A new automated scanning system can detect hidden defects in composite materials during manufacturing—in under two minutes, using low-cost cameras mounted on robot arms. Composite materials—layered carbon-fibre sheets used in aircraft, cars, and medical devices—are notoriously difficult to inspect without slowing production. Current non-destructive testing methods are expensive, require specialist operators, and often catch flaws too late, leading to wasted material and energy. The Bullfrog project combines uncooled infrared cameras, collaborative robot arms, and a deep-learning AI classifier to spot subsurface defects with over 90% accuracy during the layup stage itself. This early detection could cut product and environmental waste by up to 40%, directly supporting UK net-zero targets and aerospace efficiency programmes. If successful, the system could transform high-rate manufacturing across aerospace, automotive, defence, and electronics sectors—reducing supply-chain bottlenecks and lowering production costs. Because the solution is modular and automated, it can scale even where NDT experts are scarce. The project is a collaboration between Easol, Brookhouse Aerospace, the University of Southampton, Cranfield University, and Createc, with backing from Boeing, BAE, and the National Composites Centre.
View original technical description
The Bullfrog project pioneers a long-awaited step transformation for composites manufacturing Non-Destructive Testing (NDT); a fully automated, low-cost, non-contact in-process NDT solution that seamlessly integrates into manufacturing lines in a non-intrusive manner at critical junctures early in the composites manufacturing process. By identifying defects as early as even the layup stage, Bullfrog is poised to significantly improve composite manufacturing yield rates and rewrite industry standards, reducing product and environmental waste by up to 40% or more -- a huge leap forward for UK Net Zero initiatives and ATI's Destination Zero program. It is designed with aerospace industry needs in mind and seeks to provide a holistic solution to aerospace composite manufacturers of today. The core innovation of Bullfrog lies in its unique sequencing and optimisation of multiple state-of-the-art Active Thermography (AT) signal-processing techniques to enable low-cost Uncooled Infrared (IR) Cameras to provide Signal-to-Noise Ratios (SNR) near equivalent to Cooled IR Cameras at a fraction of the price - with none of the traditional drawbacks involved in cooling - thereby unlocking the full potential of Uncooled IR Cameras for composite NDT. Preliminary research and initial testing of the technology have already shown promising results in defect detection for aerospace-grade CFRP panels, with subsurface defect identification accuracy of over 90%, which the project looks to further refine. These Uncooled IR Cameras are rigged on portable, lightweight cobot (collaborative robot) arms with portable excitation sources. The cobot integrates machine vision to facilitate a fully automated scanning process - this combined with a deep learning AI classifier for automatic categorisation makes Bullfrog an independent modular solution that can truly scale to sector despite industry deficits of NDT experts. The entire scanning and analysis process is projected to take less than 2 minutes. Bullfrog's contribution to high-rate composite manufacturing is projected to make waves in solving supply chain shortages across the Aerospace industry and drive down costs, supporting efficiency-driven aircraft conversions, retrofits, and Aerospace R&D. This impact potentially extends far beyond Aerospace -- the Automotive, Defence, Medical, Electronics sectors and more will similarly enjoy the effects of higher yield, lower waste composite manufacture -- with the UK as its debut stage, then the world. The Bullfrog project is a collaboration between Easol, Brookhouse Aerospace, University of Southampton, Cranfield University, and Createc, and has support from major aerospace primes (Boeing, BAE), the National Composites Centre (NCC), other composite manufacturers (GTR).
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know