![[ SINGLE USE ] Harry Campbell illustration for Computerworld -- IoT weather data](https://images.techhive.com/images/article/2015/11/single_use_harry_campbell_illustration_for_computerworld_iot_weather_data-100629297-large.jpg?auto=webp&quality=85,70)
After weeks of rumors of an impending deal, IBM has agreed to buy Weather.com and other digital assets from The Weather Company for its Watson IoT unit and its IoT Cloud platform.
Specifics of the deal, announced in late October, were kept under wraps; it's expected be completed in the first quarter of next year.
The move indicates that IBM is serious about pushing into the cloud and connecting the data stored there with data being collected via the Internet of Things. IBM's goal is to use the combined technologies -- and its Watson cognitive computing system -- to provide useful data to companies in real time.
Cloud data a draw
IBM is buying The Weather Company's mobile and Web-based products, though not the TV-based Weather Channel. (The TV channel will license forecast data and analytics from IBM under a long-term pact.) IBM is particularly interested in The Weather Company's dynamic cloud data platform, which powers its mobile app and can now handle 26 billion inquiries a day.
That platform is built to collect data at massive speed and scale and can analyze information from 3 billion weather forecast reference points, more than 40 million smartphones and 50,000 airplane flights each day. The result: IBM will be collecting huge amounts of global data for deeper, more relevant insights via Watson.
Companies using IoT systems will get a "significant competitive advantage as they link their business and sensor data with weather and other pertinent information in real time," said John Kelly, senior vice president for IBM Solutions Portfolio and Research.
IBM envisions using a combination of predictive weather analytics and real-time analysis of social media chatter to, for example, help retailers get goods where they need to go before severe weather hits.
To be truly effective, intelligent systems like Watson must be able to tackle complex questions and handle vast amounts of data simultaneously. With the purchase, IBM gets weather data and analytics tools it can use in multiple industries, as well as a robust IoT platform it plans to extend even further to collect data from any source or sensor-enabled device.
"When you are talking about A.I. [artificial intelligence systems], they need two things: They've got to be good at understanding questions, and they have to know about the world," said Andrew Moore, the dean of Carnegie Mellon's School of Computer Science.
Moore, a onetime Google vice president who worked on several of the company's A.I. projects, said he thought IBM's acquisition was a smart one and added that other tech vendors will likely follow suit and move to enhance their own systems with stores of data.
John Klossner Action, not entertainment
Joel Cawley, general manager of IBM's Information and Insight as a Service unit, said the company is looking to deliver "actionable" insights, not "entertaining" ones. Such insights would come from weather data being integrated into decision making. Data that's collected from countless sensors -- and from comments and posts on places like Twitter and Facebook -- can be quickly analyzed during storms to provide better insights for users such as retailers and even emergency management operations teams.
"Accuracy in the retail environment means you can do the forecast down to the store location," said Cawley.
The number of data sources around the world "is exploding," he said, thanks to recent declines in the costs of IoT sensors used to gather information.
The trick -- and something IBM's data scientists are focused on -- is extracting relevant information from all the noisy data those sensors generate. According to Moore, as more data flows in, "the chances of something going wrong go up dramatically."
Social media insights
That's why IBM wants to analyze social media posts: They can provide context for the sterile data coming from sensors. A combination of weather data with Twitter posts has more value than if the two sources were considered separately, according to Ryan Fogt, an assistant professor of meteorology at Ohio University.
The IBM-Weather Company deal builds on a strategic alliance that the two companies announced earlier this year. Under the previous agreement, IBM licensed The Weather Company's powerful cloud data platform and collaborated with its B2B division on various products and services.
IBM's efforts to augment its Watson platform go beyond just the addition of the Weather Company's digital assets. In another late October announcement, IBM revealed that six new partners will use Watson to develop new cognitive computing apps and services. Those partners are Engage, Macaw Speech, Opentopic, StatSocial, Vennli and Domus Semo Sancus.
Katherine Noyes of the IDG News Service and Computerworld's Patrick Thibodeau contributed to this story.