The 8th International Conference on System Reliability and Safety
Sicily, Italy - November 20-22, 2024

Masoud Naseri, The Arctic University of Norway, Norway

 

Dr. Masoud Naseri is currently an Associate Professor of Societal Safety and Security at UiT – The Arctic University of Norway. Naseri has received his bachelor’s and master’s degrees in petroleum engineering and his PhD in Science specialised on RAM modelling and analysis of Arctic offshore oil and gas platforms from UiT in 2016. During his PhD, Naseri has focused on risk and reliability modelling and analysis of offshore oil and gas platforms operating under the severe and dynamic environmental conditions of the Arctic offshore, and how expert judgements can be employed for risk and reliability assessment in such demanding environments where the historical reliability data is scarce. Naseri’s current research revolves around modelling and analysis of Arctic offshore meteorological phenomena, such as vessel icing, and assessing the risks associated with maritime activities and shipping operations in the Arctic waters for supporting informed decisions with respect to icing storms and harsh weather conditions. From January 2025, Naseri will join the group Laboratory of Analysis of Systems for the Assessment of their Reliability, Risk and Resilience, LASARˆ3 (www.lasar.polimi.it) at the Department of Energy of Politecnico di Milano in Italy as recipient of a Marie Skłodowska-Curie Actions Postdoctoral Fellowship (MSCA-PF) to perform research on modelling and optimising the resilience of integrated natural gas network systems.

Speech Title: A Markov Decision Process Model for Solving the Sequential Decision Problem of Fishing Vessel Route Planning in Arctic Waters

Maritime operations in the Barents Sea, particularly pelagic fishing, face numerous safety-related challenges due to the extreme weather and sea conditions prevalent in this region. Sea-spray icing is one of the most critical threats, as the ice accumulation may jeopardise the safety of the crew and routine operations, and adversely affect vessel manoeuvrability. Such an effect may be detrimental, especially for smaller vessels, such as fishing vessels with lower residual stability, where the accumulation of topside icing may impair the vessel stability eventually resulting in capsizing, as evidenced by the capsizing of fishing trawler ONEGA with 17 fatalities in December 2020 in the Barents Sea.
While vessels may be equipped with various anti-icing and de-icing systems, efficient design and energy consumption estimation for winterization remain complex tasks. This exacerbated by the dynamic nature of metocean conditions contributing to ice accretion, as well as the uncertainties in the underlying environmental variables.
The above challenging context calls for an optimized fishing vessel route planning framework capable of handling multiple, often conflicting, objectives of catch yields, economics, safety, etc., under uncertain and dynamic operating conditions.
Route planning, also referred to as weather routing, has been extensively studied in the maritime sector, with various static and dynamic multi-objective optimization models and approaches proposed. However, the optimization of fishing vessel routes in Arctic waters, particularly the Norwegian Barents Sea with its unique adverse metocean conditions, remains underexplored, despite the fishing industry’s economic significance.
This work frames fishing vessel route planning as a sequential decision problem, by a Markov Decision Process (MDP) model. The proposed optimization framework considers the catch yield, fuel consumption and risk of sea-spray icing with associated uncertainties. The MDP-based model is solved using a reinforcement learning approach, offering a novel and adaptive method for fishing vessel route planning under the harsh and uncertain conditions of Arctic waters.