US university uses AI traps against invasive species in Florida

The University of Florida (UF) is using cheating with artificial intelligence (AI) to trap invasive species in the state, in this case two large reptiles that represent a threat to the biodiversity.

A team of scientists from the Institute of Agricultural and Food Sciences of the University of that institution uses traps to capture black and white Argentine tegus, species not native to Florida.

Argentine tegus, native to South America and known for their diverse diet and ability to hibernate underground during the winter, have quickly invaded several areas of central and southern Florida such as Fort Pierce, the area where the study is being carried out.

“If we can implement innovative solutions to remove invasive species that are effective and reduce costs, we will all win,” project manager Melissa Miller said in a statement.

In this context, the ‘smart traps’ Equipped with AI technology they have become a key tool in the control and removal strategy of these invasive reptiles.

These traps, designed to capture tegus specifically, operate in an automated and remote manner, significantly reducing the resources required for their removal.

The ability to monitor and operate traps through a web interface has proven effective in eliminating the need for daily visits to check traps, noted Miller, who is an assistant research scientist specializing in invasive species ecology at the Research Center. and Education of UF/IFAS, based in Fort Lauderdale.

For this project, UF/IFAS worked with Wild Vision Systems, who designed the traps and AI technology.

The software, which provides a phone app for users that provides remote control and monitoring of the trap, is designed to recognize tegus by identifying, among thousands of photographs, the characteristic demarcations on their skin.

The pilot project conducted in St. Lucie County, Florida from May to October 2023 was successful, capturing 15 tegus and demonstrating the potential of AI-equipped smart traps for controlling invasive reptiles.

With plans for a second phase of the project in summer 2024, the team hopes to further improve the effectiveness of the software and explore its application on other invasive species, such as green iguanas and Nile monitors.

“The use of AI for the research and management of invasive species seems promising and is consolidated as one of the next steps in the exploration of novel methods to win the battle against invasive species,” Miller explained.

With this pioneering project, UF/IFAS demonstrates how the integration of advanced technology such as artificial intelligence can make a difference in protecting the environment and biodiversity from the threats posed by invasive species.