Vessel Navigation Rule Estimation for Transparent COLREG Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems fail to provide clear compliance with various navigation rules, such as COLREG, and lack transparency in decision-making processes, posing safety and regulatory compliance risks in automated vessel navigation.
Innovation Solution
A navigation assistance system that includes a rule estimation logic storage, own ship and target data acquirers, situational awareness data generator, relative relationship specifier, and rule estimator to determine applicable rules based on the relative relationship between the own ship and targets, ensuring compliance with navigation rules and providing transparent decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated navigation systems are implemented to reduce seafarer workload, then productivity is improved, but reliability deteriorates due to lack of transparency in decision-making processes
Solution Approach 1:
The system implements feedback by continuously monitoring navigation situations and providing explanations for automated decisions. The explanation generator creates transparent feedback loops that show seafarers why specific navigation decisions are made, maintaining trust and reliability while enabling automated operation.
Solution Approach 2:
The explanation generator acts as an intermediary between the automated navigation system and human operators. It translates complex AI decision-making processes into understandable explanations, bridging the gap between automated efficiency and human comprehension of system behavior.
2Productivity
If conventional collision avoidance routes are generated using robotics technology, then productivity is improved, but reliability deteriorates due to non-compliance with COLREG rules
Solution Approach 1:
The system changes the parameters of collision avoidance by integrating COLREG-specific rules into the path planning algorithm. Instead of using generic robotics avoidance parameters, the system adjusts navigation parameters to comply with maritime regulations, ensuring both efficiency and legal compliance.
Solution Approach 2:
The navigation decision-making process is segmented into distinct components: situation assessment, rule application, and route generation. This segmentation allows the system to apply specific COLREG rules to different situational contexts, ensuring comprehensive rule compliance while maintaining operational efficiency.
3Productivity
If AI is used for decision-making from target detection to collision avoidance, then productivity is improved, but reliability deteriorates due to black box nature of AI processes
Solution Approach 1:
The explanation generator provides continuous feedback about AI decision-making processes, transforming the black box into a transparent system. It explains target detection decisions, collision avoidance choices, and rule applications, maintaining reliability through explainability while preserving automation benefits.
4Device complexity
If navigation systems do not provide clear compliance information, then device complexity is reduced, but reliability deteriorates due to safety and regulatory compliance risks
Solution Approach 1:
The explanation generator serves as an intermediary layer that adds compliance information without fundamentally changing the core navigation system. It translates internal system decisions into compliance-certifiable explanations, maintaining simple system structure while ensuring regulatory reliability.
Data Source
AI summary
A navigation assistance system includes a rule estimation logic storage storing a rule estimation logic, an own ship setting data storage, an own ship data acquirer, a target data acquirer, a surrounding data acquirer, a situational awareness data generator generating situational awareness data, based on the own ship data, the target data, and the surrounding data, a relative relationship specifier specifying a relative relationship between the own ship and the target, based on the situational awareness data, and a rule estimator estimating an applicable rule to be applied to the own ship, based on the relative relationship and the own ship setting data, in accordance with the rule estimation logic.