Example code used for the Predictive maintenace task of a Remote Terminal Unit (RTU). The main modules of this task are:

1. Data gathering (via Mqtt in a MongoDB) / simple Python scripts to subscribe to MQTT broker, store data in MongoDB and publish measurements
2. Data preprocessing
3. Time-series forecasting / using the Walk Forward Validation (WFV) method with different models (XGBoost, ARIMA, LSTM)
4. Anomaly Detection on the forecasted data (Isolation Forest, DBSCAN, Autoencoder)