10 IoT Startups You Need To Know

The Weather Company, which is now a part of IBM, has a new commercial offering called Deep Thunder, announced this week. Deep Thunder combines hyperlocal, short-term custom forecasts developed by IBM Research with The Weather Company's existing global forecast model.

Deep Thunder will use historical weather data to train machine learning models to help businesses predict the actual impact of weather, IBM said in a statement released June 15. The name Deep Thunder had already been in use by a team within IBM. This new commercial offering, however, marks the first time IBM has used Deep Thunder as a brand name, and the first time The Weather Company's capabilities have been combined into the project.

IBM already had customers under its Deep Thunder project, including The Vermont Electric Power Company (VELCO), which uses the technology to integrate and predict the availability of renewable energy sources in the grid. As a result, Deep Thunder delivers for VELCO 95% accuracy for solar power forecasting and 93% accuracy for wind power forecasting, an IBM spokesperson told InformationWeek in an email.

"The Weather Company has relentlessly focused on mapping the atmosphere, while IBM Research has pioneered the development of techniques to capture very small-scale features to boost accuracy at the hyper-local level for critical decision making," Mary Glackin, head of science and forecast operations at The Weather Company, said in a prepared statement. "The new combined forecasting model we are introducing today will provide an ideal platform to advance our signature services -- understanding the impacts of weather and identifying recommended actions for all kinds of businesses and industry applications."

[Take a deeper look inside IBM's IoT expansion efforts. Read IBM's Watson Taps APIs, Learning Centers, for IoT Expansion.]

When IBM announced its acquisition of The Weather Company in October 2015, it said it would use the infrastructure as a foundation for its Internet of Things cognitive computing effort. The resulting work earned The Weather Company the No. 2 spot on the 2016 InformationWeek Elite 100.

The Weather Company already produces regional models that provide new forecast guidance every three hours. Meanwhile, IBM Research has focused on customized models for business customers that provide hyperlocal forecasts -- at a 0.2-mile to 1.2-mile resolution -- and take into account other relevant environmental data, such as vegetation and soil conditions.

The goal is improve the understanding of how a particular weather system might affect a business, according to IBM. The Weather Company will now integrate this IBM capability and broaden access to it on a global scale.

The combination of these forecast models into Deep Thunder will help businesses precisely predict how even modest variations in temperature could affect factors, such as consumer buying behavior and how retailers manage supply chains and stock shelves. The tool can also be used, for example, by insurance companies to analyze the effects of past weather events in order to assess the validity of claims.

Utility companies could use the tool to mine and model historical data about damage caused to power lines or telephone poles, and put that together with hyperlocal forecasting to predict how many repair crews will be needed and where they should go, IBM said in its June 15 statement.

The Weather Company's models analyze data for every location worldwide using more than 100 terabytes of third-party data daily. The company's network includes the Weather Underground, made up of more than 195,000 personal weather stations.

The Deep Thunder group at IBM has been able to work with other analytics-driven projects such as Smarter Cities, according to the company's spokesperson. For instance, the new IBM Research Center in Brazil and the IBM India Research Lab are leading a Rio de Janeiro project to anticipate flooding and predict where mudslides might be triggered by severe storms. 

A new city command center integrates weather data with other city information systems to determine the best response, including where and when to deploy emergency crews, make the best use of shelters, and monitor hospital bed availability.