Скачать или смотреть A Deep Learning-Based Pipeline for Structural Diagnostics and Prognostics

#MORPHOworkshops

Workshop: Advances in Structural Health Monitoring. Innovation and value creation in aeronautics
Expert speaker: Giannis Stamatelatos, PhD Student at University of Patras

The stochastic nature of composite structures complicates analytical RUL prediction models, necessitating data-driven approaches for real-time operations. Traditional Deep Learning methods require extensive preprocessing, manual feature engineering, and mathematical constraints, limiting their use. This study proposes a novel approach for real-time RUL estimation, leveraging sensor data and pretrained models to capture complex patterns automatically. Our streamlined method processes raw sensor data, extracting features without manual effort and adapting easily to new datasets. Testing identified 1D Inception Convolutions as highly effective, achieving good results even with limited data. This work advances predictive maintenance with efficient, real-time RUL estimation for composite components.

Don't forget to like, comment, and subscribe to our channel to stay updated with the latest videos and developments from the MORPHO project. Thank you for watching!

Stay conected: https://morpho-h2020.eu/

#MORPHOProject #SHM #Aeronautics #Innovation #Workshop