The Inevitability Of Software-Defined Manufacturing

From 2003 to 2006, I worked at a contract manufacturing company as a robotics engineer. I was the first software engineer hired by the company, an opportunistic hire by a visionary CEO who saw the importance of automation in manufacturing. The CEO wanted to reduce downtime in manufacturing, improve quality, and empower the folks on the factory floor to be more efficient.

That period of my career was a fascinating experience. I was coming from a Fortune 500 energy company, where I had been a database programmer working with many highly capable engineers on scaling large data models. In that environment, continuous improvement through software automation wasn’t aspirational, it was our explicit mission. I took the role in manufacturing because I wanted the opportunity to define and deploy a software roadmap from scratch. I learned a lot during that time. As successful as the company was, software didn’t really exist inside the company, aside from an arcane enterprise resource planning (ERP) system that was poorly supported and badly used. I did everything from programming robots by hacking into them (APIs in manufacturing equipment didn’t really exist at the time, and still don’t), to developing web-based workflow software, to educating employees on how to use not only the tools I built, but software such as Microsoft Excel. Along the way, I discovered these existential truths, so to speak, as they applied to manufacturing as a whole:

  • Everyone saw the benefits of automation and wanted to automate as much as they could
  • Very few people understood the role software played in automation, even at the highest levels of the company

Fast-forward 16 years and much to my astonishment, manufacturing as a whole has not progressed. In learning about Bright Machines and our opportunity space, I encountered a lot of the same problems I faced 16 years ago. In manufacturing, the bulk of inspection remains largely manual. Instead of data being collected across the factory to be analyzed, it is mostly hostage to a particular machine, or worse, not collected at all. The concept of transforming data across the factory floor into actionable information that enables building higher quality products faster is at best an ambition. From designing a product to setting up a job, there is very little automation throughout the process of building physical products. In fact, setup and deployment take weeks, sometimes months, leading to significant product delays. That’s just the beginning of the list of problems with manufacturing today. It’s a very long list indeed.

When we compare manufacturing to other industries that have not only embraced technology, but pushed its boundaries to innovate and succeed, we can’t help but wonder why this key economic pillar remains stuck in time. I posit that this is for several reasons. Manufacturing is a demand-driven industry with low margins. For most manufacturing companies, it has simply been easier to throw humans at any given problem, knowing that labor costs can be scaled up and down based on demand. At first blush, the calculation seems rational. Investing in sophisticated hardware powered by equally sophisticated software at an industrial scale carries a lot of expense, not only in upfront costs, but maintenance, ongoing upgrades, support, and so on. Then, there’s the problem of time. Customers want things manufactured quickly. Who has time to invest in equipment set up, calibrating machines, setting up networks, securing the data, etc.? Human workers, on the other hand, can be deployed on an as needed basis.

Except that things really don’t work this way anymore. Humans, rightfully so, decided they are no longer willing to work in arduous and monotonous jobs, leading to reports of “voluntary turnover rates exceeding 300%” in some parts of the world. That is an astonishing statistic. The cycle of innovation in industry has evolved and sped up so much that having the ability to not only deliver product in near-real time, but perform meaningful reactive as well as predictive data analysis is an absolute must in order to operate efficiently in manufacturing. The increasing sophistication of the products being developed require the precision of machine automation and the power of not only software, but artificial intelligence, for higher product quality and predictability.

Which brings us to today. Manufacturing is crippled by these pain points, but ill equipped to solve them, for the same two fundamental reasons I encountered 16 years ago: manufacturing companies certainly understand the value of automation but have not historically utilized software to implement automation. Manufacturing companies are, after all, not software companies. And until now, the lack of demand for software-defined manufacturing has led to few external companies that are actually positioned to deliver holistic software solutions that act as both immediate relief as well as business accelerators to manufacturing companies. Thus, we are at a critical inflection point where manufacturing as an industry is not only ripe for disruption, it is virtually begging to be disrupted in order to save itself.

So what does disruption look like in this space? In fact, what is software-defined manufacturing, really?  Is it artificially intelligent robots? Is it data platforms with state of the art business intelligence? Is it cloud-based platforms, remote deployment and troubleshooting, machine-learning driven analytics? These things definitely comprise the concept, but Software-Defined Manufacturing is really just the beginning.

