PLC: An evolutionary matter

Has the PLC evolved over the past 10 years? Is it better? As we look back at how machine automation has changed since 1997, columnist Jeremy Pollard reflects on the evolution of the PLC.

Control Design 10th AnniversaryBy Jeremy Pollard, CET, Columnist 

The past 10 years have seen the demise of many third-party software developers that created programming software for the hardware vendors. They were there because hardware guys are hardware guys, and couldn’t write software.

Most vendors still put the power of the automation solution in a box. This way they retain their captive audience. You have to use their software, cables, and methodology. No evolution there.

Evolution is defined technically as “a gradual process in which something changes into a different and usually more complex or better form.”

As it involves the PLC, call this gradualism, maybe? Has the PLC evolved? Is it better? And to what form has it moved? Surely it’s more complex.

I’ve been involved with PLCs and automation for 30 years now. The PLC began as a very disruptive technology, a “PDP -11 in drag,” as Dick Morley likes to describe his invention. Morley is arguably referred to as the inventor of the PLC. The PDP-11 was a Digital Equipment Corp. mini-computer.

The PLC wasn’t too fancy, took a beating, had no off switch, and looked less than cool. Oh, and it was really, really expensive at the time.

What’s happened over the past 10 years to the venerable PLC and its supporting cast?

In my years of doing PLC instruction, I’ve always taught the notion that a PLC is a control system—it doesn’t matter what it looks like. It’s what it does that’s important.

A PLC is an electronic apparatus—with programmable memory, inputs and outputs, and a power supply—that executes user-entered instructions. Yeah, it was a computer, but we didn’t know it as such back then. We also needed a programming device to get those damned addresses into the thing.

We controlled discrete applications, then process applications, to the point where the lines blurred and the PLC morphed into a full-fledged automation controller. This is an important point.

Every vendor, be it Texas Instruments, Siemens, Struthers-Dunn, or Allen-Bradley, had its own way of doing things. Ten years has marked the demise of many third-party software developers that created programming software for the hardware vendors. They were there because hardware guys are hardware guys and couldn’t write software. That has since changed.

Ten years ago is when IEC-61131 was a very-young implementation of an older idea.

The proprietary nature of the hardware suppliers created a market opportunity for many companies. It also created a nightmare for users. Does the phrase “held captive” ring a bell?

It was because of this environment that a group from GM Powertrain started OMAC (open, modular architecture control). This group promoted the use of open standards and open connectivity. I believe it started the evolutionary path to where we are today.

Who would have thought that DeviceNet devices from many vendors would play together on the same bus? That in itself is an evolution.

The price/performance ratios of current day PLCs are 30 times greater than 10 years ago. The power of the micro-PLC range is equivalent to the mid-to-large range PLC of yesteryear. The difference is in I/O count. Remember, 10 years ago, 256K of RAM was expensive.

Small and distributed is the way of today. The good ol’ days saw massive PLC processors with big footprints for central control. Now the available communication layers allow for more distributed automation.

And, lest we forget about software, Windows 95 brought in a whole new era of computer-based programming software that allows us to do in five minutes what it used to take an hour to do. Ease of use is up a gazillion percent.

It’s not that we didn’t have software before. It’s just that it’s easier now. Windows allows for much more elaborate development platforms than DOS did. Learning curves are no longer steep. After all, “it’s Windows.”

We’ve evolved from automation development using desktops on carts to laptops, thankfully!

Oh, how I long for the days of sealed keyboards and cassette recorders. Until I realize I have it better now.

Morley thinks PLCs are evolving into an appliance, and that this appliance had better be Internet-enabled (really means IP-enabled). So much so that he has developed his own home-based PLC, the PERTASCO (Personal Task Controller). Ease of use prevails, and you don’t have to tell it who it is. He feels that most of what we do today is much too vendor-defined, and that the users are losing out.

The one common thread with the folks I talked to about this article was that the PLC now is the PAC (programmable automation controller), and it has its own vendor-defined definitions. Market differentiation is important, but functionally, they aren’t much different.

Mark the first point on the evolutionary curve here.

The PAC by definition is a PLC. It has everything that a PLC system has, but with differences.

“PACs have the ability to handle discrete, motion, drive, safety, batch and process control applications all from a common control platform,” says Lee Lane, Logix product manager for Rockwell Automation.

PLCs did this 10 years ago. Heck, even 20 years ago. The PLC-3 with Advisor graphics, or a Modicon 584 with Modvue graphics, were so cool! Do you remember the GA module?

Why would a new definition be required? Evolutionary, my dear reader, evolutionary.

Integration of components and of functionality is the single biggest contributor to the PAC movement. We could do the multiple-discipline stuff before, but it wasn’t pretty. Not that everything was an afterthought, but nothing ever seemed to work just quite right at first, and it was painful to make changes.

With the PAC architecture, Lee says, you can add an axis of motion in less than five minutes with software. That’s inarguable progress.

He credits the change towards a PAC environment to the requests of business, which wanted a better contribution from the factory floor. Data requirements have exploded, he says, which requires a more integrated solution with the IT guys.

Common data tags allow integrated hardware and software to easily communicate through the complete data pipe. This, Lane states, allows resources to be leveled based on the information they need.

Bill Black, controllers product manager at GE Fanuc, agrees, but for different reasons. He believes the PAC emerged as the new PLC because of a global need for reducing manufacturing costs and increasing the performance of standardized hardware architecture.

The past 10 years have brought many new tools to the marketplace, he says. These tools improve the engineer’s ability to design effectively and quickly. Thermal knowledge can be gained instantaneously, for instance, allowing a certain processor to be used in a certain footprint. Remember, there are no CPU cooling fans in an automation system.

Flash memory and surface-mount technology allowed GE to better design its non-dedicated architecture, which leads users to better software solutions, easier design and builds, and the ability to migrate to various platforms if required.

I would suggest that this platform also allows for system integration to be a top contributor.

Opto 22 has a somewhat different viewpoint to this “gradualization” of automation.

Opto 22 started as an I/O company. Rich Ryan, then president of Rockwell Software, stated in a session I chaired that every control company would become an I/O company and would stop producing processors as such. This was about 10 years ago. Opto 22 would say he was right on the money, although they do have PAC offerings.

Most vendors still put the power of the automation solution in a box somewhere. This way, they retain the captive audience. You have to use their software, cables, and methodology. No evolution there.

Benson Hougland, Opto 22’s marketing VP, suggests that the available processing power we have today allows the I/O subsystem to do the data handling on its own. Put the power in the hands that hold the process, if you will.

We couldn’t do this stuff 10 years ago. The glue that allows this to happen is Ethernet and a notion that Ethernet can be a common bus system or a universal physical media, say the Opto folks.

Hougland adds that the Ethernet I/O system can be an autonomous member of the control network, communicating directly with higher-level software applications.

Opto 22 thinks, as do many others, that the corporation of the future can’t exist without data flow.

Monitoring and control are different animals. Moving away from a central controller—for control—will be foreign to some, and scoffed at by some vendors, I’m sure.

So, the evolution of the PLC is the PAC. But it still is a PLC that has user-entered instructions, memory, and I/O. But, we can do a lot more with it, with ease, and still have our weekends free.

We evolved slowly, however. I believe the next 10 years will have a different pace. As Morley said, this all has to be Internet-enabled. This means Ethernet, IP-technology, web-based services, and the ability to have information anywhere at anytime for anybody. The current offerings from our control system vendors already are that, and a bit more.