
CETI’s project to decipher the language of sperm whales is using the AVATARS automation system to track sperm whales in the wild to increase the use of drones to collect data on their communication.
AVATARS’ new automated solution uses a drone system to track sperm whales and collect data about their communication. Sperm whales are difficult to observe because they spend most of their time underwater and only about a quarter of their time on the surface. Using machine learning, AVATARS was able to “encounter” three sperm whales off the coast of Dominica using their audio signals and data from special radio frequency (VHF) tags, while AVATARS’ autonomous control module used reinforcement learning algorithms to predict the surfacing and emergence times to “encounter” the sperm whales and control drone to increase the probability of a “meeting”. A waterproof sensor module processes radar data to pinpoint the whale’s path.
The CETI (Cetacean Translation Initiative) project is a non-profit initiative to decipher the acoustic signals of sperm whales off the coast of Dominica in the Caribbean Sea, which aims to use advances in robotics and machine learning to understand whale communication. Project researchers have already discovered a cetacean “phonetic alphabet” that may indicate that whale communication is similar to human language.
Using drones and innovative algorithms, the AVATARS system could greatly simplify the collection of data on sperm whales and lead to further breakthroughs in understanding their communication. This opens up new perspectives in marine biology and artificial intelligence research.