The IIoT still lags behind the IoT and that may always be the case

July 11, 2018


The IIoT still lags behind the IoT and that may always be the case

Is the industrial community waiting for somebody to stand up and be the Google of the Industrial IoT sector?

IoT in the consumer space has taken off, from connected thermostats to smart lighting to exercise equipment. On the industrial side, it’s growing, but not nearly at the pace of the consumer market.

Why is that? The technology is certainly there. Heck, in many applications, you need far less to handle your industrial platform than is deployed in many of the common consumer applications. So, what’s the holdup?

One reason the industrial segment lags is that it’s being deployed to conservative industry. And then there’s the fact that you’re dealing with physical equipment, sometimes pretty dated equipment. If something goes wrong with that equipment, it could be a safety concern or cause an interruption in manufacturing, thereby disrupting the revenue stream. It’s the “if it ain’t broke, don’t fix it” philosophy.

Those that are more in tune with the technology look at it differently, seeing that you’re losing revenue by not adopting the latest technology. For example, let’s look at equipment that’s can be tied to the price of energy. In a perfect world, you want to run your machinery when the cost of energy is the lowest. If your system can dynamically poll the energy company as to when that is, you can realize a significant savings. That’s a savings from an operational perspective.

What makes this somewhat tricky is that there’s no singe database, a la Google, that the industrial sector can turn to for all its needs. According to Rich Carpenter, General Manager for Controls Platforms at GE Automation and Control. “There’s no database anywhere that will give you the kind of insight that’s needed to run a factory floor, but it has to come eventually. We can’t avoid it. We’re working hard to get to that level of experience with our own equipment. The analytics of the past were mostly drawn by the physics of the equipment itself. The analytics of the future must still maintain that physical component, but it’ll incorporate the user actions. For example, the more you record when an event occurs and the actions taken, the better that database becomes and the better your ability to provide insights going forward.”

That begs the question, is the industrial community waiting for somebody like GE to stand up and be the Google of the Industrial IoT sector? This is where it gets tricky and the answer could be yes, but could just as easily be no. It’s likely that GE would handle that database for customers using GE equipment. But don’t hold your breath waiting for GE to manage a database for its competitor’s equipment. That just doesn’t make good business sense.

“I’m not sure our competitors would want that either,” says Carpenter. “People in general are nervous about sharing too much of their data and until that information is shared freely, it’s hard to realize the dream of one common platform.”

The first step is that there will be multiple “Google-like” platforms. Eventually they may consolidate to one or some smaller number, with companies like GE and Siemens taking lead roles. You already see signs of vendors like that partnering with the likes of Microsoft, Google, and Amazon. But the scope of those arrangements is still somewhat fluid. You may even see some of the associated standards bodies, like the IIC taking a key role.

In a potential implementation where the database would reside, you could envision two control loops, say an inner loop and an outer loop. The inner loop is the traditional control loop, while the outer loop is where you run your analytics and have access to information that might be outside your private network. You pull in the information you need and bring it into the inner loop and take some action based on the information. That’s one way of keeping the network secure.

GE claims they can implement the loops in one box, using a quad-core processor with a hypervisor on top of it. The secure RTOS would run on one of the processors and an OS like Linux would be on another.

It’s no secret that all security issues must be addressed before any standard platform goes mainstream. Legacy control systems that may not have any connection to the outside world had no reason to consider security. To design it properly today, it must be developed in parallel from the ground up alongside the industrial system design.