EPS uses DeepSea AI to improve accuracy of fuel consumption forecasts
Published by Alfred Hamer,
Editorial Assistant
Dry Bulk,
DeepSea Technologies, the Al-led maritime technology company, has announced it has delivered weekly fuel consumption forecasts accurate to within 1% after a 6 month project with Eastern Pacific Shipping (EPS). Measured against a baseline of standard data inputs, improved data quality reduced the inaccuracy of weekly fuel consumption forecasts to 0.8%.
Across the business, EPS teams were able to benefit from accurate vessel behaviour models, supporting improved vessel operations and maintenance scheduling, enabling chartering teams to provide more reliable speed and consumption warranties to customers.
The transformation was made possible by the provision of real-time sensor data from EPS’s vessels, coupled with DeepSea’s AI-driven Cassandra platform, which provides highly granular, real-time insights to help EPS streamline fuel consumption, ensure regulatory compliance, and reduce environmental impact.
Working together, EPS and DeepSea have driven technical advancements in the EPS fleet and developed innovative approaches to model validation that support EPS’s strategic roadmap for digitalisation and decarbonisation. The fleet owner has prioritised developing a robust data pipeline and rigorous validation process to improve fleet optimisation approaches.
Over six months in 2024, DeepSea and EPS built and implemented advanced digital twins, offering an accurate, up-to-date understanding of vessel behaviour, validating them through a combination of empirical and statistical observations and methods to confirm model accuracy.
Deploying DeepSea’s Cassandra technology, EPS can now evaluate vessel performance with precision, generating predictive insights that inform performance-based decision-making across its 300 ship fleet of tankers, bulk carriers, PCTCs, containerships, and gas carriers. Using the AI technology, EPS expects to enhance vessel and fleet performance monitoring, improve reaction time to inefficiencies and drive forward its decarbonisation efforts.
“This result is testament to the fact that the technology now exists to model vessel behaviour incredibly accurately. We can finally say that it’s a solved problem. The industry is quickly realising that the availability and quality of data is now the limiting factor.” said Dr. Konstantinos Kyriakopoulos, CEO and Co-founder of DeepSea Technologies. “For EPS, understanding the value of data is embedded in their organisational DNA, and they were uniquely prepared to leverage high-frequency sensor data to achieve next-level efficiency.”
“Performance is not the work of a single department – it's about giving the entire organisation the tools to make data-driven decisions. Good information enables good business,” said Pavlos Karagiannidis, Fleet Optimisation Manager at Eastern Pacific Shipping.
“With our unwavering focus on performance, DeepSea’s Cassandra platform empowers EPS to achieve new levels of operational efficiency and advance our sustainability goals,” added Karagiannidis.
Click here for free registration to Dry Bulk
Read the article online at: https://www.drybulkmagazine.com/environment-sustainability/14112024/eps-uses-deepsea-ai-to-improve-accuracy-of-fuel-consumption-forecasts-for-bulk-tankers/
You might also like
Vale and Petrobras announce a partnership to test fuel with renewable content on bulk carrier
The product was formulated by Petrobras Singapore (PSPL) itself in its locally leased tanks, by blending 76% fossil fuel oil from the refineries of the Petrobras System and 24% UCOME, a biofuel originating from the processing of used cooking oil (UCO), purchased in the region.