The comments below share some of my thoughts about the ever-increasing role of automation in manufacturing. These are interesting economic times, to say the least - hopefully my thoughts will stimulate the thinking of some readers as we all try to cope with the rapidly shifting financial and business landscape before us.

We all are told of how America’s manufacturing prowess has declined to where our economy is driven by consumption of mass-produced goods from the Pacific Rim. There is truth to this, and we all have our favorite reasons why. More troubling to me is the sense that we’re giving up on building and nurturing new manufacturing enterprises, because we no longer believe that we can manufacture anything competitively. There’s just too much cheap labor available overseas.

I beg to disagree. Yes, significant hurdles lie ahead, but the “inevitable demise” of American manufacturing is not inevitable at all. I submit that much of the pessimism comes from a lack of understanding and appreciating what is possible with today’s manufacturing technologies.

Mass-market products will continue to be manually assembled by low-cost labor in other countries, as well they should. This provides opportunities that raise people’s standards of living worldwide, and free up Americans to concentrate on the types of innovation we do best. Innovation such as developing clean, renewable sources of energy, and the measurement products and software for the decentralized energy networks to take full advantage of them.

Our future is not restricted to creating new technology, only then to see it manufactured elsewhere. Let me repeat, we are not priced out of the manufacturing game. Why do I say that? I seem to have been let in on something that has inadvertently been kept a secret; that is, the emergence of extremely inexpensive networked manufacturing and automation products, combined with open networking standards.

Automation costs on the low end have dropped approximately by an order of magnitude, to where projects can increasingly be driven from the ground up, no longer so expensive that they require upper management buy in from the start. I refer to this trend as “grass-roots automation”.

First, about the products: they're here today, and are far easier to set up and deploy than a generation or even a decade ago. They are already beginning to transform the ways in which we will manufacture products and deliver services, in the companies that recognize their potential and that are learning how to use them. Everyone knows how today’s hand-held electronic marvels possess thousands of times the computing power of older mainframe computers. How many are also aware of the degree to which this has also happened with shop-floor automation technology?

I have witnessed much of the development of computer based, data-driven automation over the course of my career. My first automated test system designs used HP-85 GPIB controllers programmed in HP Basic. Today, the instrument and computer costs to do the same job (with 100 times more power and sophistication) probably cost about 1/5 of what they did then, even before adjusting for inflation.

The control horsepower of a PLC chassis costing maybe $50,000 in the mid 1980s is now exceeded by small off-the-shelf integrated PLC modules costing as low as $300-$400, with an Ethernet connection to boot. Fully embedded solutions can potentially cut those costs in half again, with a bit of one-time engineering. For more demanding applications, your controller module costs will run higher, say $2000-5000 or even more if your needs are demanding. But this is a world apart from where things were a generation ago.

Second, and less understood - web standards: the emergence of platform-independent web standards and Ethernet protocols such as XML, SOAP, Ethernet/IP, Modbus/TCP, etc. They are built on top of the foundational Ethernet protocols and building blocks. Ethernet communication using these protocols coexists readily with the standard TCP/IP, HTTP, FTP, etc. traffic that is ubiquitous in a connected office or manufacturing operation.

(Note that I said that automation communication coexists ‘readily’, I didn’t say ‘optimally and safely’! Moving from simple connection to reliable and secure connection is absolutely critical, but need not be overwhelming. That’s where experience comes in to the picture.)

If you choose your internet-ready automation components wisely, the hardware and software overhead required in order to utilize these standards and protocols varies from no additional cost at all, up to very moderate additional cost. Even better yet, in most cases, expensive software licensing for proprietary automation networks is no longer necessary. Most developers will ultimately invest in quality development tools for productivity reasons, but often that is a choice and not a requirement.

What does this all mean?

Simple controllers can now automate your processes, and directly interact with the rest of your operation, in ways that used to be reserved for six to eight-figure automation solutions. They have low enough deployment costs to often allow funding out of manager’s expense or discretionary budgets, at least for 1st phase roll-outs and/or evaluations.

