Predicting the Unpredictable: Leveraging AI to Forecast Tramp Emissions

By Kristofer Maanum, Senior Product Manager, Bearing AI
It’s no secret that maritime shipping is one of the world’s largest emitters of CO2. Thankfully, the industry is taking steps to reduce its carbon footprint.
Initiatives like the Carbon Intensity Indicator (CII) from the International Maritime Organisation (IMO) and the European Union Emissions Trading System (EU ETS) are pushing shipping companies to restructure their operations in an effort to reach net zero emissions by 2030.
Shipping companies are using cleaner fuels and reducing speed to minimize emissions. They’re optimising their service routes to help ensure each vessel in their fleet meets EOY carbon targets. And, increasingly, they’re using AI to make their operations more efficient.
AI can sift vast amounts of data to find patterns no human would ever detect. Algorithms trained on millions of vessels can predict and optimise fuel usage, emissions, and even the impact of seaway variables like weather and biofouling on any voyage.
But some forms of shipping push even AI to its limits.
Filling in the Gaps
Tramp shipping companies pose a unique challenge for artificial intelligence. Unlike their liner counterparts, tramp operators face uncertainty about future routes and operational profiles of their vessels, which in turn complicates efforts to reduce emissions and comply with environmental regulations.
Artificial intelligence—while capable of immense calculations and forecasts more accurate than any human could produce—has historically struggled with tramp shipping. Regular, consistent schedules like those of liner companies lend themselves well to AI’s analytical capabilities. However, the inherent unpredictability of tramp operations presents a more difficult problem for data-backed decision-making.
Bearing AI is working to bring tramp shipping companies up to speed with the rest of the shipping industry, crafting powerful AI models trained specifically for the tramp industry’s volatile schedules.
The secret is data. Vast amounts of data.
"Our new Deployment Planner is designed to provide precise, real-time insights for tramp operators."
A New Tool for Tramp
Bearing’s Deployment Planner for tramp leverages AI models trained on millions of voyages and is tuned specifically to the volatility of tramp operations to fill in the unknown gaps of tramp vessel schedules and provides data-driven EOY emissions forecasts.
The Deployment Planner for tramp offers quick insights into the environmental impact of an upcoming contract(s). It also allows users to experiment with different vessel deployments to determine the best fit for a specific contract from an environmental standpoint.
Tramp companies can now see the forecasted impact of an individual contract (or multiple contracts) on a vessel’s EOY CII rating — allowing vessel operators to factor in the exact emissions impact of a contract in their negotiations. This enables them to manage fleet efficiency, mitigate compliance risks and reduce operational costs.
“Tramp shipping companies face a distinctly different set of operational challenges compared to their liner counterparts,” Aleksandar-Saša Milaković, Product Manager & Naval Architecture Lead at Bearing AI said.
“Our new Deployment Planner is designed to provide precise, real-time insights for tramp operators, ensuring they can make informed decisions that maximise both efficiency and sustainability, regardless of the variability in their contracts. In practice, the tool allows operators to input the contracts they expect until the EOY, while Bearing AI will fill in the gaps and forecast the expected EOY CII rating, at the same time allowing the operator the flexibility to test how potential changes in planning might affect the EOY CII outcome.”
Bearing AI’s forecasts go one step further. The Deployment Planner predicts the EOY performance for every vessel in a tramp fleet, whether or not they have contracts scheduled, by creating a performance forecast for uncontracted time periods.
These predictions provide insights into vessel emissions and help tramp operators achieve superior emissions performance despite their constantly shifting schedules.
The Deployment Planner demonstrates the power of AI for tramp shipping companies striving to enhance their environmental performance.
Bearing.ai
