Deep Learning Robotic System for Aeroengine Blade Surface Inspection

Title

Deep Learning Robotic System for Aeroengine Blade Surface Inspection

Content

Researchers at the Advanced Research and Innovation Center (ARIC) at Khalifa University of Science and Technology in Abu Dhabi developed and deployed a deep-learning-enabled intelligent robotic system for aeroengine blade (AEB) surface inspection. The system integrates vision-based deep learning with robotic automation to autonomously localise, pick, image, inspect, and return blades in real time. Yusra Abdulrahman, corresponding author and researcher at ARIC and the Department of Aerospace Engineering at Khalifa University, led the work. The team trained separate robust deep learning models on two datasets — one for blade localisation and one for surface defect detection — then integrated both into the robotic system. The system provides real-time feedback and accelerates decision-making compared to conventional manual or borescope-based inspection, which is time-consuming, labor-intensive, and susceptible to human error. Experimental results demonstrate that the proposed approach achieves a mean Average Precision (mAP) of 88.2% and reduces the inspection cycle time to approximately 4 seconds per blade, significantly improving efficiency over conventional methods. The system is designed for controlled industrial inspection environments in modern aerospace MRO (maintenance, repair, and overhaul) processes, overcoming limitations of manual inspection. The deployment target is aerospace MRO facilities where aeroengine blades are critical components requiring consistent monitoring to ensure airworthiness and operational safety. Traditional manual or borescope-based inspection methods rely on highly trained technicians and produce inconsistent results. The robotic system addresses key MRO requirements with autonomous operation. The research was funded by the Advanced Research and Innovation Center (ARIC) at Khalifa University, jointly funded by Aerospace Holding Company LLC (a wholly owned subsidiary of Mubadala Investment Company PJSC) and Khalifa University, with additional support under Award No. 8474000660. The work was published in Scientific Reports on 16 June 2026 (received 30 January 2026, accepted 9 June 2026). Type: Research prototype. Note: this is a research deployment at Khalifa University, not a commercial enterprise deployment. The current company record 'Aeroengine Blade Inspection Research' is generic; the actual institution is Khalifa University of Science and Technology (ARIC).