General Electric Launches Cloud Service Customized for Industrial Application Data

The hype that has frothed up around the idea of the Internet of Things has been amazing to watch and we’ve long wondered whether the eventual reality will bear any resemblance to the futuristic fantasies. In the consumer world it remains an open question whether people will willingly embrace having every device in their vicinity sending and receiving data, along with the privacy issues that arise, or whether the available bandwidth will be sufficient to carry all those signals even if most of them amount to real-time trivia.

All along, however, the one area of life where massively interacting devices and sensors has seemed to make sense is on the factory floor and in infrastructure systems. What companies such as General Electric (GE), Shelton, Conn., call the Industrial Internet of Things (IIoT) is a different realm of data from the kitten videos and fitness trackers familiar in the consumer world. The data handling infrastructure also has different needs.

GE moved this week to stake a claim to handling that industrial data in an ambitious new program it calls the Predix Cloud. GE proposes that the data coming from industrial machinery, much of it generated by sensors monitoring machine performance and environmental factors to provide predictive and diagnostic tools and signals, is of a different sort and requires a different kind of cloud, one built exclusively to capture and analyze machine data.

“The data we’re storing is very different from a Facebook picture, which is pretty well defined,” said Harel Kodesh, vice president and general manager for GE’s Predix software platform at GE Software, in a post on the GE Reports blog. “Some of the sensors on machines may not be working properly and their data is dirty. It needs to be cleaned, normalized, compressed and ingested in a secure and efficient manner.”

In doing so, GE goes up against established cloud service providers including some of the most formidable technology giants – Amazon, Google, Microsoft, IBM and a host of others. GE Software as it exists today is a $4 billion business made up of software offerings from a variety of GE business lines, including things like analytics and control software for the company’s jet engines, power turbines, healthcare equipment and other industrial products. GE expects to push its software revenues to $6 billion this year. Many articles in the general business press questioned whether GE Software can take on those companies on their home turf, but it’s not at all clear that GE will have to compete with them directly.

The Predix Cloud is built around GE’s Predix software platform and will be tailored to handle and analyze machine data in a secure environment. It allows for development of third-party applications for processing data within the Predix Cloud environment, but also the ability to strictly silo data according to regulatory requirements, according to GE.

Such security concerns are a key issue with the IoT. The data passing into and through the kind of cloud GE describes would be the ultimate motherlode for malicious attacks aimed at control or disruption of critical infrastructure ranging from power grids to industrial and chemical processing plants to transportation systems and so on. If GE can deliver the level of security it proposes, it should make a place for itself in the land of clouds, but technology press seems to be reserving judgment.

“By inviting in some of the most high-value hacking targets, GE may be creating the most tempting jackpot in the history of cloud computing,” Klint Finley wrote in Wired. “GE is well aware of the sensitive nature of the data and applications it proposes to host. In fact, it’s making security its number-one selling point. The big question is whether it can actually deliver on that promise.”

GE said it will start by building two data centers in the U.S. – one on each coast – and then pursue further expansion abroad, particularly in regions where there’s more industrial than consumer data flow.