Under what circumstances do we stop using PLCs and PACs? Will we ever stop? What creates the need to migrate away from them, and/or why will we stay with them?
These are things that make you go, “Hmmm.”
I was on a conference call with a network company who was trying to figure some of the future out. Good luck with that, I thought. They sent me a list of questions that they wanted to talk about and my first reaction was to forget about it. Our industry reacts in years, not months. Processes don’t change overnight, and products are only new and improved based on packaging. They sell this as a feature, but in reality there is less product in the new package.
But I digress.
Do you know what a Kubernetes is? Well, I didn’t, so, with Google by my side, I discovered that it is a computer application that is a container system for automating computer application deployment. Could part of our future be deployment of SCADA/HMI and PLC/PAC applications?
There is really minimal difference between the PLC/PAC world and the computer world, except for packaging. The newest ControlLogix PAC from Rockwell Automation has four environments as part of the hardware, which means that one of those environments can run computer-based software such as Windows 10. That means that the HMI, such as Inductive Automation’s Ignition or FactoryTalk View Machine Edition, could run locally on the PAC without the need of a secondary computer, much like Opto 22’s Groov.
I have heard from various visionaries about a software-based PLC that could run in the cloud and thought that, while this innovative, they are really just nuts. There’s no way in the world I would ever run control software over the Internet.
But wait. We run control networks over Ethernet, so where’s the difference? That’s for another column.
We have continually believed and developed applications for the least common denominator, which is the maintenance department at 3:00 in the morning. This is meant as no disrespect to those who man those departments, but more of a mention that the job is the hardest on the planet at 3 AM. In order for them to be successful with minimal support at that time in the morning, they need to be very familiar with the systems that are in place.
Typically the technology changes are incremental, not orders of magnitude, because of the 3 AM demons. The engineering person will develop and design based on not wanting to be called at 3 AM to help solve a problem.
Note that remote access and remote monitoring with the Internet has moved the goal posts, but no one want to have disrupted sleep.
Our systems are going to become much more complex as we move forward. The PLC/PAC will still be the mainstay of the control spectrum, but, based on the questions I was asked, there are people out there who want to play in our sandbox and take things to a totally different universe.
One question I was asked was whether customers in general feel comfortable in using Raspberry Pi single-board computers to run single-source applications? My answer revolved around fixing versus replacing.
A maintenance electrician will typically know little about fixing a Raspberry Pi device with I/O and communication outboard modules. But he will understand that it has failed if told or has the ability to discover it has failed, and replace it with another black box. Could PLC/PAC environments come to that? Will knowing the programming software become redundant? While I don’t think so in the short term, where we might be in 10 years escapes me.
There was a large emphasis on IIoT and all things IIoT. What are they using and what use cases are most prevalent? IIoT is not concerned with refrigerators and stoves here. They are stand-alone valves and sensors that have the ability to be edge devices as well since they can compute and communicate with upward mobile platforms such as PLCs but also the cloud.
Can a PLC provide the analytics required for this modern-day phenomenon called big data? Do we need SCADA systems to provide additional use cases along with simple graphic interfaces?
Or will the PLC/PAC develop into an analytical machine along with control? The aforementioned ControlLogix can run analytical software along with HMI code with the PLC program all in the same box at the same time.
Real-time analytics will be a buzz term for years to come because we all want to know right now. Five minutes from now may be too late.
I fear that the role of the PLC is going to become so much more important than it is now, which will take a better-trained maintenance department, or at least a better troubleshooting environment, which may include the engineering department as a core component of that environment.
We are moving fast. Long live ladder logic.