Title webinar

Knowledge-Based vs. Data-Driven?

Fault Detection and Diagnosis in Buildings: Theories and Practical Challenges

 

Description

Due to faults and improper controls, 15%-30% of energy may be wasted in buildings. It is crucial to develop fault detection and diagnosis (FDD) algorithms for building energy systems. Over the past decade, the emergence of artificial intelligence has sparked a surge in interest towards data-driven FDD approaches, which account for approximately 70% of all related publications. Despite this academic enthusiasm, the adoption of data-driven FDD in the market has been sluggish. According to reports from the Lawrence Berkeley National Laboratory in the US, conventional rule-based or expert knowledge-based FDD methods still dominate the market.

 

This disjunction between academia and industry prompts questions about the barriers hindering the real-world implementation of these advanced algorithms. What challenges impede their adoption, and what viable solutions can be pursued? In the webinar, Chujie will share updates and insights from the ongoing efforts to develop FDD tools for the B4B living labs.

Speakers:  Chujie Lu (TU Delft) - Postdoctoral Researcher in the Faulty of Architecture and the Built Environment. He has a multidisciplinary background with a PhD in Computer Science and a BEng in Building Environment and Energy Engineering.