With the International Maritime Organization (IMO) announcing a goals-based approach to emissions reductions for sulphur, nitrogen and carbon oxides over the next two decades, marine data intelligence business GreenSteam has called on shipowners and operators to use vessel data to build a fuller picture of unnecessary fuel consumption.
The Denmark headquartered company has stated that the IMO 2020 sulphur cap is just the beginning of greenhouse gas emissions (GHG) goals with reductions targets of 85 per cent by 2050 and 40 per cent by 2030 the long-term visions.
While companies are considering hardware-based technologies to cut GHG emissions, such as Flettner rotors, electric drivetrains and air lubrication, GreenSteam believes the industry is expected to adopt machine learning as a means of accurately quantifying each operational lever that can cut fuel consumption and reduce emissions as part of a multi-solution approach. The company says that the shipping industry cannot ignore any area of operational fuel wastage, as it considers each area of new technology.
Data collected from the ship together with metocean and AIS data indicates where and why the vessel uses excess fuel. These insights allow the crew to adjust operations, minimising fuel wastage, reducing cost and cutting emissions. On average, fuel savings of 5 per cent are typical and, in some cases, 20 per cent fuel savings have been observed.
GreenSteam reports that traditional or legacy methods of vessel data analysis are still very common across the industry despite the fact that some of these methods exclude 90 per cent of vessel data. The company says that this is because analysts don’t have access to the methods or processing capacity to consider all the multi-dimensional factors affecting performance and therefore need to use quite drastic sampling to simplify their task. Critical data that might highlight worrying symptoms of excess fuel consumption, can take longer to become apparent or might be overlooked entirely under this data sampling regime.
According to GreenSteam, machine learning enables data analysis of almost all vessel data, with powerful computing technology “connecting the dots” and identifying the relationship between each of the 13+ factors affecting vessel fuel efficiency. This inclusive and accurate approach can highlight fuel savings opportunities 2-4x higher than traditional or legacy data analysis.
“Vessel performance is highly complex involving multiple and often inter-related factors. In order to identify and measure the true level of fuel wastage it is vital that all data is used and analysed. Vessel owners and operators who are not using data analytics informed by machine learning may only be working from 10 per cent of the critical information they need to make decisions on their vessel and fleet operations,” said Simon Whitford, COO of GreenSteam.
“Each tonne of HFO fuel wasted enters the atmosphere as three tonnes of carbon dioxide. Machine learning provides a solid foundation for clear, actionable advice empowering ship owners and operators to cut fleet-wide GHG emissions,” he said.