jump to content

NRN AEM research into non-destructive testing wins top prize at international conference



Dr Abhishek Kundu from Cardiff University is celebrating after securing best conference paper at this year’s European Workshop on Structural Health and Monitoring Series (EWSHM) which took place last month.  


The NRN funded research aims to address the challenging issue of detection and characterisation of miniscule, often barely visible, damages in complex composite structures within the aerospace industry. The paper, 'Probabilistic method for damage identification in multi-layered composite structures'  having been chosen for its innovative and advanced research in non-destructive testing highlights the complex damages that could, if left, severely jeopardies the operational safety of these real life applications.


As aerospace systems become smarter and more interconnected due to revolutionary frameworks such as the Internet of Things (IoT), the industry is becoming more and more reliant on innovative research and resources to improve its operational safety and services.


Using the latest ultrasonic acoustipublicize02.pngc emission (AE) technology with a distributed network of on-board sensors, Dr Kundu and team detected miniscule damages (such as delamination, fibre breakage) sustained in these intricate structures using machine learning algorithms and advanced signal processing techniques. With this the team have been able to create a data-driven map of vital sensor signal features to the localized damage characteristics, which is ideal for deployment in real-time monitoring environments in safety-critical operations with minimal human intervention.


“The next generation of smart structures are being conceived within an IOT (internet of things) framework with the aim of structural models being driven by model and real-time operating data and build with self-monitoring, controlling and possibly mitigating technology.

This paper proposes an intelligent data-driven damage identification methodology which can be used as a blackbox detection algorithm coupled with on-board sensors for smart composite structures. This technology is slim enough to be deployed in real-time monitoring environments. The detection scheme can potentially be updated at regular intervals with real-time operational data so as to account for structural degradation and minimize the reliance on a baseline pristine state of the operating structures.” Says Dr Kundu.

As technology within the field of computational modelling advances, along with rapid evolution of high-performance digital signal processing platforms, it is possible to embed such intelligent monitoring systems in real life smart structures. 

This not only lowers safety risks but has the potential to economically benefit industry.


“The economic benefit coming from the technology is evident in the fact that expensive and time-consuming interventionist manual inspection of structures would be minimised which will also ensure minimal out-of-operation structural down-time.”


 “The challenge of having a reliable, predictive model for damage characterization in an operating environment that is rife with uncertainty has been addressed in this paper; it provides excellent accuracy under different load types. The efficient, data-driven, framework for damage identification is ideal for deployment in real-time monitoring environments in safety-critical operations with minimal human intervention” explains Dr Abhishek Kundu


Dr Abhishek Kundu who previously secured an NRN AEM industrial fellowship with Airbus to further his research into early-stage design and optimisation of aircraft has said how this industrial fellowship contributed extensively to the success of the research that has been presented in the winning paper.


“The NRN Sêr Cymru’ s 18 month industrial fellowship award for working with Airbus on early-stage aircraft design under uncertainty has contributed directly to the work produced in this paper. The research on Bayesian inverse problems helped in the creation of a set of toolkits on the probabilistic damage detection framework which were used for this research.


The fellowship supported participation at meetings as well as networking events which helped in building important academic links in addition to key connections with industrial contacts interested in next-gen structural health monitoring technology which gave valuable input to this project.


Following on from this a full research grant proposal is now in the pipeline based on the NRN research reported at the EWSHM conference.”







News Image
News Image 2
News Image 3