12 Oct The AI that helps whales: the Whale Safe project
Every year, along the West coast of the United States, 80 minke, humpback and blue whales (the latter endangered) are killed by the ever-increasing vessel traffic. In fact, due to climate change, whales are forced to move closer and closer to the coast in search for food, ending up in the route of ships. Together with his team, Douglas McCauley, director of the Benioff Ocean Initiative at the University of California, has developed a solution to this problem: Whale Safe. It is an artificial intelligence that aims to reduce collisions between whales and ships along the California coast.
How Whale Safe works
Designed to prevent fatal ship collisions with whales, Whale Safe maps and analyses whale and ship data in near real-time. The data collected are divided into three categories:
- Visual whale data, collected mainly by trained observers who record whale sightings on board vessels. When whales are spotted, their location, species, and behavior are recorded with the Whale Alert and Spotter Pro apps and are transmitted to a database when the boat returns to dock. These observations are also combined with data collected during monthly aerial surveys of the Santa Barbara Channel shipping lanes.
- Acoustic whale detections: using a hydrophone (an underwater microphone) and thanks to an artificial intelligence that recognizes them, the vocalizations of blue, humpback and minke whales are identified. Information about detected sounds are transmitted via satellite to scientists who review them.
- Updated habitat model of the blue whale, which using data collected from 104 minke whales tagged by satellite analyses how suitable a habitat may be for the specimen and makes statistical hypothesis of its location.
These three sources are supplemented by the Whale Presence Rate (divided into low, medium, high, very high). The aim of the Whale Presence Rating is to provide the shipping industry, natural resource managers, and the public with a data-driven assessment of whale presence to reduce the risk of ship collisions with whales.
In addition to sharing whale data, the tool detects international shipping lanes, voluntary vessel speed reduction zones and shipping activity detected by Automatic Identification Systems (AIS). In addition, Whale Safe has a reporting system that summarizes AIS data to determine which vessels and companies best abide by voluntary speed recommendations to protect endangered whales. Finally, the tool allows users to receive email updates on both whale data and safe navigation for them.
Results and future prospects
It seems that this solution is working well: in 2021, the first year of operation along the Santa Barbara Channel, no accidents were recorded. But that is not enough: in addition to wanting to expand monitoring to all the most at-risk areas of North America, Whale Safe is currently looking to implement the system directly into the navigation components of future ships. One thing is certain: this is going to be a great benefit for whales.
This example shows how AI is able not only to adapt to human needs, but also to act in favour of the environment and animals, taking the first steps towards a future of coexistence and mutual respect between these elements.