Software-Defined Manufacturing will happen simply because it has to – it is the immediate cure to manufacturing’s already existing pain points. The true disruption in manufacturing will involve not disrupting manufacturing per se, but actually disrupting the very idea of software-defined manufacturing itself. And it will happen by industrializing all the technologies that make up software-defined manufacturing, deploying them as a scalable platform and delivering them to customers in a service-based model that grows and modulates with the needs of the business. True disruption is extending software-defined manufacturing to a hardware/software ecosystem, with minimal to nonexistent single points of failure, where multiple components work harmoniously with the single purpose of enabling fast, high quality delivery at lower cost; where data is assembled, collected and turned into predictive analytics, and artificial intelligence is effectively used to solve repetitive human tasks.

When will this happen? At Bright Machines, the call to innovation has been answered, and the transformation in manufacturing has already begun. For us, software-defined manufacturing is just the beginning, the building blocks of delivering an ecosystem of products that will not only disrupt but redefine an entire industry. It’s an extraordinary challenge and truly a generational opportunity. And it’s Day One of our own journey to change the world.

— Nick Ciubotariu, SVP software engineering, Bright Machines

‘Flex Ability’

There aren’t many women in charge of major EMS companies today. Indeed, a quick look at the CIRCUITS ASSEMBLY Top 50 shows there are none.

Women are among the leadership teams at some top companies. Creation Technologies, for instance, has a female CFO and chief culture and people officer (read: HR). But no woman has occupied the top spot at a major EMS since Gayla Delly suddenly and unceremoniously left Benchmark in fall 2016.

Which makes it all the more exciting to see Flex naming Revathi Advaithi chief executive of the EMS company. Advaithi, 51, has impeccable credentials. She is an engineer with an MBA, and was wooed to Flex from Eaton Corp., where she headed the
largest division of the $20-billion company.


Flex has all sorts of incentive to go after a rising star like Advaithi. Its big bet on the consumer market with Nike cratered, and the company’s stock went with it. The stock price dropped about 46% in the past year, much worse than the industry average (8% loss). Flex has wound down its Nike manufacturing operations in Guadalajara, taking at least a $30 million hit.

Industrial, on the other hand, is a growth market. Based on the most recent quarter, it represents a $6.6 billion a year business for Flex, and is growing in double-digits. Moreover, as an end-market it remains stubbornly captive, with estimates of just 20% EMS penetration. Advaithi could help unlock that potential. Her standing as a director with defense giant BAE, another mostly untapped market by Flex, couldn’t hurt either.

It’s refreshing and overdue to see a woman on top in our industry. According to Fortune, only about 5% of the Fortune 500 companies have female chief executives. Notably, those firms include GM, Lockheed Martin, IBM, Oracle and General Dynamics — all major customers of the electronics industry. If we are serious about opening the door to the next generation of engineers, we need role models with all kinds of backgrounds. When a woman looks for her future in the crystal ball, it’s only right to see a woman looking back.

Foxconn Whiplash

You are to be forgiven if you have whiplash from the multiple changes in direction of Foxconn last week. The world’s largest ODM and EMS company announced it was essentially pulling out of Wisconsin, scaling down its much publicized multi-million square foot campus in favor of a couple of small R&D centers. Then, after pressure from the US government, it quickly reversed course once again, saying the plans were still on.

Wisconsin taxpayers might feel a little like Charlie Brown getting the football yanked out from under him again. Not only does it look ever-less likely Foxconn will create anything close to the 13,000 local jobs it promised, but towns like Mt. Pleasant are already on the hook for hundreds of millions of dollars, the net effect of bonds it issued to pay for the initial construction. And if Foxconn doesn’t deliver, the state must pick up whatever the municipalities cannot pay back.

In any case, when it comes to Foxconn, actions speak way louder than words. Let’s wait to see whether anything actually gets built before commencing with the back-patting.

STI: Small Size, but Big Service

We are thrilled to announce STI Electronics as our EMS Company of the Year for 2018.

CIRCUITS ASSEMBLY selects one company each year for this distinction. In making the determination, we look at profitability and sustained excellence among their peers over a period of years. We also look at the company culture and uniqueness of their business or service model, and assess whether we think it is sustainable over time and across generations of management.

In that regard, the evidence strongly supports STI. It is downright stunning to see the breadth of services the Madison, AL-based company offers, especially given its size (under $25 million). On a daily basis, STI performs traditional SMT, box build, failure analysis (it has a complete lab), cleanroom die bonding, and operator training. And at the end of this month, it will introduce an OEM product it is developing with a third party.

Read CIRCUITS ASSEMBLY’s profile on STI Electronics at circuitsassembly.com and in the February issue of PCD&F/ CIRCUITS ASSEMBLY.

