Wien Energie’s predictive maintenance initiative combined network sensor readings with historical work records and field-service feedback. Its workflow shows how data preparation, fault labeling and machine-learning classification can help maintenance teams prioritize potential problems in a large district heating system.