Distinguishing Process Change, Instrument Fault and Data-Path Corruption in Streaming Telemetry: A Reference Implementation and Benchmark
The paper tackles the 'three-cause problem' where a reading deviates from expectation due to process change, instrument failure, or data corruption. It proposes a robust system using a bank of online detectors—including Kalman filters and statistical tests—combined with an inspectable attribution rule. This method successfully detects anomalies in synthetic and real industrial signals while accurately attributing the root cause.