Machine learning for reduced emissions from shipping
Machine learning for reduced emissions from shipping
Background: The efficiency of a marine engine is varying due to factors such as load conditions, fuel quality, air temperature, and wear and tear. These variations make it hard to detect and isolate performance issues from all the "noise". Recent machine rooms are fitted with more and more sensors.
Objective: The task to be studied is to use machine learning to filter out known variations (for instance load condition and fuel quality), so that performance issues can be detected, and the cause of the issue can be isolated. This will help ship operators to reduce their emissions by correcting performance issues, which will reduce emissions. Machine learning will be used in combination with physical models. The task will be done with data from a Wallenius Willhelmsen vessel and in collaboration with SINTEF Ocean.
Collaborator: SINTEF Ocean and Wallenius Willhelmsen