THere is a thread of thought, from science fiction movies to Stephen HawkingThis suggests that artificial intelligence (AI) can be deadly to humans. But conservationists are increasingly turning to artificial intelligence as an innovative technology solution to tackle the biodiversity crisis and mitigate climate change.
A recent report prepared by Wildlabs.net I found that AI was one of the top three emerging technologies in conservation. From trap camera and satellite images to audio recordings, the report notes: “Artificial intelligence can learn to identify images among thousands of rare species. Or identify animal call after hours of field recordings—dramatically reducing the manual work required to collect vital preservation data.”
AI helps protect species as diverse as humpback whales, koalas and snow leopards, and supports the work of scientists, researchers and rangers on critical tasks, from anti-poaching patrols to species monitoring. With machine learning (ML) computer systems that use algorithms and models to learn, understand and adapt, AI is often able to do the work of hundreds of people, and get results faster, cheaper, and more efficiently.
Here are five AI projects that contribute to our understanding of biodiversity and species:
1. Stop the Hunters
Kafue National Park in Zambia is home to more than 6,600 African savannah elephants and covers an area of 22,400 square kilometres, so stopping poaching is a major logistical challenge. Illegal fishing in Lake Itzi-Teji on the park border is also a problem, and fishermen disguise themselves as fishermen to get in and out of the park undetected, often under cover of darkness.
Connected Preservation Initiative, from international rangers game (GRI), the Zambia National Parks and Wildlife Department and other partners, are using artificial intelligence to boost traditional anti-poaching efforts, creating a 19km virtual fence across Lake Itzi-Tizi. FLIR infrared (FLIR) thermal cameras record each boat passing in and out of the park day and night.
The cameras were installed in 2019, and were manually monitored by security guards, who could then respond to signals of illegal activity. FLIR AI is now trained to automatically detect boats entering the park, increasing efficiency and reducing the need for constant manual monitoring. Flying waves and birds can also trigger alerts, so the AI is taught to eliminate these false readings.
“There have always been insufficient resources to secure protected areas, and having people watching multiple cameras 24/7 just isn’t expanding,” says Ian Hood, GRI’s special technical advisor. “Artificial intelligence can be a game-changer, as it can monitor illegal boat crossings and immediately alert teams of rangers. This technology has enabled a handful of forest rangers to provide round-the-clock monitoring of a massive illegal entry point across Lake Itzi-Tese.”
2. Track water loss
Brazil lost More than 15% of its surface waters In the past 30 years, a crisis has emerged only with the help of artificial intelligence. The country’s rivers, lakes and wetlands are facing increasing pressures from growing population, economic development, deforestation, and exacerbation of effects from the climate crisis. But no one knew the extent of the problem until last August, when the MapBiomas aquatic project released its results using ML, after processing more than 150,000 images created by NASA’s Landsat 5, 7 and 8 satellites from 1985 to 2020 across 8.5 million square metres. kilometers of Brazilian territory. Without artificial intelligence, researchers cannot analyze water changes across the country at the scale and level of detail required. AI can also distinguish between natural and man-made bodies of water.
The Negro River, a major tributary of the Amazon and one of the world’s 10 largest rivers by volume, has lost 22% of its surface water. Brazilian part of The Pantanal, the world’s largest tropical wetland, has lost 74% of its surface water. These losses are devastating to wild animals (4000 species of plants and animals live in the Pantanalincluding jaguars, tapirs and anacondas), people and nature.
“AI has given us a shockingly clear picture,” says Cassio Bernardino, head of the WWF’s MapBiomas aquatic project in Brazil. “Without AI and machine learning techniques, we would never have known how dire the situation would be, let alone have the data needed to convince people. We can now take steps to address the challenges that surface water loss poses to biodiversity and amazing Brazilian communities.”
3. Find whales
Knowing where the whales are located is the first step in taking measures such as marine protected areas to protect them. It is difficult to visually locate humpbacks across vast oceans, but they special singing It can travel hundreds of miles underwater. in National Oceanic and Atmospheric Association (kind) fisheries in the Pacific Islands, voice recorders They are used to monitor populations of marine mammals on remote and hard-to-reach islands, says Anne Allen, a research oceanographer at Noaa. “In 14 years, we have collected about 190,000 hours of audio recordings. It would take a very long time for an individual to manually identify whale vocalizations.”
In 2018, NOAA entered into a partnership with Google’s artificial intelligence for social interests Vital Acoustics Team Create an ML form He can recognize the song of a humpback whale. “We have been very successful in identifying humpback song from the whole data set, and establishing patterns for its presence in the Hawaiian and Mariana Islands,” Allen says. “We also found a resurgence of a hunchback song in Kingman reef, a site that has not previously documented the existence of a hunchback. This comprehensive analysis of our data would not have been possible without artificial intelligence.”
4. Koala protection
Australian koala numbers are in serious decline due to habitat destruction, domestic dog attacks, road accidents and bushfires. Without knowing their numbers and whereabouts, rescuing them is a challenge. Grant Hamilton, Associate Professor of Ecology at Queensland University of Technology (QUT), created the AI save hub with federation and Land Care Australia Funding for koalas and other endangered animals. Using drones and infrared imaging, an AI algorithm quickly analyzes infrared snapshots and determines whether the heat signature is a koala or another animal. Hamilton used the system after destroying Australia Forest fires in 2019 And the 2020 To identify surviving koala populations, mainly on Kangaroo Island.
“This is a game-changer project to protect koalas,” says Hamilton. “Powerful AI algorithms are able to analyze countless hours of video footage and identify koalas from many other animals in dense bush. This system will allow Landcare groups, conservation groups and organizations working to protect and monitor the species to survey large areas anywhere in Australia and send data us at QUT to process it.
“We will increasingly see the use of artificial intelligence in conservation,” he adds. “In this current project, we simply couldn’t do it with the same speed or accuracy without AI.”
5. Count the species
Saving species on the verge of extinction in the Congo Basin, the world’s second largest rainforest, is a huge task. In 2020, data science company Appsilon has teamed up with the University of Stirling in Scotland and the National Park Agency in Gabon (ANPN) to develop Mbaza. AI Image Classification Algorithm To monitor biodiversity on a large scale in the Loubé and Ouaka National Parks in Gabon.
Conservationists were using robotic cameras to capture species, including African forest elephants, gorillas, chimpanzees, and pangolins, which then had to be identified manually. Classifying millions of images can take months or years, and in a country about 150 elephants are lost every month for fishermenTime is important.
The Mbaza AI algorithm was used in 2020 to analyze more than 50,000 images collected from 200 camera traps spread over 7,000 square kilometers of forest. Mbaza AI rates up to 3000 images per hour and has an accuracy of 96%. Conservationists can monitor and track animals and quickly detect anomalies or warning signs, enabling them to act quickly when needed. The algorithm also works offline on a regular laptop, which is useful on sites where there is no or poor internet connection.
“Many forest mammals in Central Africa are threatened by unsustainable trade, changes in land use and the global climate crisis,” says Dr Robin Wittock, Postdoctoral Research Fellow at the University of Stirling. “Appsilon’s work on the Mbaza AI app enables conservationists to quickly identify and respond to threats to biodiversity. The project began with 200 camera traps in Lobi Waka National Parks in Gabon But, since then, hundreds of others have been deployed by various organizations across West and Central Africa. In Gabon, the government and the National Parks Agency aim to deploy cameras across the country. Mbaza AI can help all of these projects speed up data analysis.”