In 1998, when Paul Rogers started at GE, implementing optimization software at a coal-fired power plant was easier said than done. Management understood and worked with GE to develop the software. Within the plant itself, though, the vast majority of employees didn't know how to use a computer, let alone software, and were very suspicious of the system.
These days, says Rogers, now GE's chief development officer, the tables have turned. Smartphone-toting plant employees know firsthand how technology changes their lives as consumers — and they want to know why the industrial environment isn't like their home environment.
"They want to optimize equipment, and that's a sign that the world is ready," Rogers says. Put another way: "My daughter has radically different experiences about how the world works."
Industrial Internet Solving Obvious Problems and Unknown Ones
The tech industry sees the industrial Internet driving productivity and efficiency in a variety of verticals. In GE's case, it's the industries where it sells hardware as well as data management and analytics systems. These verticals include oil and gas, transportation and healthcare.
Each industry poses different challenges, says Rogers, who sat down with CIO.com earlier this year. The answers may seem obvious to technologists, but that's not always the case with industrialists. GE's aim, Rogers says, is to show industries how to first optimize assets and, having done that, optimize operations.
[ Commentary: The Industrial Internet Is the Next Great Economic Revolution ]
Oil firms, for example, used to "monitor" their wells by sending an employee out in a truck. Since most wells are in remote locations, and unmanned, it typically took three weeks to discover and fix pumps that had stopped running, Rogers says. For the "black gold" industry, this represented huge losses.
Placing sensors on the wells, or assets, solves the obvious problem: Companies know much sooner if a well runs dry. This also brought additional benefits, Rogers says — data from the sensors, stored in the cloud and analyzed by Web-based research and development tools, lets firms study the overall efficiency of an oil field.
The industrial internet has a similar impact in transportation. Here, again, the initial focus was an asset (in this case, freight cars). Firms started by optimizing individual routes — cresting hills, for example, to save fuel — before turning to entire networks, where trains can spend up to two-thirds of their time idling.
Big Data, Cloud Driving Industrial Internet
These operational examples help demonstrate the industrial Internet's ability to address what GE sees as $150 billion in annual inefficiencies in the various industries the company covers. Taking this "new approach to efficiency," as Rogers describes it, isn't about making individual assets more efficient. The real value is in improving efficiency across an enterprise — and that exercise requires big data and cloud environments able to handle "a ton of information."
Take gas turbines. They tend to be efficient as they are, so fomenting improvement requires a "hyper-sophisticated approach," Rogers says, running models with "incredibly complex algorithms." On a laptop, it would take three weeks for a typical query to be answered; distributing the same query among cloud-based processors performs a calculation in a fraction of a second.
In healthcare, meanwhile — where estimates suggest as much as $1 trillion in annual spending in the United States is redundant, unnecessary or fraudulent — Rogers says there are plenty of opportunities. These include inventory management, hospital bed management, talent management and the omnipresence of proprietary platforms that make it difficult for providers and patients to share data.
The industrial Internet often parallels with the consumer Internet, but there's one key difference. The data being shared in consumer-to-consumer interactions — dollars, preferences, names and the like — makes sense to the naked eye, Rogers says.
In industry, on the other hand, data points such as pressure, output and weather are both complex and contextual; the temperature reading on a gas turbine's exhaust means something very different than the internal temperature of a locomotive, Rogers says.
It's a challenge, to be sure, but getting there means using technology that already exists, not waiting for something new. "It's already here," Rogers says. "It's about taking that data and turning it into something meaningful."
Brian Eastwood is a senior editor for CIO.com. He primarily covers healthcare IT. You can reach him on Twitter @Brian_Eastwood or via email. Follow everything from CIO.com on Twitter @CIOonline, Facebook, Google + and LinkedIn.