Marine biologists are inching closer to understanding the ins and outs of sperm whale communication. But in order to decode what the cetaceans are saying, they must first need to find them and know where they will surface. This is no easy feat, since sperm whales can dive over 10,000 feet andstay way below the surface for up to 60 minutes. Their habitats themselves stretch for thousands of miles.
Now, scientists from Project CETI (Cetacean Translation Initiative) and Harvard University are proposing a new method for finding sperm whales and predicting where they will surface using autonomous robots and a rich combination of sensor data. The method is detailed in a study published October 30 in the journal Science Robotics.
[Related: Sperm whales may have their own 'alphabet.']
This method is a chance for scientists to test out new algorithms, sensing data, and artificial intelligence in a challenging environment. First launched in 2020, Project CETI is working to collect vocalizations to decipher how sperm whales communicate with one another. They have used tags affixed to the whales to track them in real-time and customized drones to follow their behaviors.
The new study uses various sensing devices, including aerial drones equipped with very high frequency (VHF) signal sensing capability. These devices can use leverage signal phase and the drone's motion to imitate an "antenna array in air." This can help estimate which direction pings from one of the tagged whales CETI is tracking is traveling.
According to the team, this shows how to predict when and where a sperm whale may surface using sensor data to predict their dive behavior. With that data, Project CETI can now create algorithms for the most efficient route for a drone to or encounter a whale at the ocean's surface. In the future, this could be applied to conservation methods, including helping ships avoid hitting whales when they are surfacing.
They call this method the Autonomous Vehicles for Whale Tracking And Rendezvous by remote Sensing (AVATARS) framework. It uses two interrelated components-autonomy and sensing. Autonomy determines the positioning commands for autonomous robots deployed to increase the chances of visually spotting whales. Sensing measures the angle-of-arrival from whale tags to better inform the decision-making process of where to deploy the robots. The data taken from autonomous drones, surfaced tags, underwater sensors, and whale motion models from previous biological studies of sperm whales are all put into the AVATARS autonomous decision-making algorithm, which then aims to minimize missed opportunities to rendezvous with sperm whales.
[Related: Sperm whale clans tell each other apart by their accents.]
A similar and more well-known application of this kind of time-critical rendezvous method is in rideshare apps. They use real-time sensing to note the changing paths and positions to connect drivers with potential riders. When a rider requests a ride, the app can assign a driver to get to the rider as quickly and efficiently as possible. According to the team, Project CETI's new method works in a similar way, by tracking the whales in real-time with the eventual goal of coordinating the drone's rendezvous to meet the whale at the surface.
"I'm excited to contribute to this breakthrough for Project CETI," study co-author and Harvard University computer scientist Stephanie Gil said in a statement. "By leveraging autonomous systems and advanced sensor integration, we're able to solve key challenges in tracking and studying whales in their natural habitats. This is not only a technological advancement, but also a critical step in helping us understand the complex communications and behaviors of these creatures."