# Long Horizon Transformer Quantile Fault Prediction for Multi Site Industrial Predictive Maintenance

arXiv: https://arxiv.org/abs/2609.04840 · HTML: https://mlsanity.com/paper/2609.04840/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 3 of its 30 references that are in the map.
- First posted on arXiv 4 Sep 2026.

## Builds on (most cited first, 3 of 3)

- [Long-Range Transformers for Dynamic Spatiotemporal Forecasting](https://mlsanity.com/paper/2109.12218/) (2021)
- [Anomaly Detection in Predictive Maintenance: A New Evaluation Framework for Temporal Unsupervised Anomaly Detection Algorithms](https://mlsanity.com/paper/2105.12818/) (2021)
- [Boosted Enhanced Quantile Regression Neural Networks with Spatiotemporal Permutation Entropy for Complex System Prognostics](https://mlsanity.com/paper/2507.14194/) (2025)

## Abstract (arXiv)

Long-horizon predictive maintenance requires models to distinguish slowly evolving degradation from normal operating-regime variation over planning windows measured in days rather than hours. This paper evaluates whether an explicit conditional-quantile representation provides an informative classifier interface for this problem. The proposed TQRNN30d framework combines a dual-stage quantile regression neural network (QRNN) feature extractor with a multi-stream temporal fusion classifier. Each hourly word of 81-channel machine behaviour is mapped to a 324-dimensional quantile-state representation, and 720 ordered hourly words form the 30-day document supplied to the long-horizon model. The classifier fuses quantile states with dynamic covariates, channel-level static metadata, and a 168-hour latent-history stream using gated residual processing, causal recurrent encoding, and metadata-conditioned cross-modal attention. A bounded instability-aware signal derived from sustained one-word-ahead prediction-error divergence provides auxiliary memory modulation at the longest horizon. Evaluation uses a machine-disjoint 43/14/15 train/validation/test allocation across 72 machines in nine manufacturing facilities. At 30 days, TQRNN30d achieves 79.97% F1, 80.18% recall, 81.82% precision, 82.39% accuracy, and 0.820 ROC-AUC. It leads all 18 evaluated baselines at the 7-, 14-, and 30-day fixed-threshold comparisons, with the largest F1 advantage at 14 days. The results support held-out-machine performance within the observed homogeneous nine-facility fleet, but do not establish unseen-site, cross-equipment, or cross-sector generalisation.