Google and Intel team up on new AI weather-prediction tool

Published by:
Posted on: July 31, 2020 6:36 pm EDT

If weathermen can’t get it right, perhaps AI has a better shot

Artificial intelligence (AI) has the ability to receive enormous amounts of data to learn and analyze things in a way that humans can’t. Data collection is the fuel behind AI and, for weather forecasting, the gathering and processing of data from multiple sources in real-time are what allows a more accurate prediction. ClimaCell, a weather technology company, is changing the industry of weather prediction through AI micro-forecasting that is possible thanks to the company’s proprietary big data collection and analysis platform. This tool has been optimized to run on Google Cloud and enabled by Intel processors to offer some of the world’s most accurate micro-weather data.

The data that feeds the platform is collected from a variety of sources like wireless signals, connected cars, airplanes, drones and Internet of Things (IoT) devices. In order to gather this data in a more efficient way, ClimaCell runs its platform on the Google Cloud compute-optimized instance that features second-generation Intel Xeon scalable processors. This power is necessary for ClimaCell to be able to quickly process the massive amount of data that is being collected.

The power of AI is being enhanced by high-performance computing so ClimaCell is able to maximize its predictive services to deliver a holistic and highly accurate model of approaching weather conditions. This service is being already used by a major airline company, JetBlue, that accesses ClimaCell’s services to determine changes in weather conditions before they even pose a safety risk to passengers or a delay in the flight. JetBlue claims that ClimaCell’s predictions for micro-forecasting has allowed the company to save nearly $50,000 every month.