Ted Hisokawa
March 04, 2025 08:58
A new AI model, Atlanteans, developed by AI2, improves the surveillance of world fisheries and fauna, helping environmentalists to protect natural resources against illegal activities.
The innovative AI model targets illicit fishing
In a revolutionary development for conservation efforts, the researchers have introduced an open source IA model named Atlanteans, designed to monitor world ships of the sea. Developed by Allen Institute for IA (AI2) in Seattle, this model aims to fight illegal fishing by analyzing more than five billion GPS signals from around 600,000 ocean ships Nvidia.
High precision and real -time alerts
Atlantes has an impressive prediction of approximately 80% in determining the activity of a ship. Integrated into the AI2 maritime maritime maritime surveillance platform, it can alert the authorities within 15 minutes of detecting potential illegal fishing activities. This capacity was demonstrated when the Argentina navy intercepted and inflicted a fine a ship for illegal fishing, following a Skylight alert.
Technological backbone of Atlantean
The AI model, made up of 4.7 million parameters, is built on the basic GPUs of the Nvidia H100 and Pytorch tensor. It processes data from the automatic identification system (AIS), which is compulsory for most ships, from January 2022 to June 2024. Training involved maritime experts annotating more than 15 million signals to improve the accuracy of the model.
Wider implications for global fishing
Illegal, unconvertised and unregulated (INI) fishery costs the world economy up to 23 billion dollars per year, representing around 20% of world fishing catches, according to Financial Transparency Coalition. African waters are particularly affected, where local communities strongly depend on fishing for subsistence and employment.
Extension of AI to fauna conservation
AI2 plans to extend the use of Atlanteans beyond maritime applications, integrating it in Earthranger, a platform that aggregates data from various sources to monitor fauna. This includes monitoring elephants, rhinos and wild dogs to alleviate human life conflicts. The system will be formed to predict the behavior of elephants, helping to reduce clashes between elephants and farmers.
Future perspectives and the impact of conservation
Using large sets of elephant movements, AI2 aims to proactively deal with the life conflicts of humanity. Jes Lefcourt, director of Earthrant, underlined the potential to save elephant lives by predicting their movements and preventing conflicts with humans.
The infrastructure supporting the classification of fishing ships is also applicable to forecasting the behavior of elephants, presenting the versatility of AI in conservation. This innovative approach could considerably improve the protection of natural resources and biodiversity in the world.
Image source: Shutterstock
(Tagstotranslate) ai
