SMM Chair at the EMI 2023 conference

Kiran Bacsa, doctoral candidate at SEC/ETH Zurich, participates as a finalist of the EMI Dynamics Committee Student Paper competition, with his work on Structural Dynamics Learning using a Supervised Variational Auto-Encoder (SVAE), which is further presented as part of the regular conference program during the Friday, June 09, Afternoon Sessions, 14:15 - 14:35.

Kiran presents a novel approach using a Supervised Variational Auto-encoder (SVAE) to learn the structural dynamics of a monitored structure and classify its response. By training the SVAE on sensor measurements and auxiliary classification of structural parameters, a compact and expressive representation of the structure's dynamics is learned, which can be generalized across different stiffness categories. The SVAE outperforms other models in reconstructing unseen response measurements, making it a valuable tool for monitoring and controlling the health of dynamic structures. 

Wei Liu, doctoral candidate at SEC/NUS, has been selected as a finalist of the EMI Structural Control and Health Monitoring Committee Student Paper competition for his research on estimating unknown inputs in monitored engineered systems. The proposed method utilizes a Partially Observable Markov Decision Process (POMDP) approach to identify system inputs that replicate measured outputs, enabling the estimation of operational loads and system reactions. The method is demonstrated on a simulated vehicle and a real-world wind turbine, showcasing its potential for enhancing the safety and performance of structural health monitoring and informing structural system design.