As with Big Automation, this new generation of products can:

• Obtain work instructions and recipes from central databases or company web services on demand, as orders reach a given station.
• Automate the collection and storage of shop floor data, or simply make specified manufacturing statistics directly available through direct connection using an Internet browser.
• Notify key personnel directly of alerts or alarm conditions through email messages.
• Increasingly perform these types of tasks without needing the assistance of dedicated Windows PCs with all their added costs, complexities and temperamental nature.

Even better, unlike Big Automation, implementing in a phased-in or design-build fashion is much less problematic, since the costs are so low. Intelligent “control nodes” are easily deployed in a standalone fashion, then later networked using standard Ethernet switches and routers. Industrially hardened networking gear is now available in abundance, at some increased premium to simple office networking products. Ethernet automation controllers are sufficiently modular that they are readily re-deployed should your plans change, as they most certainly will.

This is not an abstract set of ideas that look good on paper, but don't quite cut it in practice. Our clients’ total out-of-pocket costs for low-cost automation solutions, combined with my programming services (which may be significant) are still often much less than the cost a box of parts and vendor software licenses alone from a “real automation vendor”. It’s your money, and in general there’s less of it around these days.

Why is this “inadvertent secret” not better known and understood?

Certainly the information about these products is all over the internet, and the trade press carries news of the products and technologies, but I don’t see them doing much to convey the larger picture that is emerging, as I’m trying to do here. My take is that much of the industry is still dominated by a handful of established, big name companies, and this trend is not in their favor. Many of these companies still run on a manufacturing, sales and distribution model of 30 years ago; automation is one of a few industries that still do. These companies are by no means going away, but things are changing.

Another reason as to why you don’t hear more is that, (as vertically integrated vendors will be quick to point out), either you or your system integrator must be sufficiently skilled to identify and deploy these new capabilities. This requires a much better understanding of Ethernet networks and protocols than control engineers typically have had.

It will take time for these new essential skills to be disseminated through the ranks. But the potential increases in return on investment for cell-based automation are just too significant to pass up any longer.

Platform-specific protocols and software are still essential. I use them often, and even prefer them in PLC/PAC applications where a single vendor’s products are used throughout. Such solutions are simpler to deploy and maintain, and the manufacturer’s Ethernet communications protocols usually come with the package, even for the lowest cost solutions.

Please understand that low cost automation solutions using freely available web standards are not a panacea. You can still pay 5-20X more for your automation systems, and many will continue to do so, for valid reasons such as:

• Your automation controllers must work reliably in extremely harsh environments, over wide temperature ranges, and only certain products are up to the task.
• You have exacting process tolerances, where no compromise in measurement speed, precision, accuracy, and traceability is allowed, and you must use the best measurement devices available. Superior measurement performance usually means instrument-grade products, not automation controllers with analog options. (But then again, think open protocols between meters and PLCs – I bet you’ll find one available, and then you have the best of both worlds!)
• Your company has standardized company-wide on a more expensive platform, such as Rockwell Automation/Allen Bradley controllers and software, and has no intention of changing. Too much mixing and matching is not a good thing.
• You work in a heavily regulated industry such as medical devices/pharmaceuticals, and only certain automation vendors meet your exacting expectations for reliability and security, and for user authentication control that restricts unwanted access to critical program code and controller settings.

If your operation fits into one of the above, the new generation of lower cost automation products may not yet be for you, at least not in a major way. On the other hand, thousands upon thousands of small to medium scale manufacturing cells or lines, both existing and on the drawing table, are not so demanding that they can’t take advantage of low cost automation solutions. Many of these projects will never come to fruition without reasonably priced intelligent control.

Maybe you’re tasked with developing a manufacturing plan for your operation, and are evaluating “Made in the USA” vs. other alternatives. Perhaps your only chance to compete favorably against manual off shoring, is through the use of automation, but only if the start-up costs are not prohibitive. The message here to you is very simple - if you’ve looked at automation solutions in the past and the numbers just weren’t there to keep it stateside, maybe you should look again.

Whatever your needs may be, I work with automation products across the entire spectrum of price and performance. Most of them to perform quite well if they are used within the limits of what they’re designed to do. I’m not in any manufacturer’s “camp”. My job is to steer you toward the least expensive total solution that is adequate to the task, both now and into the future, at least as far as you can see there today.