I'm looking at yet another iPhone charger cable slowly shredding itself near the connectors at either end. I'd think this was inexplicable, if I didn't know that cable jackets, shielding, wires and other internal cable layers and materials can get damaged if they're assembled at different tension levels. A silly mistake, unless Apple wants me to buy more cables? I guess you have to do whatever to reach $1 trillion in market capitalization.
Oh well, it's a minor annoyance, and one consolation is it crystallized for me a trip I just took down the memory lane of industrial networking, and a debate at its core. As part of Control's new Educational Video Series, I just put together the content for the "Process control basics: industrial networking" video. However, because my approximately 2-MB mental capacity is usually focused on what I have to do in the next 10-20 minutes, I had to go back through several years of our former quarterly Industrial Networking magazine, which was reinvented as Smart Industry magazine in 2015.
This process took me through the evolution of point-to-point hardwiring to fieldbuses, Ethernet, wireless and on to today's Internet protocol (IP) networks built on Ethernet, which is the foundation for all the Industrial Internet of Things (IIoT), cloud computing, big data analysis, virtualization and other sexy, software-based digitalization activity going on above. It's been amazing to watch the progress that industrial networking's developers, integrators and suppliers have made toward simpler, faster, more widespread, easier to program and more powerful networks.
Now, I fully recognize the evolution toward IIoT, virtualization and the cloud is pretty much inevitable due to the process monitoring, control and efficiency gains that go with it. No doubt, simple, secure, inexpensive and always-on networking will assist IIoT and the cloud as they march toward always-on process analytics and optimization. As I think I've said before, these systems are looking more and more biological, and appear to be widening the radius of the operations and business systems they'll likely bring together. Along with advanced process control, big data analytics, closer-to-real-time simulation and artificial intelligence, there's little doubt these tools will form the independent nervous systems of many process applications in the future.
However, while the software and digitalization party ramps up in the penthouse, it remains crucial to not forget the mostly Ethernet infrastructure, switches, cables, connectors and other components chugging away in the basement—and making everything above it possible. Just as all software has to run on a microprocessor somewhere, every IIoT, cloud service, virtual application and other form of digitalization has to run on a chip-filled server somewhere.
This means potential pesky, real-world bottlenecks, slowdowns, power interruptions and crashes that someone better have the know-how to find and fix when the need arises, as it always does. This is familiar territory, of course, for process control engineers and every other plant-floor professional, whose primary responsibility is uptime and who often labor unheralded in support of high-flying operations and performance goals.
What's more important? The operating, content communicated and analytical calculations performed, or the infrastructure they travel over? This is just like the industrial network debate I mentioned earlier. What comes first, the data or the wire? Yin or yang? Mind or body? Chicken or egg?
The answer is these two seemingly opposing sides go together, and are part of the same larger whole. It's only the inexperienced and uniformed that can't see it, and think of them as truly separate and opposed. This is mostly because of their different backgrounds and histories in previously separate technical disciplines.
Likewise, the evolution of industrial networking isn't about battles won or lost, but just about the most efficient path data can take from the field or process to the minds of its users.