In a significant stride towards trustworthy artificial intelligence in engineering, a team of researchers led by Doniyor Mukhtorov from the Department of Computer Engineering at Gachon University in South Korea has developed a platform called AIVeritas. This platform is designed to validate, explain, and certify AI models, ensuring they meet international standards. The research, published in the IEEE Access journal, addresses a critical gap in the systematic validation and certification of AI models, particularly in engineering fields.
AIVeritas integrates four key components: data-quality analysis, model-performance evaluation, explainability diagnostics, and governance-based certification. These components work together to provide measurable, auditable, and reproducible assessments of AI systems. The platform aligns with several ISO/IEC standards, including ISO/IEC 24028, 23053, 25023, 4213, and 17065, ensuring that AI models are reliable, traceable, transparent, and ready for certification.
Mukhtorov emphasized the importance of each module in maintaining the trustworthiness of AI systems. “The elimination of any module led to a decrease in trustworthiness scores,” he noted, highlighting the indispensable nature of each component. The platform’s rigorous verification capabilities, including clause-level standards mapping, trust-index computation, lifecycle traceability, and explanation-stability analysis, make it a robust tool for assessing AI models in various engineering contexts.
For the maritime sector, the implications are substantial. AI models are increasingly used for vessel performance prediction, route optimization, and operational decision-making. However, the lack of a systematic framework for validation and certification has been a significant hurdle. AIVeritas offers a unified and regulation-ready pathway for assessing the validity, robustness, and explainability of AI models, thereby enhancing trust and reliability in maritime AI applications.
The commercial impacts are equally promising. By ensuring that AI models comply with international standards, AIVeritas can facilitate the adoption of AI technologies in the maritime industry. This can lead to improved operational efficiency, reduced costs, and enhanced safety. Moreover, the platform’s ability to provide measurable and auditable assessments can help maritime companies meet regulatory requirements and gain a competitive edge.
In summary, AIVeritas represents a significant advancement in the field of AI validation and certification. Its alignment with international standards and comprehensive assessment capabilities make it a valuable tool for the maritime sector. As AI continues to play a pivotal role in engineering and operational decision-making, platforms like AIVeritas will be instrumental in ensuring the trustworthiness and reliability of AI models.