Windward applies AI models to vessel fuel consumption - Smart Maritime Network

Windward has announced the launch of the Windward Vessel Fuel Consumption API, an artificial intelligence-powered system providing fuel consumption assessments to allow maritime stakeholders to optimise chartering decisions and manage their carbon footprint. The service is based on a deep learning model that calculates and assesses the fuel consumption of any given tanker using multiple data sources, inputs, and behavioural features, including speed, meetings, port calls, weather conditions, hull fouling, engines, and fuel types, with more than 10 years of historical voyage data. The company says that the system can deliver assessments with up to 95% accuracy on average per voyage, provided with actionable insights to support further decision making. The datasets were built, tested and verified in collaboration with companies participating in Windward's Data for Decarbonisation Programme (D4D), a partnership aimed to increase transparency and foster collaboration within the maritime industry by leveraging the power of big data and AI to decrease carbon emissions. The service is seen as a supplementary digital source of fuel consumption data alongside noon reports, which typically involve the crew measuring and reporting data manually, making the report susceptible to human error and potential falsification.