TA1 - Exposé magistral 3 / Tutorial 3

Maritime supply chains face growing pressure to balance operational efficiency with environmental sustainability. This tutorial presents data-driven and machine learning (ML) approaches to address these dual challenges in the context of the St. Lawrence Seaway, a critical corridor for Canadian and international trade.

We demonstrate how ML models can revolutionize the management of maritime operations by enabling accurate, real-time predictions of vessel estimated times of arrival (ETAs), speeds, emissions, and underwater noise levels from the moment a ship enters Canadian waters. These predictions support comprehensive monitoring of maritime traffic and environmental impact, forming the backbone of intelligent decision-support systems for maritime supply chain operators.

A central focus of this tutorial is the role of operational optimization, in particular, just-in-time (JIT) navigation policies, as a practical, low-cost strategy for reducing fuel consumption, greenhouse gas emissions, and underwater noise pollution. We further explore the impact of this strategy on marine ecosystems, with a particular emphasis on the protection of marine mammals in the St. Lawrence River.