Digital technologies are increasingly transforming the field of Non-Destructive Examination (NDE), enabling enhanced defect detection, data interpretation, inspection efficiency, and decision support. Machine Learning (ML) and Artificial Intelligence (AI) techniques offer significant opportunities to improve the consistency, reliability, and scalability of inspection processes across a wide range of industries, including marine and offshore.
The growing adoption of AI-enabled NDE systems introduces new considerations relating to data quality, model development, validation, explainability, cybersecurity, human oversight, and regulatory compliance. While AI tools can augment inspection capability and support informed decision-making, responsibility for inspection outcomes remains with qualified personnel and organisations operating within established engineering and quality assurance frameworks.
This Guidance Note provides recommended practices for the development, verification, validation, deployment, and governance of ML and AI technologies used in NDE applications. It outlines key principles for ensuring technical robustness, transparency, traceability, and confidence in AI-assisted inspection systems. The recommendations are intended to support industry stakeholders in achieving safe, effective, and responsible implementation of AI while maintaining compliance with applicable standards, codes, and certification requirements.
The guidance is advisory in nature and does not replace mandatory regulatory, certification, or qualification requirements. It is intended to complement existing NDE standards and industry best practices by providing a structured framework for the application of ML and AI technologies throughout the inspection lifecycle.