Client story: Maersk Tankers
Forecasting market prices with 20% more accuracy
Deploying reliable forecast models for freight prices to maximize revenue in the shipping industry
Maersk Tankers facilitate the global trade of energy to meet the world’s energy needs. They are prominent in the product tanker industry, operating one of the largest fleets of vessels and have 3,000 employees worldwide.
Maersk Tankers lead the industry in digitization, using algorithmic trading to position vessels where the customers need them the most and support pool partners in optimizing value. With 2021.AI, Maersk built reliable forecast freight prices and maximized revenue opportunities.
Download the case
Download the PDF-version of “Forecasting market prices with 20% more accuracy” client story
The challenge
Tanker shipping is a dynamic industry that demands the highest standards in planning from both the officers on the ships and managers onshore. Over the past few years, Maersk Tankers have transformed from a traditional tanker company into a service company. As such, they continuously strive to support decision-makers with innovative solutions to make better-informed business decisions, including where and when to position a global fleet of tankers.
To achieve this, one of the digital products Maersk Tankers has launched is a reliable and accurate model to forecast freight prices across 20 market segments to optimize the positions of their global fleet of tankers.
··········
Together with 2021.AI we are creating some of the most advanced and ambitious forecast models for the shipping industry. This collaboration offers new ways of increasing efficiencies running our global fleet. We are leveraging 2021.AI’s expertise in developing advanced forecast models.
The solution
The 2021.AI’s data science team analyzed the existing forecasting model and data, addressing shortcomings and creating the first baseline to benchmark improvements for the new models. The team developed and validated more than 20 new models, applying algorithms to the data and training the data set in a series of 2 weeks sprints. The sprints were propelled leveraging 2021.AI’s Grace AI Platform, which helps boost the collaboration between data scientists through a common workflow and fast-tracking the project leveraging the latest technology advances within AI and machine learning.
Additionally, new data sources were added to improve the overall data quality, making the new models predict the future freight market based on historical data from an extended and enriched dataset going five years back in time.
The results
After Maersk Tankers’ chartering team validated the new forecast models, the models were swiftly taken into production, improving accuracy by 20% on average. As a result, better-informed decisions have paved the way for Maersk Tankers to optimize the positioning of their global fleet of vessels.
Project highlights
- The Grace AI platform provided a shared workspace for developing a forecasting model
- The forecasting accuracy of the existing model improved measurable business value by 20%
About Maersk Tankers
Maersk Tankers facilitate the global trade of energy to meet the world’s energy needs. The company are leading players in the product tanker industry, operating one of the largest fleets of vessels and employing 3,000 employees worldwide. Maersk Tankers lead the industry in digitization, using algorithmic trading to position vessels where the customers need them the most to cover their transportation needs and support pool partners in optimizing value.