Environmental data-gathering technology has proliferated in recent years. But how do you derive meaningful insights from myriad data sources? A new AI-powered platform aims to solve this problem. OlmoEarth, developed by the nonprofit Allen Institute of AI (Ai2), is a platform that integrates multiple artificial intelligence models that have been trained on approximately 10 terabytes of environment observation data. The open-source platform, launched in November, helps extract actionable insights from satellite as well as sensor data. The platform allows researchers as well as organizations to use their own data to customize a foundational model and use it to monitor trends such as forest loss or mangrove health without having to build models from scratch. “It’s intended to democratize access to this kind of technology in a no-code kind of way,” Patrick Beukema, the OlmoEarth lead at Ai2, told Mongabay in a video interview. The motivation behind building the platform was to drastically reduce the time scientists spent parsing through humongous volumes of data to get meaningful information from it. “What we set out to do was to flip that on its head and really go from them spending months to literally days to get the same sort of information,” Ted Schmitt, senior director of conservation at Ai2, told Mongabay in a video interview. Mangrove tree rising out of crystal clear turquoise water on the tropical beach of Havelock Island, Andaman Sea, Andaman and Nicobar Islands, India. Image by Vyacheslav Argenberg via Wikimedia Commons (CC BY 4.0). Beukema and Schmitt spoke…This article was originally published on Mongabay

From Conservation news via this RSS feed