Google Cloud today released environmental APIs for its Maps Platform that help software developers build tools for analyzing solar and air quality information using artificial intelligence (AI), machine learning (ML), aerial imagery and environmental data.

This move feeds into the hyperscaler's goal to help its customers and partners collectively reduce greenhouse gas (GHG) emissions by a gigaton by 2030, in turn lowering Google's scope 3 emissions.

"We do feel a responsibility to lead this transition to a more sustainable future by making information accessible and by driving innovation," Yael Maguire, Google Cloud VP of geo sustainability, told reporters. "With insights into how the planet is changing, we can build sustainability tools and support other organizations to do the same," he said.

Google Cloud already offers information on transportation- and building-related emissions, and AI is core to how the hyperscaler approaches climate challenges. "We really take that to heart as a key responsibility," Maguire said.

The hyperscaler's solar API began in 2015 as Project Sunroof, which estimated how much solar power could be generated and the associated cost savings based on granular geographic areas. The latest version of the API uses computing and mapping resources "to make detailed rooftop solar potential data available for more than 320 million buildings across 40 countries," Maguire said.

More specifically, an AI model trained by Google extracts 3D information about each roof's geometry based on aerial imagery, taking into account trees and shade. The API also considers historical weather patterns and energy costs for the area.

As commercial demand for solar panels continues to rise, "businesses like solar installers can see how much sunlight buildings receive and potential energy savings before even visiting the area," he explained. The technology could also support business owners and homeowners, like Maguire himself, when planning to install solar panels on residences or offices.

"We have been seeing solar get more and more accessible, so we saw a lot of opportunity and potential to bring this information and technology to businesses," he explained.

Based on the air quality layer added to Google Maps last year, the air quality API provides data on air quality for 100 countries, including pollution heatmaps and details on specific types of pollutants present. To stay up to date, the API validates and organizes terabytes of data per hour from sources like government monitoring states, satellites, meteorological data and sensors.

The API also takes into account live vehicle traffic data "to understand congestion data and car volume in an area, and we use ML to predict different pollutants in that area at a given time," Maguire explained. Health care, automotive and transportation industries can use this tool to share timely air quality information and help individuals adapt to the effects of climate change.

Although it's possible to determine local air quality without Google's API, Maguire distinguished the hyperscaler's approach by its "hyperlocal" granularity in terms of space and time. "The notion that you can now get information that is within meters of your location and that will update every several minutes" is Google's differentiating value proposition, he said.

"Some of the data sources that are out there, actually, are using some of this data that we're providing," Maguire added, though he was unable to share the specific organizations currently using the air quality API.