Case Study: Airbus achieves automated pipe defect detection and rapid deep‑learning deployment with MathWorks (MATLAB)

Airbus Uses Artificial Intelligence and Deep Learning for Automatic Defect Detection

Airbus faced the challenge of building a robust end-to-end AI system to automatically detect defects in aircraft pipes, including locating ventilation holes and wires to measure distances and angles to industry standards. To prototype and develop deep learning models quickly, Airbus worked with MathWorks and used MATLAB (with support from MathWorks Consulting Services) for interactive labeling, model design, and training.

MathWorks provided an integrated MATLAB-based workflow for semantic segmentation, real-time defect visualization, and automatic translation of MATLAB code to CUDA for direct embedded deployment. The solution enabled rapid prototyping and testing, delivered a 50× boost in deep learning throughput over three years, and achieved fast, deployable defect detection—demonstrating clear operational and development speed improvements from MathWorks’ tools and services.

Airbus

Nicolas Castet

Airbus