PhD student wins ASCE Structural Health Monitoring and Control Committee student paper competition

By Peter Murphy

Published June 25, 2020

The Engineering Mechanics Institute (EMI) Structural Health Monitoring and Control (SHMC) Committee, featuring prominent researchers in the area of SHMC, held its annual paper competition last month. Seyed Omid Sajedi, a civil engineering PhD candidate earned first place. 

Human intervention working with AI

“This research paves the way for a reliable implementation of deep learning models in vibration-based damage diagnosis and making risk-informed decisions. ”
Xiao Liang, Assistant professor of research
Department of Civil, Structural and Environmental Engineering

“The recent advances in artificial intelligence (AI) have greatly impacted structural health monitoring,” Sajedi says, “I am a big fan of AI, and am confident in its great potential for the future. However, we should not forget that such data-driven models can make mistakes.”

In his paper, Risk-informed Semantic Damage Segmentation for Large-scale Vibration-based SHM, Sajedi describes an alarm system in AI or deep learning frameworks to allow for human intervention if necessary. This system allows researchers to utilize the advancements in AI, but gives them the flexibility to step in and take control. 

“The decision-makers will know how much they can trust their automated damage diagnosis systems and therefore minimize the consequences of occasional mistakes,” Sajedi says, “I believe that the way we look at the problem and the solution we are offering facilitate the industrial integration of AI in structural health monitoring.”

Sajedi poses with bull statue in alumni arena.

Seyed Omid Sajedi

According to Sajedi’s advisor, assistant professor of research Xiao Liang, this research could provide major advantages to decision-makers in structural health monitoring. “This research paves the way for a reliable implementation of deep learning models in vibration-based damage diagnosis and making risk-informed decisions,” Liang says, “for example, decision-makers will receive warning signals for potentially inaccurate condition assessments.”

Sajedi and Liang are currently developing a method to incorporate this system into a building structure. According to Sajedi, the next phase of this research will focus on signal processing and 3D neural network models to diagnose damage in bridge structures. 

The American Society of Civil Engineers' (ASCE) EMI SHMC committee is made up of researchers in the area of structural health monitoring. This is a significant achievement for Sajedi according to Liang. “Winning this competition indicates the high quality of Omid’s research, which successfully gets his peers’ attention,” Liang says, “the feedback was very positive, and he was able to learn what others are currently working on.”