GE wants to be a big data company. In a presentation in San Francisco Tuesday, the industrial giant announced a platform of products, including predictive software products, Hadoop-based big data software for ingesting and managing industrial data and a relationship with Amazon Web Services to share industrial data in public clouds.
All of this is key to its industrial intent vision, where connected sensors on machines talk to the cloud and companies harness the power of industrial data in real time to automate industrial processes. GE has estimated that connecting devices to the “industrial internet” could boost global GDP to the tune of $10 trillion to $15 trillion by 2030.
GE’s new data and cloud products
GE is building big data software called the Historian that uses Hadoop to manage time-series data to help industrial customers track their rising industrial data. GE’s Bill Ruh, VP of the Global Software Center, pointed out that industrial data is growing at twice the rate of other types of data. For example, GE generates about 5 terabytes of data a day in its labs.
The Hadoop part of the software allows the data to scale across multiple nodes, while the time-series component helps manage the influx of tiny pieces of data that comes in almost constantly. Time-series data isn’t huge, but it’s always coming in, adding up to millions and billions of records over a relatively short amount of time depending on how often it is collected.
The partnership with Amazon (Amazon CTO Werner Vogels attended the event, and he’ll also speak at our Structure event tomorrow in San Francisco) means the cloud giant will be the first cloud provider on which GE will deploy its industrial internet platform. It’s not clear yet, if Amazon will use GE’s Hadoop software in its cloud or if there are just some API links being built.
Update: Vogels and Ruh say that the relationship means that customers can ship and store their data designed for the GE Predictivy software and the Historian platform to Amazon’s cloud. So customers can do it on-premise and/or in the cloud. Pivotal will help build some of the software connectors that will make it possible for customers to use this data where they want without concerning themselves about where it is headed.
In some ways the demonstrations that GE showed off, are taking direct design strategies from consumer applications such as Facebook, and its software options, called Predictivity are designed to connect the data coming in from machines to people in user-friendly ways.
The goal behind all of these products is to bring the internet of things back to the enterprise realm. It’s nice to connect your home, but when you can connect power plants you can drive a lot more results in terms of energy efficiency and even cost savings. And because the money is there, we’ll see a lot of interesting software to solve the problems associated with managing, analyzing and running predictions against data.
“Now for the first time I think we’re going to see innovation coming out of the industrial space and not just the IT space, “said Paul Maritz, the CEO of Pivotal, a company that GE recently invested $105 million in.
I’ll update the story with more information after the event.