RUB » Bibliotheksportal

Gadha Narayanan, Prathap Valluru

• The building sector in Germany consumes 35% of total energy, with residential buildings accounting for 21%. Predictive Maintenance (PdM) is crucial in detecting potential failures in heating systems and reducing unintended energy consumption. This article investigates various PdM techniques through extensive literature research, offering a strategic roadmap for future research. For complex heating systems with ample data, a hybrid approach combining model-based and datadriven methods is recommended. It efficiently analyzes data, simulates system behavior, and provides accurate predictions. For simpler systems, a knowledge-based approach relying on expert insights is suitable. Implementing PdM in building heating systems faces challenges like data quality, system complexity, and resistance to sensor installation costs. Limited maintenance access and data integration difficulties also arise. Research opportunities include exploring cost-effective virtual sensors, developing accurate predictive models for complex systems, and effectively utilizing BIM for PdM. Addressing these challenges and seizing research opportunities will pave the way for more affordable, reliable, and efficient predictive maintenance in building heating systems. By leveraging PdM, the building sector can optimize