Advancements in AI-Driven Navigation and Collision Avoidance Systems for Maritime Applications

The maritime industry is increasingly leveraging artificial intelligence (AI) to enhance navigation safety and mitigate collision risks at sea. This paper provides a comprehensive analysis of recent advancements in AI-driven navigation and collision avoidance systems for Maritime Autonomous Surface Ships (MASS) and conventional vessels. We extend prior work by introducing detailed mathematical modeling, numeric substitution, and MATLAB-based simulations. The study integrates collision risk assessment via Closest Point of Approach (CPA) and Time to CPA (𝑇𝐶𝑃𝐴), trajectory optimization with Model Predictive Control [MPC], Artificial Potential Fields (APF), and cybersecurity anomaly detection using Kalman filters. Results demonstrate that CPA analysis identified a critical near-collision (𝐷𝐶𝑃𝐴 = 8.84 m, 𝑇𝐶𝑃𝐴 = 11.22 s), MPC maintained safe separations > 16 m, APF successfully guided vessels to targets, and spoofed AIS signals were detected with > 95% accuracy. Figures generated from MATLAB simulations illustrate each methodology. This extended work contributes by merging theoretical models with practical simulations, highlighting both technological promise and cybersecurity challenges.