Optimization of travel, estimated arrival time and fuel consumption through artificial intelligence

The CSL Group Inc. (CSL) is a shipowner that provides services for the handling and transport of dry bulk goods. It delivers more than 78 million tons of dry bulk per year. Its clients work mainly in the construction and steel, energy and agri-food industries.

Context of the project

In the initial phase, the CSL Group’s project aims to develop and apply, through artificial intelligence, a deep learning solution to the group’s entire Canadian fleet (16 vessels operating on the St. Lawrence Seaway in 2022). In a second phase, the CSL Group will develop a collaborative solutions architecture to house data from other Québec shipowners in order to feed information into and strengthen the deep learning models. As a result, the models for predicting fuel consumption, arrival time and transit optimization will be made available free of charge to all shipowners that will feed their own data into the model.

Positive fallout

The development of advanced algorithms will produce and fine-tune three distinct models:

• a predictive fuel consumption model based on real-time data making it possible to clearly identify actions that influence consumption and pinpoint the most advantageous scenarios;
• a model to accurately estimate the time of arrival of ships (ETA) at various check points (ports, locks, etc.) by integrating real data from ships, as well as physical data (tides, winds, currents, etc.) and regulatory data (speed restriction zones, lock schedules, etc.);
• a model to optimize navigation by providing recommendations to the crew to maximize fuel economy and minimize constraints associated with the estimated time of arrival of vessels.

This project will improve the CSL Group’s competitiveness and, where applicable, the competiveness of Québec shipowners by reducing expenditures associated with fuel. The CSL Group estimates the savings to be between $1.6 million and $4.5 million annually for its fleet alone. It will also help reduce greenhouse gas emissions from CSL’s fleet by optimizing fuel consumption. A decrease of between 1,182 and 2,955 tons per year is estimated by CSL.