Turning down the volume: how artificial intelligence could cut ship noise
As people continue to gather in expanding cities, noise pollution has been described as the next public health crisis. In the ocean, this situation is aggravated as sound travels further in water than it does in air. With the sound from a ship’s propeller hitting up to 170 decibels, the equivalent of a rocket engine at lift-off, and travelling up to 160 km before fully dissipating, the survival of marine life is threatened. Imagine what it’s like to be a marine animal with container ships, ferries and other commercial vessels passing overhead. When affected by noise, whales and many other marine animals struggle to communicate, eat, and reproduce.
In response, a project developed by Clear Seas and funded by Transport Canada’s Quiet Vessel Initiative and Mitacs’ Accelerate Program brings together a research team of engineers and marine biologists at the University of British Columbia to solve this serious problem by harnessing the power of artificial intelligence. To help them with their work, the research team have invited experts from across North America to join a collaborative design workshop in Vancouver, BC on June 1-2 to develop innovative ways to tackle this issue of underwater noise from ships.
Designing the ship of the future
“Most of the projects looking at mitigating ship-source noise have taken the perspective of the ship and how to make the ship quieter,” Clear Seas Executive Director Paul Blomerus says. “This project represents a paradigm shift by bridging the gap between ship design and marine biology. This work envisions ships that are a more sympathetic part of the marine environment, and the ship will ‘think’ like the marine mammals they encounter using the power of AI.”
The multi-year project is led by Dr. Rajeev Jaiman and Dr. Jasmin Jelovica from UBC’s Naval Architecture and Marine Engineering program, in collaboration with Dr. Andrew Trites and Dr. David Rosen from the UBC Marine Mammal Research unit. Lead researcher Dr. Jaiman says he named the research project “MELO” as a play on words; the project will “create a more MELOdious ocean soundscape based on an improved understanding of the sensitivity of the marine environment.”
This initiative puts machine learning at the forefront to develop a noise prediction toolkit that will allow ships to tune and adjust their noise in real-time based on the presence and type of marine mammals. The research team is glad to have the support of the Canadian Coast Guard and will be testing the accuracy of their predictions using data gathered from the Canadian Coast Guard research vessel Sir John Franklin in summer 2023.
Source: Clear Seas