A Systematic Literature Review: Machine Learning Approaches in Building Condition Assessment & Survey Systems

Building condition assessment plays a vital role in ensuring effective facility management and maintenance planning. Traditional methods, however, are often labor-intensive, prone to human error, and lack standardization, limiting their efficiency and reliability. The emergence of machine learning (ML) offers a transformative approach, enabling automated inspections and improved accuracy. This paper presents a systematic literature review (SLR) to explore the current state of ML applications in building condition assessment and survey systems. Following the SALSA (Search, Appraisal, Synthesis, Analysis) framework, 223 studies from leading databases such as Scopus and IEEE Xplore were meticulously analyzed based on strict inclusion criteria. The review reveals that Support Vector Machines (SVMs) and Convolutional Neural Networks (CNNs) are the most frequently applied ML techniques, achieving up to 65% time savings compared to traditional methods. While these advancements demonstrate the potential of ML, challenges remain in key areas such as data standardization, model generalization, and scalability. Additionally, the integration of ML systems with existing building assessment frameworks poses significant hurdles that must be addressed. This study provides a comprehensive overview of current ML-based advancements, highlighting their implications for more efficient and reliable building assessment processes. It also identifies critical research gaps, particularly in scalability, integration, and cross-domain adaptability, to foster widespread adoption of these innovative technologies in facility management practices.