-Series Foundation Models
• Predictive Maintenance
• Covariate Time Series
• Data Scarcity

Innovation

• Retrieval-Augmented Generation
• Covariate Time Series
• Regime-Aware Retrieval
• Physics-Informed Knowledge Base
• Agent-Driven Context Augmentation