January Issue Highlights

The January issue of of PRINTED CIRCUIT DESIGN & FAB and CIRCUITS ASSEMBLY is now available. Our cover story, from Skyworks, looks at
as-shipped vs. mounted height for BGA and LGA packages.

When a component is surface-mounted to the motherboard, the x- and y- dimensions do not change. Not so for the height. LGA height increases; BGA height decreases. A new study shows how an increase in as-shipped thickness can enable greater electrical performance and reduce quality risk.

This month’s other highlights include:

  • Understanding schematics
  • Using Maxwell’s equations to solve transmission line problems
  • Determining Df and Dk tradeoffs among various laminates
  • Bare board x-ray inspection
  • Busting the myth of PCB design at the college level
  • A profile of EMS firm Green Circuits
  • Ten steps for achieving good DfX
  • The latest happenings among the IPC Designers Council chapters
  • And Peter Bigelow asks if smaller manufacturers outmaneuvering the big ones.

Check it all out here.

Jabil on Tariffs

Jabil chief executive Mark Mondello said what we’ve all been thinking about the US-China trade tariffs.

On a conference call with analysts, Mondello called the issue “a big deal.” He underscored how Jabil could benefit if customers start to move manufacturing from China, as the EMS is well-positioned with factories all over Southeastern Asia, including Malaysia, Vietnam, Singapore and Taiwan — not to mention Mexico and Eastern Europe. And he broke down the potential impacts:

  1. “If the tariff and trade issues get resolved, that’s great.”
  2. “If the trade and tariff issues create some choppy seas and a storm here and there, that’s really good for (Jabil), because … there’s nobody that has our scale that can move product around with the agility and the flexibility that we can and, in fact, we do that all the time.”
  3. “If the trade tariff issues become some nasty hurricane, it’s going to be bad for all.”

Indeed.

Steve Jobs’ Biggest Legacy?

The decision of Foxconn to enter the semiconductor manufacturing market gives additional heft to the premise that the US created a monster determined to swallow everything in its path.

As reported by Nikkei Asian Business today, Foxconn is working on a potential joint venture with its Sharp subsidiary to “invest” as much as $9 billion in the new plant, which would be the company’s first foray into IC development. (We put “invest” in quotes, because 1. the gulf between Foxconn’s reported investments and its actual investments tends to be oceanic in size and 2. in this case, the investment is reportedly coming from the Chinese government.)

Foxconn already is likely the world’s largest consumer of chips, so getting into the OEM business would cause reverberations among its major suppliers. Moreover, it returns us to the sad refrain: What is Foxconn’s end-game? The company dominates the electronics supply chain from boards to assemblies to box build, makes other components (connectors, displays, motherboards, etc.),
operates retail stores, invests in 5G … you name it.

Personally, I blame Steve Jobs. The iPhone was a revelation, for which Jobs deserves every ounce of credit he has received. But in looking for assemblers, he could and should have looked further than Foxconn. There simply is no major company in the electronics industry today that is more aggressive and yet has a worse record of worker treatment than Foxconn. I’ve worked in the industry since 1991. Foxconn remains the only company that I’ve ever received direct complaints from its employees about their treatment. (And that came from US workers. I can only imagine what their Chinese counterparts might say.)

And yes, I realize it was Michael Dell, not Jobs, who gave Foxconn and Terry Gou its entry into the US computer industry. But it was Apple that gave Foxconn its biggest stage, boosting the Taiwanese company from a third-party motherboard maker to a partner in the most revolutionary electronics device the world had seen to that point.

When criticized for his reliance on Foxconn, Jobs would fire back that the US didn’t have the engineers to build what Foxconn could build. But I don’t think it was an issue of talent, or availability. I think it was an issue of greed. Jobs couldn’t acquire the volume of talent needed at the price he wanted. Foxconn could.

And so that’s Steve Jobs legacy. Foxconn is a $150 billion company and growing. Its revenues are larger than any of its customers. And, being traded on the Taiwan Exchange, it has access to financial markets without the transparency of public companies in the US or Europe. A monster is present among us, and will eventually devour us all.

Patty and the Professor Flashback: Uptime Part 4

Folks, the adventures of The Professor continue … 

So far the meeting with The Professor had proven very valuable, John thought. He was anxious to hear the other suggestions that The Professor had. The Professor began to speak. 

“Changeovers are what really hurts ACME’s uptime and, hence, productivity,” The Professor commented.

Pete was surprised. “Even you were impressed with our system of having a white board to document the logistics’ status for each future job,” said Pete.

