COLREGs Compliant Collision Avoidance and Grounding Prevention for Autonomous Marine Navigation

Maritime Autonomous Surface Ships (MASS) are increasingly regarded as a promising solution to address crew shortages, improve navigational safety, and improve operational efficiency in the maritime industry. Nevertheless, the reliable deployment of MASS in real-world environments remains a significant challenge, particularly in congested waters where the majority of maritime accidents occur. This emphasizes the need for safe and regulation-aware motion planning strategies for MASS that are capable of operating under dynamic maritime conditions. This paper presents a unified motion planning method for MASS that achieves real time collision avoidance, compliance with International Regulations for Preventing Collisions at Sea (COLREGs), and grounding prevention. The proposed work introduces a convex optimization method that integrates velocity obstacle-based (VO) collision constraints, COLREGs-based directional constraints, and bathymetry-based grounding constraints to generate computationally efficient, rule-compliant optimal velocity selection. To enhance robustness, the classical VO method is extended to consider uncertainty in the position and velocity estimates of the target vessel. Unnavigable shallow water regions obtained from bathymetric data, which are inherently nonconvex, are approximated via convex geometries using a integer linear programming (ILP), allowing grounding constraints to be incorporated into the motion planning. The resulting optimization generates optimal and dynamically feasible input velocities that meet collision avoidance, regulatory compliance, kinodynamic limits, and grounding prevention requirements. Simulation results involving multi-vessel encounters demonstrate the effectiveness of the proposed method in producing safe and regulation-compliant maneuvers, highlighting the suitability of the proposed approach for real time autonomous maritime navigation.

💡 Research Summary

The paper addresses a critical gap in autonomous surface ship (MASS) navigation by presenting a unified motion‑planning framework that simultaneously guarantees real‑time collision avoidance, compliance with the International Regulations for Preventing Collisions at Sea (COLREGs), and protection against grounding. The core of the approach is a convex quadratic program (QP) formulated in velocity space. Collision avoidance constraints are derived from the Velocity Obstacle (VO) concept, but the authors extend the classic VO to a robust version that explicitly incorporates bounded uncertainties in both the estimated position and velocity of surrounding vessels. Position uncertainty is modeled as an inflation of the obstacle’s geometry, while velocity uncertainty is treated as a Minkowski sum with a bounded set, resulting in an enlarged, conservative collision cone.

Regulatory compliance is enforced by translating COLREGs maneuvering rules into linear half‑plane constraints. For each encountered vessel, the VO boundary provides left‑ and right‑tangent lines; the side required by COLREGs (port‑side or starboard‑side) determines which half‑plane is retained. This yields a set of linear inequalities that guarantee the chosen velocity respects the prescribed maneuvering direction.

Grounding prevention is achieved by integrating bathymetric data. Since shallow‑water regions are typically non‑convex, the authors employ an integer linear programming (ILP) set‑cover formulation to approximate these areas with a small number of convex polygons. Each polygon is expressed as a linear inequality, allowing the grounding constraints to be incorporated seamlessly into the same QP.

Speed limits are approximated by an inscribed polygon to preserve linearity, and a reachable‑velocity set accounts for the vessel’s dynamic limits over a finite time step. The final QP minimizes the Euclidean distance between the feasible velocity and a reference velocity that represents the desired nominal motion, subject to all linear constraints. If the feasible set becomes empty, the algorithm safely reduces speed to zero. The commanded velocity is further limited by a maximum allowable acceleration per control cycle.

The authors validate the method through simulations based on IMA‑ZU encounter scenarios involving multiple vessels and shallow water. Results show that the robust VO reduces collision risk compared to a deterministic VO, the COLREGs constraints eliminate any rule violations, and the convexified bathymetry prevents grounding. The QP solves in under 10 ms on a standard workstation, demonstrating real‑time capability.

The paper contributes a computationally efficient, safety‑oriented planning scheme that respects maritime regulations and environmental constraints. Limitations include potential conservatism from large uncertainty bounds and the need for offline preprocessing of bathymetric convex approximations. Future work will explore probabilistic risk‑based cost functions, online updating of shallow‑water approximations for dynamic tides, and field trials on actual autonomous vessels.