A new project uses AI to improve the identification and measurement of fish species in Africa’s Nile Basin. The software can help scientists understand fish population density more quickly than human observers. It’s part of a larger effort to harness AI to improve sustainability across a wide range of industries. “The promising thing about AI is that it now allows us to do tasks that would be time-consuming or impossibly complex using traditional methods, with considerably more speed and efficiency,” Andrew Dunckelman, head of impact and insights at Google.org, the search giant’s charitable arm, told Lifewire in an email interview.
Something Fishy
The UN’s Food and Agriculture Organization is working to improve access to the AI technology that monitors fish stocks. Getting more information about fish species could help build algorithms to identify species and their locations and recognize any changes. The UN estimates one-third of all fish stocks are now overfished and are no longer sustainable. To help keep fish stocks safe, University of Florida researchers are also using AI to make sure fishermen aren’t catching endangered species. The AI models estimate the locations of endangered species where fisheries operate, which helps commercial fishers avoid fishing in those areas. “AI is not a silver bullet to all of our problems,” Zachary Siders, the scientist who developed the application, said in the news release. “We have to keep in the front of our minds that the decisions we allow an AI system to make have real consequences for livelihoods of the fishing industry as well as irreplaceable species.”
AI Keeps Watch
It’s not just fish that AI is keeping an eye on when it comes to the environment. Climate TRACE, the world’s near-real-time greenhouse gas (GHG) monitoring platform, is helping identify where emissions are coming from and pinpointing where decarbonization efforts should be focused. There’s also Restor.eco, an open data restoration platform hosted on Google Earth. It provides scientific data and high-resolution satellite imagery to allow researchers to analyze the restoration potential of any place on Earth. Essentially, the program can map out land to predict where trees can naturally grow. Dunckelman said that Google has found that programs achieve their goals faster with AI. He noted the case of BlueConduit, an organization that emerged out of the Flint, Michigan, water crisis. The group built a machine learning platform that uses data about the age of homes, neighborhoods, and known lead service lines to predict whether a home is serviced with lead pipes. “In the past, the only way to know this would be to physically dig [at] each site and inspect for lead pipes, which is costly and time-consuming,” Dunckelman said. “Through the introduction of machine learning, BlueConduit can now quickly predict with greater accuracy whether a home is serviced with lead lines, which can drive policy decisions that have a substantial impact on both public health and government resources.” But not everyone agrees that big tech companies can necessarily solve the planet’s problems through AI. Eric Nost, an assistant professor at the University of Guelph who researches how data technologies inform environmental governance, said recent studies have raised concerns about the amount of energy and other resources required to implement AI hardware and software. “I suspect many researchers will find it hard to translate AI-based findings into actual policy or decisions if that AI hasn’t been developed with policy and decision-makers in mind, especially in light of the challenges to explaining how an AI arrives at its results,” he told Lifewire in an email interview. AI for sustainability is still in its infancy, too, Dunckelman acknowledged. The field still lacks sufficient data sets and models needed to drive progress. “For example, we all know there are emissions happening in the world, but we don’t really know where they come from,” Dunckelman added. “All we have is what the emitters themselves say they’re doing, which is imperfect.”
