When Done Right, AI Can Be a Powerhouse for Environmental Conservation

The capabilities of artificial intelligence (AI) are proving to be beneficial to wildlife. Conservation teams and researchers are integrating AI into their environmental protection practices and using it to conserve natural resources, monitor ecosystems long-term, and measure the impact of natural disasters.

Conservation relies on biodiversity measurements that are timely and accurate, and AI can support monitoring and measurement systems. That type of integration is working for the World Wildlife Fund (WWF), a conservation organization working with communities in almost 100 countries to protect wildlife and ecosystems. 

"We bring in data from all over the world, and the number of sensors that are pouring data in, the specificity of the data … and the speed is just accelerating," Dave Thau, the global data and technology lead scientist at WWF, told TriplePundit. "You need artificial intelligence to understand all of that information. Both on a global scale, which we work on, and also a local scale, which we work on, as well.” 

How to use AI for natural resource conservation

WWF has many uses for AI, including partnerships with academic, private-sector and non-governmental organizations, as well as people in house who are working with AI, Thau said. 

One of its many AI-integrated projects is Eyes on Recovery. Launched with a range of local, on-the-ground partners and supported by Google.org, the project was created to monitor the impacts of the Australian bushfires in 2020. 

During the bushfires, more than 12 million hectares of forest land were destroyed and an estimated 1.25 billion animals were killed. After the fires, the Eyes on Recovery team installed more than 600 camera traps in bushfire affected areas to monitor wildlife recovery. This allowed researchers to monitor important local and endangered species like the Kangaroo Island dunnart, a mouse-sized marsupial that is challenging to spot given its small size. 

The footage captured by the cameras is sent to an online tool called Wildlife Insights, which uses Google’s AI technology to sort and analyze the images. The team used initial photos from the project to train the AI to accurately identify different animals. Anyone with camera traps can add photos to the platform to contribute to the publicly-available data. 

“We can get a sense of which species are coming back, which are not, and what we can do to help those that are coming back,” Thau said.

Another example of an AI-enabled project is Forest Foresight. Last year, WWF-Netherlands and its partners collaborated with computer scientists and AI professionals to build an AI model that can predict illegal deforestation so action can be taken before it occurs. The model uses satellite images combined with data like population density, to study how forest loss happened in similar places in the past. That information helps it to predict areas that are at risk and alert local authorities and stakeholders. 

The tool predicts deforestation up to six months in advance with 80 percent accuracy, according to WWF. In Gabon, Forest Foresight also prevented illegal gold mining on 74 acres of forest land. 

"That system is being applied in several countries around the world, has been successful in finding places where deforestation is likely to increase, and has enabled organizations to intervene before it happens," Thau said.

Benefits and drawbacks of AI integration 

Still, AI integration can pose challenges and create social and environmental risks. Improper use of tools can skew data, create opportunities for greenwashing, dismiss data rights of local communities and spark language barriers.

WWF faced challenges figuring out where to use AI, how best to use it, and how to work with new applications, Thau said. 

But using AI comes with benefits and positive long-term impacts, too. Including understanding the effects of WWF projects and knowing what's actually working, Thau said. 

"We use artificial intelligence to understand where we can be impactful, what areas could benefit from conservation efforts, and understand risky events." Thau said. 

In addition to its technical capabilities, AI can enable engagement between organizations and local communities. Best practices and recommendations will ensure that all decision-makers are able to participate and understand the technology's insights. Leveraging the technology for environmental protection and conservation practices is making these collaborations possible.