CANCELADA

Unfortunately, due to the war between Israel and Iran, Prof. Bortman is forced to postpone his trip to Brazil. Right now there are no flights to and from Israel. 

If the situation will change in the coming few days, he will reschedule his trip and we will let you know. 

We hope for better times to come. 

Thank you for your understanding,

Tami Matus

Research Administrative Support

Office of Vice President and Dean for R&D

Ben-Gurion University of the Negev

******************************************

Palestrante: Prof Jacob Bortman, Ben Gurion University, Negev.

Dia 30.06.2025 (segunda-feira), 14:30, auditório do CEPER, sala S-05, Fonseca Telles.

O Prof Bortman é full-professor da Ben-Gurion University, Negev, Diretor do laboratório "Prognostic Health Monitoring" PHM e autor de mais de 80 artigos em periódicos e congressos internacionais.

New Practical Approaches for Modern PHM

Prof.  Jacob Bortman,  PHM  Laboratory,  Department  of  Mechanical  Engineering,  Ben-Gurion University of the Negev, P.O. Box 653, Beer-Sheva 8410501, Israel.   

Abstract   

Prognostic health monitoring (PHM) by vibration signal analysis for rotating machinery is widely used for condition-based maintenance. Currently, machine learning, physical-based and signal processing algorithms cannot be implemented successfully for severity estimation and remaining useful life (RUL) estimation of rotating components of airspace vehicles (like airplane, helicopters etc.), where very few if any examples of faults can be used during the development phase of the algorithms.   

PHM  by  vibration  signals  can  be  implemented  in  four  stages:  fault  detection,  fault  location estimation, fault severity estimation, and RUL estimation. In the first stage, the algorithm tries to detect if there is a fault in the rotating components. In the second stage, it tries to localize the origin of the fault, for example, whether it is in the inner or outer race of the bearing. In the third stage, the algorithm estimates the fault severity, for example, by estimating its geometries. In the last stage, it estimates the RUL of the rotating components.

Several physical and signal processing algorithms have been developed over the last decades and can currently be used to address the two first stages. However, these approaches are very limited for fault severity estimation and, hence, also for RUL estimation.

In the presentation new approaches to overcome these difficulties will be presented:

a.  New algorithms based on validated physics-based models

b.  Hybrid AI modeling approaches 

c.  New sensing approaches based on FBGs and small embedded cameras

d.  Digital twin used for CBM

In  the  presentation  practical examples  from the  aeronautical and transportation industries  will  be presented