Marine Collision Avoidance Using Historical Navigation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current collision avoidance systems for marine vessels, especially autonomous ones, lack reliable alternatives to traditional manned vessel navigation, as they struggle to accurately predict and respond to collision risks in dynamic maritime environments, relying heavily on human expertise and rules like COLREGS which are not fail-safe in all situations.
Innovation Solution
A computer-implemented method and system that utilizes real-time data from multiple vessels to identify collision risks, employing historical navigation data to determine and provide collision avoidance maneuvers, incorporating vessel-centric coordinates, normalized with respect to vessel manoeuvrability, and using models like artificial neural networks or polynomial expressions to assess and mitigate risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional collision avoidance systems rely on COLREGS rules and human expertise, then they can provide operational guidance, but they cannot be absolutely fail-safe and cannot precisely prescribe responses to all situations
Solution Approach 1:
The system creates a virtual copy of the maritime environment by mapping real-world vessel positions, courses, and speeds into a simulated collision scenario. This virtual representation allows the AI model to analyze and predict collision risks without physical trial and error, enabling reliable decision-making across diverse situations.
Solution Approach 2:
The collision avoidance system performs preliminary risk assessment and maneuver prediction before actual collision threats materialize. By continuously analyzing real-time vessel data and predicting potential collision scenarios in advance, the system can prepare and recommend avoidance maneuvers proactively rather than reactively.
2Productivity
If autonomous vessels are deployed with limited or zero crew, then operational costs are reduced, but reliable alternatives to human navigation expertise are required
Solution Approach 1:
The system replaces human navigational expertise with an AI-based decision-making model that processes vessel data and predicts collision risks algorithmically. This substitution eliminates the need for human crew while maintaining or improving navigation reliability through consistent, data-driven decision-making.
Solution Approach 2:
The collision avoidance system enables vessels to autonomously assess their own collision risks and generate avoidance maneuvers without external human intervention. The AI model processes the vessel's own navigation data and that of surrounding vessels to self-determine appropriate corrective actions.
3Measurement precision
If real-time data from multiple vessels is processed to identify collision risks, then collision detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The system extracts only the essential collision risk parameters from the raw real-time data of multiple vessels, such as relative positions, courses, and speeds. By focusing on these critical variables rather than processing all available data, the system maintains high detection accuracy while reducing computational complexity.
Solution Approach 2:
The system creates a simplified virtual model of the maritime environment that replicates only the necessary elements for collision risk assessment. This virtual copy allows complex multi-vessel interactions to be analyzed through streamlined computational representations rather than full-scale simulations.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
Described is a collision avoidance method for a water-based vessel, comprising: obtaining real time data relating to the path of two or more vessels; identifying a collision risk between the two or more vessels; determining if the collision risk is above a predetermined threshold, wherein when the collision risk is above the predetermined threshold, determining one or more collision avoidance manoeuvres on the basis of historical navigational data which corresponds to the real-time data; providing the one or more collision avoidance manoeuvres to an operator of the one or more vessels.