IBM Is About to Become the Best Weather Forecaster Ever
In addition to Jeopardy champion and doctor, you can now add meteorologist to the list of hats IBM’s Watson supercomputer wears.
IBM confirmed today that it will acquire the digital assets of the Weather Company, the corporate parent of the Weather Channel and Weather.com. Terms of the acquisition were not disclosed, but the deal includes essentially all of the Weather Company’s assets other than the Weather Channel television station, including Weather.com, the Weather Channel mobile apps, the Weather Underground website and, perhaps most importantly, Weather Services International , a division that sells weather data to companies such as airlines and the insurance industry.
One possibility IBM has floated in the past is the correlation of retail sales trends with weather patterns to help companies make decisions.
The Weather Company and IBM may seem like an odd pairing at first, but IBM already has a partnership with the Weather Company to jointly sell its weather data services and incorporate that data into its cloud-based (no pun intended) Watson services. The acquisition will allow IBM the ability to take fuller advantage of the Weather Company’s data, plus control of a mobile app that’s installed on tens of millions of smart phones around the world.
The core idea of Watson is that users can ask questions in natural language—such as “what will the weather be like in San Francisco next week”—and the system will provide an answer. Although it started out as a proof-of-concept that just answeedr Jeopardy questions, IBM has invested more than a billion dollars to commercialize the technology through a cloud service that businesses and governments can use to analyze data and conduct research. The more data the Watson system has, the more questions it can answer accurately. To help researchers at the Memorial Sloan-Kettering Cancer Center, for example, IBM fed medical text books and journals into the system. More recently, it signed a partnership with Twitter to pipe every single tweet into into Watson.
“The whole mission is to provide better predictive capabilities in the hands of our clients,” says Glenn Finch, global leader of big data and analytics for IBM Global Business Services.
The specific examples Finch cites are fairly banal. For example, he says that car insurance companies could advise their customers to put their cars inside during a hail storm, which could ultimately save them millions of dollars in insurance payments if those customers actually comply. But he imagines more grand possibilities in the future as the company’s AI algorithms attempt to correlate data from disparate sources. “Imagine you mash up weather data and Twitter data and economic data [in a specific city]—you get this signal that produces unbelievable analytic power,” he says.
One possibility IBM has floated in the past is the correlation of retail sales trends with weather patterns to help companies make decisions. For example, certain types of businesses, might see sales spike right before a snow storm as customers prepare for the weather, while others will likely see massive declines of patronage during the storm itself. Given enough data, customers might be able to use Watson to make some informed decisions, such as how much inventory to order or how many people to schedule for work during a storm.
IBM will also feed data that it collects elsewhere into the Weather Company’s system, perhaps helping it make better forecasts. “I am 100 percent confident that we are going to uncover vast troves of data and improve the accuracy of our forecast,” says Weather Company chief information and technology officer Bryson Koehler. For example, IBM offers a cloud service that allows developers to upload and analyze data from various “Internet of Things” devices such as environmental sensors. Some of that data could find its way into the Weather Company’s models, if the owners of that data allow it.
While the Weather Company’s corporate data services are the most logical fit for IBM, the deal does more than just beef up Watson’s data pool. Since selling off its personal computing business in 2005 IBM hasn’t had much presence in the consumer market. Weather.com and its mobile apps, however, will give IBM access to a massive audience—the Weather Channel mobile app has been installed between 50 million and 100 million times on Android alone, according to its listing within Google’s Play Store. And the default weather app on iOS is Weather Channel-branded.
Once again, the use cases the companies are imagining so far are fairly simple. For example, Koehler says IBM could offer local or state governments the ability to offer severe weather alerts or other emergency information through an app that many of their citizens already have installed.
Although IBM is tight-lipped as to how much it’s paying for the Weather Company, the deal has been valued at around at least $2 billion, according to The Wall Street Journal. The company is now owned by a consortium that includes NBCUniversal, Bain Capital and and the Blackstone Group that paid at $3.5 billion for it in 2008.