[2F4] Distinguished overview speaker: Towards effective machine health monitoring
J Bortman
Despite the growing adoption of Industry 4.0 technologies, a significant gap remains in the full implementation of condition-based maintenance (CBM) across critical industries. The Predictive and Health Management (PHM) Laboratory at Ben-Gurion University is focused on closing this gap by developing advanced diagnostic and prognostic methods for mechanical systems such as bearings, gears, shafts and hydrodynamic components.
This presentation highlights recent innovations, including the development of analytical models for gear and bearing monitoring, the application of fibre Bragg grating (FBG) sensors and the integration of oil debris monitoring (ODM) systems. A novel spall damage simulation mechanism, developed in collaboration with the US Air Force Research Laboratory (AFRL), is also introduced.
A central innovation featured in this work is Camera-as-a-Sensor, a compact visual sensing system embedded with artificial intelligence (AI) models, capable of operating in harsh and inaccessible environments. This system has been validated through successful deployments by NASA, US government agencies and various defence and industrial programmes.
AI plays a key role in modern PHM systems. In addition to surveying state-of-the-art AI techniques, this work presents a new hybrid physical-AI algorithm designed for real-time damage tracking in mechanical components.
The presentation also explores the digital twin (DT) paradigm, emphasising its value in predictive maintenance. A recent case study involving railway systems demonstrates DT use to iteratively validate CBM strategies through synchronised physical and digital experimentation.
Together, these technologies represent a significant step towards realising a scalable, accurate and fully integrated CBM ecosystem.
This presentation highlights recent innovations, including the development of analytical models for gear and bearing monitoring, the application of fibre Bragg grating (FBG) sensors and the integration of oil debris monitoring (ODM) systems. A novel spall damage simulation mechanism, developed in collaboration with the US Air Force Research Laboratory (AFRL), is also introduced.
A central innovation featured in this work is Camera-as-a-Sensor, a compact visual sensing system embedded with artificial intelligence (AI) models, capable of operating in harsh and inaccessible environments. This system has been validated through successful deployments by NASA, US government agencies and various defence and industrial programmes.
AI plays a key role in modern PHM systems. In addition to surveying state-of-the-art AI techniques, this work presents a new hybrid physical-AI algorithm designed for real-time damage tracking in mechanical components.
The presentation also explores the digital twin (DT) paradigm, emphasising its value in predictive maintenance. A recent case study involving railway systems demonstrates DT use to iteratively validate CBM strategies through synchronised physical and digital experimentation.
Together, these technologies represent a significant step towards realising a scalable, accurate and fully integrated CBM ecosystem.
