Background
When an accident occurs at sea involving one or more vessels, the cause of the accident will in 4 out of 5 cases be attributed to the crew on the bridge. The type of causal explanation where the captain's decisions led to the accident is a simplification of reality. In the report "Safety Analysis of EMCIP Data" from EMSA (2022), the analyzes show that what we immediately address as "human errors" is the result of a far more complex interaction between behavior and technology.
A modern ship's bridge is equipped with a large number of technical instruments with a limited degree of integration. The disadvantage of the limited integration is that the cognitive load on the bridge crew can lead to reduced "situational awareness" and misinterpretations in the risk assessment of a situation. Therefore, there is a need for a new generation of decision-support systems that can perform bridge data integration "behind the scenes".
The integrated information is made available so that it provides decision-supporting data for the bridge crew and supports safe navigation. The new wave of technological development related to the implementation of the autonomous ship has the potential to deliver on that promise. However, the current technological solutions are still in a developmental stage, where the systems are designed to comply with basic parts of the COLREGs, which do not include "good seamanship", which is central.
Project
AI-NAVIGATOR develops the first decision support system that can analyze a navigation scenario based on both COLREGs and a data-driven modeling of "good seamanship" which provides navigation suggestions that reduce the risk of accident.
Galeazzi describes good seamanship as “the ability of human navigators to resolve navigation scenarios within and beyond the COLREGs regulatory framework by using navigation best practices”.
AI-NAVIGATOR uses methods within artificial intelligence to develop a computational model of "good seamanship" by using previously unanalyzed data from maritime academies' simulators. Output is the development of a structural model which increases the bridge crew's "situational awareness" and improves risk assessment in relation to collision.
Expected results
- A validated intelligent decision support system (navigation), which is presented to relevant stakeholders.
- Publication of a research dataset of the "good seamanship" model.
- Design of an evaluation procedure for assessing the level of "good seamanship" targeting the decision support system AI-NAVIGATOR.
The project is carried out in collaboration with Svendborg International Maritime Academy.
Facts
As a commercial fund The Danish Maritime Fund works with an overall objective to engage via grants and funding in activities that provide growth and development of the Danish maritime sector. Thereby indirectly supporting job creation in Denmark. They are engaged in four strategic focus areas, i.e. Research, Promotion of the Danish Maritime Sector (The Blue Denmark), Early Stage Funding of Innovational Commercial Projects (typically start-ups), and lastly Competence Building of Employees of The Blue Denmark. The overall themes in all they do are hovering around digitization, automation, green transition and thereby inherently building on the UN 17 Sustainable Development Goals.
