Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study evaluates the effectiveness of the univariate forecasting foundation model, TimesFM, for zero-shot multivariate time series anomaly detection (MTSAD) using the Secure Water Treatment benchmark. Despite the potential of foundation models for generalization, the results show that both proposed strategies—using TimesFM as a per-feature forecaster and as an embedder for outlier detection—fail to outperform established baselines, primarily due to the model's ability to capture temporal dynamics that obscure persistent anomalies. However, the research uncovers that while TimesFM struggles with MTSAD, it can effectively identify distribution changes at anomaly boundaries, suggesting a potential application in change-point detection.
Zero-shot foundation models may excel at capturing temporal dynamics but fall short in reliably detecting multivariate anomalies, revealing a critical gap in their application.
Multivariate Time Series Anomaly Detection (MTSAD) is essential for reliability and safety in domains such as industrial process monitoring and financial risk management, yet conventional approaches rely on application-specific models that are costly to train and hard to scale. Foundation Models (FMs), pre-trained on broad data with strong zero-shot generalization, have recently become available for univariate time series forecasting, raising the question of whether they can address MTSAD without task-specific training. We investigate the zero-shot application of a univariate forecasting FM, TimesFM, to industrial MTSAD on the Secure Water Treatment (SWaT) benchmark, evaluating two strategies: treating the FM as a per-feature forecaster with thresholded prediction errors, and as an embedder whose intermediate representations feed standard outlier detectors. Neither of our proposed setups is competitive with established baselines; embeddings reveal only partial separation between normal and anomalous segments, insufficient for reliable detection. The cause is that the FM is too effective at capturing temporal dynamics, yielding low error even within fully anomalous windows, so persistent anomalies become indistinguishable from normal behavior. However, these observations yield valuable insights: the error peaks at anomaly boundaries, indicating FMs reliably detect distribution changes. We conclude that the proposed naive zero-shot FMs are unsuitable for MTSAD but promising for change-point detection.