Engineers advance sensor technology for next generation machine wear sensing

Researchers from the University of Southampton have been awarded over £1m to study the early detection of contact distress for enhanced performance monitoring and predictive inspection of machines such as planes and cars.

A multi-disciplinary team, led within the School of Engineering, will miniaturise existing sensing technology as it applies the latest machine learning techniques to the field of tribology, the science and engineering of interacting surfaces in relative motion.

The project, which is a partnership with General Electric Company, Schaeffler KG, Senseye and Shell, aims to shed light on the fundamental principles of early wear. Researchers will focus on developing and investigating how well electrostatic micro-sensing arrays with embedded electronics detect tribological transitions related to wear and friction in machine component contacts.

Professor Robert Wood, Principal Investigator, says: ‘This grant will allow us to build on 20 years of research at Southampton into electrostatic based condition monitoring of tribological contacts to allow far better temporal and spatial resolution and thus earlier detection of distress.”

The four-year Engineering and Physical Sciences Research Council project will draw upon expertise from Dr Terry Harvey from the National Centre for Advanced Tribology at Southampton (nCATS) together with Professor Mahesan Niranjan and Dr Nick Harris from the School of Electronics and Computer Science (ECS). The grant will build on existing collaborations from two Royal Academy of Engineering Visiting Professors to nCATS from Schaeffler and GE Aviation.

Dr Harvey explains, “We have been working on electrostatic sensor for many years now, delving into the fundamental of what of the sensors are seeing but all at the macro-scale.  This project will allow us to push the technology forward by developing arrays of micro-electrostatic sensors that will us detect charge at far higher resolutions than previously possible but also effectively map it. This will mean that the amount of information being processed is multiplied and this is where machine learning will play a vital part in the project.”

Researchers identified that the combination of data from arrays of embedded sensors close to the wear surface, being developed within ECS, and the application of advanced machine learning techniques to the resulting data streams offer the promise of more accurate and earlier prognostic data for machines despite the variation of use cases that may be seen.

“New sensors will be the link that enables the practical application of machine learning to mechanical systems, which will unlock much more than just new information on wear,” Dr Harris explains. “The combination of experience from nCATS and ECS, together with the industrial partners, will allow new approaches to condition monitoring and predictive maintenance across a broad spectrum of industry.”

Professor Niranjan says, “The subject of machine learning is of increasing importance in a wide range of multi-disciplinary problems involving large and complex datasets. Here in Southampton, we are proud of our research in the subject, as well as its integration into our taught programmes at undergraduate and masters’ levels. In this project, I am particularly excited about using modern machine learning algorithms combined with domain knowledge and data gathered from specifically designed instruments and experiments.”