Do Time-Series Foundation Models Pay Off for Industrial Monitoring? A Cost-Aware Empirical Study
2026-08-24 • Machine Learning
Machine Learning
AI summaryⓘ
The authors studied different models for monitoring industrial equipment to find problems early. They compared classic methods, compact neural networks, and advanced time-series foundation models (TSFMs) across several tasks like detecting anomalies and forecasting failures. Their results showed that simpler models sometimes performed better or similarly to TSFMs depending on the task and resource limits. The authors suggest TSFMs should be chosen case-by-case, not as automatic upgrades. They also highlight differences in computational cost and responsiveness between the approaches.
Time-series foundation modelsAnomaly detectionAutoencodersForecastingAUROCAUPRCC-MAPSS datasetMIMII datasetResidual diagnosticsZero-shot learning
Authors
Guan-Hua Wen, Kuan-Yu Chen
Abstract
Industrial monitoring models must detect operationally relevant deviations while satisfying target-specific data, calibration, and resource constraints. Time-series foundation models (TSFMs) promise reusable representations and zero-shot forecasts, yet evidence for their deployment value remains mixed when task definitions are heterogeneous and lightweight baselines are competitive. This work presents a protocol-aware empirical assessment across three settings: a C-MAPSS degradation-risk proxy, normal-only training for anomalous-sound detection on MIMII, and BDG2 forecasting-residual diagnostics with synthetic target perturbations. We assess classical one-class methods, compact neural autoencoders, residual forecasters, MOMENT-small, Chronos-T5, and TimesFM 2.5 in terms of anomaly-ranking performance, risk-horizon sensitivity, residual forecasting and perturbation sensitivity, and local implementation cost. Across 100 C-MAPSS engines evaluated out of fold, TCN-AE reaches fold-weighted AUROC/AUPRC 0.9570/0.8960, compared with 0.7310/0.3080 for MOMENT reconstruction; paired engine-cluster bootstrap confidence intervals exclude zero for both differences. Across five matched MIMII pump evaluations, OCSVM also exceeds MOMENT reconstruction in AUROC and AUPRC. On a fixed 12-meter BDG2 panel, TimesFM 2.5 has the lowest aligned forecast error and the highest synthetic AUROC point estimate, although synthetic AUPRC is similar across TSFM and fitted residual models. Same-device measurements show that MOMENT incurs higher latency, peak allocated VRAM, and serialized state-dictionary size than TCN-AE. Under the evaluated frozen and zero-shot settings, TSFMs are task-dependent deployment options rather than default replacements for fitted lightweight models.