In the heart of Ávila, Spain, within the historic walls of the Nuestra Señora convent, a groundbreaking study is revolutionizing the way we approach heritage conservation. Pablo Sanz-Honrado, a researcher at the Institute of Physical and Information Technologies Leonardo Torres Quevedo (ITEFI), CSIC, has pioneered a novel methodology that combines 3D point clouds and machine learning to assess deflection in historic timber slabs with unprecedented accuracy.
Traditionally, 3D point clouds have been instrumental in creating detailed planimetries and computational models, but Sanz-Honrado’s research pushes these capabilities further into the realm of diagnostics. “We saw an opportunity to leverage the high resolution and accuracy of 3D point clouds for more than just visualization,” Sanz-Honrado explains. “By integrating Artificial Intelligence, we can now automatically detect and evaluate structural issues like beam deflection.”
The methodology employs a multi-resolution Random Forest classifier to identify beams and their faces, achieving a remarkable 99% accuracy in beam detection. Deflection is then computed using connected component and minimum bounding rectangle algorithms. This high degree of automation not only saves time but also enhances precision, offering significant advancements in heritage conservation.
The implications of this research extend beyond cultural heritage into the broader construction and energy sectors. For instance, automated deflection analysis could streamline inspections in commercial buildings, ensuring structural integrity and safety. In the energy sector, this technology could be adapted to monitor the structural health of renewable energy infrastructure, such as wind turbines and solar panel arrays, where timely diagnostics are crucial for maintaining efficiency and safety.
Sanz-Honrado’s work, published in the journal ‘Informes de la Construccion’—translated to ‘Construction Reports’—highlights the potential of combining advanced technologies for innovative solutions. As the construction industry increasingly embraces digital transformation, this research paves the way for more efficient, accurate, and automated diagnostic tools.
The future of heritage conservation and structural diagnostics looks promising with these advancements. As Sanz-Honrado puts it, “This is just the beginning. The possibilities are vast, and we are excited to explore how this technology can be further applied to preserve our cultural heritage and enhance structural safety across various industries.”