“You are correct,” responded The Professor. “However, a changeover takes you about 2-3 hours and you have one or two changeovers per line per day,” The Professor added.

 “We have a high product mix business. It’s what we do,” said John.

“The good news is that you can cut your changeover time to 30 minutes,” shared The Professor.

“How?”  asked John increduously.

 “By using feeder racks,” explained The Professor. “These racks allow you to set up the component reels for the next job while the current job is running. Admittedly they cost about $30,000, but they will pay for themselves in weeks. Right now you lose more than two hours per changeover loading the feeders onto the component placement machines. With the feeder racks, you just roll them and lock them in place,” said The Professor.

Pete moaned, “We already have feeder racks. We only used them once, because they stick on the carpet when we move them.”

This comment caused The Professor to groan internally, but he hid it well. He had noticed the frayed carpet near the component placement machines.

John was beside himself. “It’s a good thing we are not The Professor’s students……I don’t think we would be heading for an A,” he thought. John responded to Pete’s comment, “Pete, let’s get facilities to remove that rug and start using  the feeder racks ASAP.”

Patty listened to all of this with comical fascination. She had harassed Pete about using the feeder racks several times. While the meeting was going on she drew a sketch of The Professor, who is notoriously camera shy. Oh, and she decided on the restaurant, Aujourd’hui in nearby Boston. Maybe they can pick up a Red Sox game while they’re there.

Epilogue: Six months later ACME’s uptime was a respectable 30.4%. John never had to buy another line. The improved productivity enabled ACME to increase their market share.  Patty’s dinner and ball game were a complete success. She handled her victory modestly and she and Pete became best friends. Pete also joined the ranks of The Professor’s many admirers.

Dr. Ron’s note: I know that a story like this must seems too comical to be true. Every point and the associated uptime numbers, lost time etc, are all based on a real situation with no exaggeration. The Epilogue, however, is ficticious, as is the Patty/Pete friendly (?) conflict. The names have been changed to protect the innocent (guilty?).

What is your uptime??

Cheers,

Dr. Ron

Going Mobile

As those who view our websites on their phones know, we have rolled out new mobile versions of circuitsassembly.com and pcdandf.com.

The new versions are optimized for smartphones and tablets, and are designed to present news, press releases and, of course, our technical content in a much more user-friendly way. As always, we’d love your feedback.

Meng’s Rollup

The arrest and possible extradition of a high-rankling Huawei official should be of concern to anyone doing business abroad. It is bound to have a domino effect as other nations line up to wreak havoc on strategic competitors to their respective domestically based corporations.

Or will it? This has been standard operating procedure for China for years. Whereas Moscow specializes in kidnappings for ransom, that’s Finance 101 compared with the Bear’s doctoral dissertation. China’s motive is longer in range — and the detainment longer in duration. What prevents many countries from acting in such rash fashion is the inevitable broadside to their reputation. China doesn’t mind the public relations hit, provided its broader objectives are met. And that objective is complete control over its economy and security. To the Chinese government, gulags are a feature, not a bug.

Canada rolled up Meng Wanzhou at the request of the US government, which cited an unsealed indictment against the Huawei CFO. It is widely believed Huawei is a front for the Chinese government, in part because its founder (Meng’s father) is a former Chinese intelligence office. Huawei denies the charges, but the US knows of what it accuses: In 1999, no less an entity than the Taliban had approved a plan for Afghan Wireless Communications — essentially a front for the American government — to build out the phone and Internet system in Afghanistan. If not for infighting in the US government, America could have had the entire country tapped.

When contemplating these latest events, consider these issues:

  1. The US has not yet indicated why it wants Meng. (The early buzz is the company is shipping illicit gear to Iran, in violation of international sanctions.) What happens the next time an adversary decides to nab an American? Who has the moral high ground? Do we trust the government — any government — enough to take it at its word? Or is Meng a pawn in a bigger, as-yet to be disclosed play?
  2. What will be the cost to US businesses that do (or want to do) business in China?
  3. How far is the US willing to go in terms of disclosing what it knows about Huawei’s operations? Sometimes it’s more useful to allow the behavior to continue in order to monitor it surreptitiously. Also, alerting others could give them a leg up on determining where their vulnerabilities lie, and lead them to close those gaps.
  4. There’s a trade war ongoing between the US and China (really), causing several major electronics ODMs to consider relocating factories from both nations, not to mention higher costs to consumers. Some, like Foxconn chairman Terry Gou, think the effects will last for years. Will Meng’s arrest lead to further economic isolation and barriers among the world’s two largest economies?