A Hierarchical Ensemble Pipeline for Anomaly Detection in ESA Satellite Telemetry

TL;DR

This paper presents a hierarchical ensemble pipeline combining shapelet and statistical features for anomaly detection in ESA satellite telemetry, achieving strong generalization on ESA-ADB challenge data.

Contribution

It introduces a novel hierarchical ensemble approach with multi-level masking to improve anomaly detection in multivariate satellite telemetry data.

Findings

Effective detection of subtle anomalies in satellite telemetry.

Strong generalization demonstrated on ESA-ADB benchmark.

Hierarchical modeling enhances anomaly detection accuracy.

Abstract

A hierarchical ensemble pipeline is introduced to address anomaly detection in multivariate telemetry data provided by European Space Agency (ESA). The method integrates shapelet-based and statistical feature extraction, per-channel modeling, intra-channel stacking, and a final cross-channel aggregation. The pipeline is trained and validated using time-series cross-validation and two-level masking strategies to prevent information leakage. Results on the European Space Agency Anomaly Detection Benchmark (ESA-ADB) challenge demonstrate strong generalization, highlighting the effectiveness of hierarchical modeling in detecting subtle anomalies in realistic satellite telemetry.