Silicon Valley AI developer takes ship performance solutions “out of stealth mode”

Japan's K-Line is to deploy Bearing's Performance Analysis Dashboard on 300 ships.

Japan’s K-Line is to fit 300 ships with an AI (artificial intelligence) based performance analysis solution developed by Palo Alto, Calif., based Bearing Inc., a company backed by AI pioneer Andrew Ng’s AI Fund and by Mitsui & Co., Ltd.

With this high profile order booked, Bearing says that it is now “exiting stealth mode” to bring AI-powered operational efficiencies to shipping through highly accurate ship performance models that promise improved efficiency, safety and reduced greenhouse gas emissions.

Bearing says its performance models allow it to predict fuel consumption, speed, and other performance factors much more accurately than existing solutions on the market. For example, Bearing’s typical prediction accuracy for fuel consumption is said to be over 98% per voyage, compared to a typical existing accuracy of 80%).

Bearing’s team has deep roots in artificial intelligence (AI) and scaling large-scale data-driven products. Core to Bearing’s technology is a platform that that takes in data from sources including ship sensors, satellite positioning and weather data. Bearing combines this data with deep learning technologies to build hyper accurate models that predict vessel performance in a wide range of operating conditions and serve as the foundation for a range of products, including:

Smart Routing Engine: Automatically analyzes all potential routes for a given voyage and recommends the optimal route and speed profile for fuel efficiency, market opportunities and safety. Notably, the Smart Routing Engine continuously evaluates all possible route options and automatically adapts to changing weather conditions in real-time. Bearing built the Smart Routing Engine to integrate into existing workflows smoothly, so captains and operators can quickly receive and react to recommendations. The Smart Routing Engine also accurately projects the real-time profitability of a given voyage, removing much of the guesswork that exists in the industry today, says Bearing..

Bearing’s Smart Routing Engine is already deployed with its initial partner, Japan’s Mitsui OSK Lines Ltd. (MOL), which collaborated on the initial testing and launch of the solution.

“MOL, at its highest organizational levels, continuously monitors the condition of our fleet to ensure optimum operational efficiency and prudent, safe navigation by combining the technologies of Bearing as well as other existing and new solutions,” said President of MOL (Americas), Katsumi Nagata.

Performance Analysis Dashboard: Allows operators to see accurate trends in efficiency for their entire fleet. The dashboard proactively highlights actions that operators can take to improve fleet efficiency and allows operators to calculate the ROI for different actions, such as hull cleaning. Traditionally, measuring vessel performance against an objective baseline has been challenging as vessel performance can vary significantly even on a single voyage due to factors like weather. Bearing says its technology allows operators to understand their fleet in a clear, quantitative way, thereby empowering them to make the right decisions.

K-Line, which now plans to deploy the solution on 300 ships. collaborated closely with Bearing on its development and testing of the Performance Analysis Dashboard.

“As a result of verifying the performance analysis by utilizing Bearing’s technology, K-Line confirmed a much higher level of data accuracy compared with the existing performance analysis tool,” said General Manager of K-Line’s Advanced Technology Group, Shingo Kameyama. “High accuracy of ship’s performance analysis is essential for all safety and energy-saving measures in ship operation. Bearing’s technology contributes to our safe, economical operation and greenhouse gas reduction, which is one of the most significant maritime industries’ challenges.”

Ship Profile API: Enables shipping companies and other service providers to tap into Bearing’s deep learning models via an easy-to-use API to power their own tools. Bearing has partnered with ZeroNorth, a spin-off of Maersk Tankers, to develop and launch this API. Bearing’s API is currently helping to power ZeroNorth’s Optimize platform, allowing tramp shippers to optimize their operations and improve overall profitability.