We are proud to share that AIS Field successfully presented its latest research at the British Institute of Non-Destructive Testing (BINDT) 2024 Conference, held in Telford, United Kingdom. The annual BINDT conference is a respected global event in the field of non-destructive testing and condition monitoring, bringing together industry experts, academics and innovators to exchange knowledge and explore the latest technological advancements.

Our presentation, titled “Development and Comparison of RNN, LSTM and GRU Neural Network Models for Automated Eddy Current Inspection of Heat Exchanger Tubes Using Real-World Industrial Data,” marks a significant milestone in our ongoing efforts to integrate artificial intelligence and machine learning into non-destructive testing (NDT) and robotic inspection systems.

In this study, our team explored the use of advanced deep learning architectures to automate the analysis of eddy current inspection data collected from industrial heat exchangers. Using real-world data from refineries and power plants, we developed and benchmarked multiple neural network models, including RNN, LSTM and GRU, to evaluate their performance in detecting and classifying defects in non-ferrous tube materials. The results demonstrated that AI-driven systems can achieve high accuracy, consistency and speed, offering a promising step toward reducing human error and improving overall inspection efficiency.

We extend our sincere appreciation to Metin Oktay Balaban, our software lead and machine learning engineer, for his exceptional work and leadership throughout this research. We would also like to thank Sait Kökoğlu from Integrity NDT for his valuable technical support, which played a key role in making this study possible.

Participating in BINDT 2024 has been an inspiring experience for our team, and we are excited to continue advancing the frontiers of AI-driven inspection technologies.

Stay tuned for more updates from AIS Field’s R&D initiatives.