Are we headed toward Singularity? Or just more PowerPoints?
Thus says IPC:
The Hermes Standard Initiative is happy and glad to announce that IPC has confirmed to recognize The Hermes Standard to be the successor to “the SMEMA Standard” IPC-SMEMA-9851, which has been the only globally accepted and broadly established standard for machine to machine communication in SMT with regard to PCB handover. Accordingly, The Hermes Standard was assigned an IPC naming code: It can now officially be referred to as IPC-HERMES-9852.
This strong acknowledgment means a lot for further enhancements of the global footprint and acceptance of The Hermes Standard. With The Hermes Standard having shown an impressively fast start from early drafts to worldwide awareness, it is now entering the second stage in global market penetration.
Customers and vendors alike will strongly benefit from this new level of standardization progress. Obstacles with regard to implementing “smart factory” features will be dramatically reduced or removed. Electronics manufacturers joining the movement will see another strong boost in their long-term competitiveness.
Thus says the Oxford Classical Dictionary, third edition, revised:
Hermes is an ingenious god, expert in both technology and magic. From his birth onwards, he was skilled in trickery and deception …he is ‘prince of thieves.’ Most often he uses his power in mischief, illusion, and mystery… He is characterized by a great variety of functions. Above all, he is a messenger god.
So, what’s the message?
Industry 4.0? Connected Factory Exchange (CFX)?
I’m just too jaded. A neo-Luddite. Analog guy in a digital world, but even so, please. What is so captivating about a barcode, er, a digital twin?
To be sure, I recognize marketing concealed in bright shiny objects. Life Experience 587.0.
I have some notion how confidence men and women practice their trade. Caveat emptor 255.0.
Slick packaging too. Helps to sell more stuff. Add groupthink. Put a sticker on your machine and call it compliant to some standard nobody’s ever heard of.
(“They” must know what they’re doing, after all!) Salesmanship built on lingering – but dwindling – respect for authority 101.0.
They tell us we’re going to have smart factories, using intelligent software.
Who knew there were dumb factories and dimwitted software before now?
Good thing we have “them” to set us straight.
Good thing we’re test engineers. Scrutiny is our business. Trust but verify.
Software is also our business. Software allows us to examine CAD files for testability; create working test programs for debug; apportion limited test resources (probes, sensors, etc.) on a universal, bed-of-nails grid; establish whether an assembly is suitably designed for JTAG testing; and fine-tune an automatic x-ray inspection (AXI) program before production commences. Highly customized software is indispensable to our mission. No one would accuse it of being dumb.
On the failure analysis side of the house, software enables us to take thousands of individual x-ray projections, recompile their two-dimensional digital impression (pixels) into their three-dimensional analogues (voxels), and analyze all their x, y and z-axis glory as to why the product they represent doesn’t work. Completely nondestructive and in living color. Like magic. Software enables that. Also, not dumb.
Intelligent software notwithstanding, sometimes it fails, prompting emails to tech support, which invariably marvels at the uniqueness of our problem.
Stumped. Despite those big-company resources, nobody seems to have a ready solution. Or cares. Our company, raising the question, is small and insignificant. Perhaps if we featured a fruit logo, things would be different.
“We’re going to have to write a ticket for this one.”
Consigned to oblivion.
Now we are confronted with Industry 4.0. Software on a grand scale, all linked. Machine-to-machine communication, for those with bushels of money. And really big customers. And continuous, high-volume manufacturing. And clients who never sleep and feed their insomnia and obsessions by viewing yield data at 3 a.m.
Did I mention you need lots of money for this?
Industry 4.0 is the Next Big Thing. Whole trade shows and reams of publications are devoted to hyping anxiety. Are you ready?
For what? Is this really something new and worth celebrating, or is it just stealthy planned obsolescence?
I can’t help but be reminded of the 19th Century classic by Charles Mackay, Extraordinary Popular Delusions and the Madness of Crowds. Pure coincidence, I’m sure.
Tulips. South Sea Islands. Dot coms. Interest-only mortgages. All had their times. We know where they went.
What is Industry 4.0? Ask 20 people who claim to know; you’ll likely get 20 different answers. But rest assured, most if not all have started their “journey” to Industry 4.0. Didn’t somebody once suggest that the journey is the reward?
It appears to have something to do with sticking lots of sensors in production lines, so you get petabytes of real-time data about the vital signs of those production lines. This data deluge informs you exactly when that production line goes off the rails. Process engineers get interrupted during the dinner hour when the casino lights on top of the pick-and-place machine flash urgent amber, merging to red. Danger Will Robinson.
