It has been more than 30 years since automation went “digital.” Digital networks were around for some time before that, but the networking of “smart” field devices really began around 1987. HART, an analog-digital hybrid, was one of the first and simplest and is widely used today.
HART was soon eclipsed by fieldbus during the 1990s, resulting in the “fieldbus wars.” Many years of infighting proved that the market is the best decider. By 2000, industrial Ethernet was following in fieldbus’ footsteps, in an almost identical number of variants. We have a variety of networks, and have learned to live with the consequences.
Fifteen years on, Ethernet remains the main networking candidate for automation plants. Although alternatives like wireless and Gigabit Ethernet have appeared, they are really “subset” solutions. Some newcomers have turned up—IO-Link for example—but not much more progress seems to be possible at the Ethernet level.
But, technology cannot be stopped. What is next for networks?
The obvious answer is the industrial Internet, commonly referred to today as the Industrial Internet of Things (IIoT). But IIoT is not a new network; it is more an extension of Ethernet offering a way to communicate with the cloud and delivering a higher level of functionality.
IIoT relies on the plant Ethernet as much as any control or monitoring system, but in a new and interesting way. For instance, it works with almost any industrial Ethernet (and probably fieldbuses too in the future). It will be the greatest leveler of the network playing field in 30 years. Think of it as a single standard; I prefer to think of it as the first “shadow network.”
The precision, time-critical performance of real-time Ethernet occupies only a part of a network’s activity with ample bandwidth for non-time-critical functions. This hidden, shadow network supports IIoT with lean protocols designed specifically for transferring simple data at high speed into the cloud.
Importantly, this technology is not just open to big automation vendors, but also original equipment manufacturers and end users. With cloud applications burgeoning (many offered as a service paid for on an as-used basis), an IIoT future is available to everyone. It is potentially disruptive for traditional suppliers, but IIoT is an attractive option for anyone brave enough to grasp the opportunities now.
It is a time for fresh thinking and new business strategies for vendors and users. Small vendors particularly have an opportunity to install products on an as-used basis, the “equipment as a service” scenario. End users never actually buy a device; they simply use it while the device vendor monitors its performance. This lets the vendor refine products and performance as time goes by and supports extra functionality, such as predictive maintenance.
Yes, there are challenges and plenty of questions to resolve. For example, what happens to your data in the cloud? Who owns it? What about security? What does the data mean? How best can a company use it?
Industry 4.0 and the concept of cyber-physical systems are an enticing way IIoT data may soon be used in manufacturing. Another is artificial intelligence services that bring predictive maintenance programs (and much more) within reach of even the smallest machine builders.
IBM’s Deep Blue overcame the world’s best human players at chess, and more recently, Google’s AlphaGo faced the world’s best Go player and won! IBM’s Watson is making headlines today in automation. Major information technology suppliers—long seen as expensive and proprietary-solution driven—are moving rapidly to provide cost-effective IIoT packages suiting all pocketbooks and needs.
Despite all the “intelligence” in the world, none of these opportunities would exist if it were not for the shadow network. A sad truth is that digital networks are still not used everywhere. Organizations stuck in an analog world should look carefully at what they are missing. Without a shadow network to get data into the cloud and back, IIoT will remain closed, and all those opportunities may stay unavailable!
A version of this article also was published at InTech magazine.