Tragedy narrowly averted: In the darkness of a clear winter night on November 8, 2018, the 133-meter-long frigate KNM “Helge Ingstad” was involved in a significant maritime incident. Taking part in a NATO exercise and headed towards Scotland, its journey in the Hjeltefjord north-west of Bergen was abruptly interrupted due to a violent collision with the “TS Sola”, an oil tanker roughly twice its size. With 137 crew aboard, the ship was virtually invisible on AIS signaling systems yet noticeable by radar to surrounding vessels.
The impact nearly severed 50 meters from the starboard side of the Helge Ingstad, leading to seawater inundating soldier’s bunks and live electric cables protruding hazardously. Miraculously, there were no fatalities, and only seven were injured as all were successfully evacuated. The tanker managed with minor abrasions, averting an environmental catastrophe.
Could AI have changed the course?: Prof. Ingrid Bouwer Utne, an authority in marine safety and risk at the Norwegian University of Science and Technology (NTNU), posits that if artificial intelligence had been deployed on ships like the Helge Ingstad, the understanding of on-board conditions could have been significantly enhanced. Enhanced AI systems could aid in bridging the gaps of misinterpretation and oversight by vessel captains, as seen in the Helge Ingstad incident where the oncoming oil tanker was not properly identified.
Advancing autonomous systems: Emphasizing integrated decision-making, Prof. Utne has been instrumental in research projects such as ORCAS and UNLOCK, which aim to embed a deep understanding of risk in autonomous military vessels, drones, and underwater robotics. These projects, funded by the Norwegian Research Council, strive to incorporate sophisticated control mechanisms that can evaluate and respond to risks akin to human reasoning.
Integrating risk analysis: At the heart of ORCAS and UNLOCK initiatives is the development of risk analytic models that are intertwined with algorithm design from the onset, contradicting traditional late-stage risk assessments. Such foresighted integration enables autonomous systems to make safer, more informed decisions and may signify a paradigm shift in how artificial intelligence is utilized for maritime safety. Prof. Utne acknowledges the current embryonic state of autonomy development but remains hopeful for its accelerating progress and its potential to prevent accidents similar to the unfortunate near-tragedy of KNM “Helge Ingstad”.
When discussing the potential for artificial intelligence (AI) to prevent naval collisions, it’s important to highlight a few key aspects:
Key Questions:
– Can AI systems effectively interpret complex maritime scenarios better than human crews?
– What are the limitations of current AI technology in dynamic and unpredictable environments, such as the open sea?
– How can AI-enhanced navigation systems improve maritime safety?
Challenges and Controversies:
– Reliability: AI systems must be extremely reliable to be trusted in life-threatening situations.
– Accountability: In the event of a failure, it can be challenging to determine accountability between the AI developers, operators, and the autonomous system itself.
– Ethical considerations: Decisions made by AI in critical moments may raise ethical questions, especially if they involve prioritizing actions that could impact human lives.
Advantages of AI in Avoiding Naval Collisions:
– Continuous Monitoring: AI systems can monitor and process data continuously without the fatigue that affects human operators.
– Advanced Data Analysis: AI can analyze vast amounts of data from different sources to identify risks and suggest evasive actions.
– Decision-making: Improved AI systems can make quick decisions based on complex risk assessments.
Disadvantages of AI in Naval Operations:
– Technological limitations: Current AI may not be able to handle unexpected variables that human experience can navigate.
– Dependency: Over-reliance on AI could lead to skill atrophy in human operators, potentially making them less effective in situations where AI fails or is unavailable.
For further reading on artificial intelligence and its applications, the following link to the primary domain (without subpages) provides a general overview that might be helpful: Norwegian University of Science and Technology.
Please note that this advice is based on general information about artificial intelligence and maritime safety and should be adapted to the specific details of the KNM “Helge Ingstad” incident and Prof. Ingrid Bouwer Utne’s research as necessary.