Reliability assessment of autonomous maritime navigation systems under uncertainty

Maintaining the reliability of autonomous maritime navigation systems (AMNS) is of utmost importance for the safety, efficiency, and sustainability of ocean and offshore operations. Conventional reliability analysis approaches tend to be overwhelmed by maritime environments' complexity and uncertainty. The Fuzzy ELECTRE method is employed to evaluate and rank different autonomous navigation systems (ANS) based on key reliability factors. Five navigation system alternatives conventional GPS-based navigation, inertial navigation systems (INS), integrated GPS/INS hybrid systems, LiDAR-based navigation, and AI-driven predictive navigation systems are considered. The evaluation is performed using seven reliability-based criteria: fault detection and diagnosis capability, human error resilience, system downtime and maintainability, uncertainty handling, redundancy mechanisms, energy efficiency, and compliance with safety-critical regulations. The results provide a ranked assessment of the most reliable ANSs, offering insights into their strengths and limitations in real-world maritime applications. The results of this study support the progress of reliability analysis in ocean and maritime engineering by presenting a structured decision-making framework that enhances the selection process of autonomous navigation technologies. The proposed methodology can assist maritime engineers, policymakers, and industry stakeholders in identifying the most robust and fault-resilient navigation solutions for future autonomous maritime operations.