Machine intelligence, computer vision, and fluid mechanics for Canadian ice-breakers

A model ship passing through ice-covered water, photo credit: NRC, Canada.

Photo credit to National Research Council Canada (NRC).

We live in the north, and global warming is true. The ice-covered ocean will soon be more available, and more dangerous at the same time, to be explored. Canada needs autonomous driving ships that can navigate in the ice-covered water to deliver goods and find resources. Much different than a self-driving car on the highway, a ship navigating in ice does not have a lane to guide the vehicle. Instead, how the ship should move is mainly determined by the situation of ice (e.g., thickness, size, layout, age, and strength of the ice floes). We aim at developing an information fusion system leveraging our expertise in computer vision, machine intelligence, dynamics and controls, and fluid physics to improve the safety of the next generation arctic shipping.

This project is in collaboration with National Research Council Canada (NRC).