🛰️ SCADA Anomaly Detector (LSTM Autoencoder)

An LSTM autoencoder that flags abnormal pump behaviour from multivariate SCADA signals — trained on normal operation only, no fault labels. Anomalies reveal themselves as high reconstruction error.

Model: LSTM Autoencoder · Domain: EMS SCADA · The architecture behind my M.Tech thesis (Objective 3). Trained on synthetic SCADA — swap in plant historian data for production.