It also offers nano-level traceability when things go wrong. Taken in the extreme, it will allow somebody to identify which BGA ball crack, on that board, in that SMT line, during that shift, at that fateful hour, caused the plane to crash. Eureka. Note the plane still crashed.
But good news: It provided all manner of actionable data!
This data torrent will be automated, as will process adjustments responding to the truth as the sensors tell it. Minimal human intervention will be required. Minimal human employment will be needed. What remains of a factory workforce will dwindle to a cybermetric vanishing point. All will run by itself. Singularity.
This is progress?
This much is certain: Industry 4.0 has spawned a thousand PowerPoints.
We are engaged (means) with making the world a better place (ends). It justifies everything. “They” told us so. Yet another messianic tendency from the tech world.
Publications refer to activities under the auspices of Industry 4.0 as “smart manufacturing.” Tools like additive manufacturing, 3-D printing, augmented reality, data aggregation, M2M, SCADA, cloud monitoring. What preceded this was, supposedly, unenlightened, semiliterate manufacturing. Data-driven insights are the differentiator. Like we lacked data before.
Here’s a question: If a product comes from smart manufacturing, will that increase consumers’ willingness to buy it? Is this a rehabilitation of what the economists call Say’s Law, in which supply creates its own demand?
Maybe we’d be better off calling the new way of thinking cool manufacturing. Practiced by the cool kids congregating on the far side of the metaphorical schoolyard. Sounds a lot like Fear of Missing Out. I got your real-time data right here, in every color of the rainbow, and monetized, to boot.
Why do we need this?
What are we going to do with it?
Who, above all, pays the bill?
Forty years ago, back in the Dark Ages, I studied something called linear programming in college. Originally developed during the Cold War as an offshoot of doomsday studies euphemistically known as operations research, the methodology takes a bunch of independent variables – call them inputs – and mashes them into a stack of simultaneous algebraic equations, manipulated in shorthand called matrices, and spits out a series of dependent variables – dubbed outputs – to solve vital problems in aircraft manufacturing, oil refining, drug dispensing and delivery, military strategy, cake baking. The method reduced human activity to an algorithm, providing the optimal, mathematically elegant solution to life-and-death problems. Consider: optimal warfighting.
Optimal for whom?
Today’s version, discharged to civilian life and dubbed Industry 4.0, sounds like a warmed-over edition of those Cold War simultaneous equations. Sharpen the variable sufficient to obtain the optimal real-time solution to the problem of production maximization. Sharpen them enough and one cuts costs. Cut costs enough and one maximizes profits.
The term “optimization” carries its own compulsion. It antiseptically describes a better state, desirable in itself, devoid of humanity. Saving one-third on 3% margins is a lot of money, if our plant is assembling 3 million boards a day.
What if you’re building one of something?
What does all of this have to do with the test software I was telling you about earlier? That’s the thing: I’m not sure.
What I am certain about is this latest peel of the techie onion is making some people money. Graphic designers and PR types, to be sure. Beyond that, I don’t know. Three percent margins still prevail. That doesn’t leave much left over. And pick-and-place machines with embedded sensors are rather expensive. ROI, anyone?
Lots of new tickets to write if it breaks.
I am sure our company continues to test boards the same way we tested them a year ago and, presumably, will be more or less testing them in much the same manner one year from now. Perhaps in greater numbers a year from now, with the same defects (reported by Industry 4.0 sensors). Rising levels of defects to be detected, powered by rising levels of automation, are good for business. Thank you.
Here’s an idea: Emphasize the design, or the essential service, rather than the data or the Lean manufacturing that birthed it. Then desire will create its own demand. As it should.
Like software that works. Which returns me to the beginning of this column. Why is it that so much we feed back to tech support travels a one-way path, and we never hear another word? Why is it so many questions stump the tech support folks?
Maybe they’re obsessed with Industry 4.0. Adding their names to those who come up with a new definition. Or a repackaging of what we know to sell better. We sure as hell don’t know better.
Do they? What if the data belched by the 4.0 firehose is wrong? Who will know? How will they decide?
They have it down to a science. That’s why They are in charge and not us.
Which makes us lab rats.
Hermes was a messenger. He was also a thief.
firstname.lastname@example.org. His column runs bimonthly.is president of Datest Corp. (datest.com);