StormImpact Predicts Outages and Damage to Utility Infrastructure
Back in 1752, when Benjamin Franklin walked into a thunderstorm to fly his kite, few people, perhaps even Ben himself, likely imagined where his weather experiments would lead. Storms and the power infrastructure of the United States have been connected ever since—and not always for good. Storms are the top reason there are power outages in the U.S.
Fast forward 270 years and hundreds of experiments later to meet the innovators at StormImpact, a new company leveraging spinout technology from The Ohio State University (OSU). StormImpact provides predictive tools that help utilities across the country minimize outages and reduce repair costs caused by weather events.
Thunderstorms, tropical cyclones, tornados, and winter storms create power outages impacting millions of people each year.
The U.S. power grid is made up of more than 7,300 power plants, nearly 160,000 miles of high voltage power lines, and distribution transformers that connect 145 million customers throughout the country.
Using the superpowers of cloud computing and customer models developed by global experts, StormImpact characterizes the risk of storm-related damage to the distribution of electricity—where most of us interact with utilities. These lines bring power into our homes and businesses.
“One of the ways that electrical utilities mitigate the risks associated with major weather-related power outages is through using decision-support tools to optimize their storm preparation and response,” said Dr. Steven M. Quiring, Co-Founder, and President. “Utility companies want to be prepared, but not over-prepared,”
Over the past 15 years, Dr. Quiring and his associates have developed StormImpact’s prediction platform powered by artificial intelligence (AI) to better predict weather-related power outages.
“Our data-driven machine learning models use past weather history and expansive predictive models produced by the national weather agencies,” Dr. Quiring said. “There is enormous history to draw on—historical forecasts, radar data, satellite data, and observations from weather stations that can be leveraged to understand what happened in past events.”
StormImpact marshals the forces of cloud computing and proprietary AI algorithms to help utilities pre-position crews and assets to the right place so they can quickly respond to alerts. Utilities need to start making those decisions days before a major weather event, and they need to be able to respond when things change.
Industry changes create real-time opportunities for StormImpact.
“Forces are working together to make ours a very timely business,” said Dr. Quiring. “Many utilities have aging workforces. Before those people retire, we are creating machine learning models to help capture the complex connections and relationships that the human brain can make.” Another factor, he says, are the regulatory commissions in most states that have responsibility for overseeing utility operations. There have been more storm outages; which creates more scrutiny.
“We also have advances in machine learning, cloud computing, and data from the utilities,” he said. “StormImpact is presently serving utilities in the Midwest, East Coast, and Southeast. Most contacts come from utility conferences and workshops. We are active in industry organizations and through research at Ohio State. Our best and most effective advertising is word of mouth from existing customers.”