Machine Learning for Structural Health Monitoring: Bridging Research and Practice

AI Researcher

Hedral Inc.

Imagine if bridges and buildings could tell us when they’re damaged before disaster strikes. Structural health monitoring (SHM) systems promise exactly that—using sensors and artificial intelligence to detect problems early, saving lives and money. With today’s advanced cameras and vibration sensors, we can collect massive amounts of data about how structures behave. Machine learning excels at finding patterns in this data that humans might miss.

So why aren’t these systems everywhere? Despite a decade of exciting algorithmic breakthroughs in academia, real-world adoption remains frustratingly low. The gap between what works in the lab and what practitioners actually use continues to widen. This talk bridges both worlds, drawing from my academic research and industry experience.

We’ll explore cutting-edge machine learning approaches, including both vibration-based and vision-based methods, that show real promise for monitoring structural health. But we’ll also confront the practical barriers I’ve encountered in deployment: the scarcity of failure data for model training, lack of standardization across projects, unclear value propositions for asset owners, and models that struggle to transfer across structures. The future of SHM isn’t just about better algorithms; it’s about building systems that actually work in practice.

Dr. Kareem Eltouny is an AI and structural engineering researcher specializing in machine learning-driven structural health monitoring. A Fulbright Scholar, he earned his PhD (2024) and MS (2019) in Civil Engineering from the University at Buffalo and his BSc (2014) from Suez Canal University.

Dr. Eltouny previously worked at Simpson Gumpertz & Heger, where he conducted applied research and developed machine learning-based monitoring tools for critical infrastructure, including nuclear facilities, offshore platforms, rail systems, defense infrastructure, and cable-stayed bridges. His research spans vibration- and vision-based monitoring, deep learning, and robotics, and has been funded by the NSF, USDOT (PHMSA), TCRP, and OESI. He has published in leading journals, including Computer-Aided Civil and Infrastructure Engineering, IEEE Robotics and Automation Letters, and Earthquake Engineering & Structural Dynamics, and received first place in the 2021 International Competition for Structural Health Monitoring. 

Dr. Eltouny is currently an AI Researcher at Hedral Inc., working to bridge AI innovation and practical deployment in the built environment.