Posts Tagged ‘Manufacturing Automation’

In our prior post, we pointed out that China is on the verge of becoming the world leader in the production, sale and implementation of robots, with a stated goal of producing at least half the nation’s own robots for manufacturing by 2025.  The takeaway from that view, outlined recently in Bloomberg BusinessWeek might be that the world has much to fear from the ascendance of this wave of Chinese bots.

But a recent counterpoint to such a robot apocalypse offered by Greg Ip of the Wall Street Journal suggests that in fact, robots aren’t destroying enough jobs, fast enough.

In short, Ip points out that by enabling society to produce more with the same workers, automation like robots becomes a major driver of rising standards of living – in effect, a productivity boost.  While some say that “this time is different” because the technological change is so profound they fear that millions of workers will be out of work or at best consigned to more menial tasks… Ip says the evidence shows we’re moving in exactly the opposite direction.

He notes that while the U.S. “has many problems, job creation isn’t one of them.”  Job creation has averaged 185,000 per month this year and unemployment is down to a ten year low.  Wage gains are even up, slightly.  Ip says that “if automation were rapidly displacing workers the productivity of the remaining workers ought to be growing rapidly.”  Instead, worker output per hour has been dismal in most sectors, including manufacturing.

When slow-growing occupations are compared to fast-growing ones in data going back to 1850 (a proxy for job creation and destruction driven by ‘technology’), they find that churn relative to total employment today is the lowest on record.

Ip’s point is that the past was, in fact, much more ‘convulsive’ than today’s job churn.  American consumption he notes is gravitating toward goods and services whose production is not easily automated.  Societies increasingly are devoting “a growing share of their income to consumption in sectors where productivity [is] stagnant.”  The idea is that robots can replace fewer things that go into GDP than we think.

As examples he cites medical breakthroughs in new, more expensive treatments rather than cheaper existing treatments, and that child-care work has soared because parents won’t leave kids in the care of a robot.  Over the past decade, “low productivity sectors” including education, health care, social assistance, leisure and hospitality have added nearly 7 million jobs, whereas information and finance, where value added per worker is 5 to 10 times higher, have cut or barely added jobs.

His conclusion: We need a change in priorities.  Instead of worrying about robots destroying jobs, we need to use them more, especially in low-productivity sectors.  While robots may one day replace truck drivers, “it’s more urgent to make existing drivers, now in short supply, more efficient,” and to be more concerned about reducing the labor, and thus the cost of energy, rather than worry about jobs added in areas like solar power.  The alternative, notes Ip, “is a tightening labor market that forces companies to pay ever higher wages that must be passed on as inflation.  And that, he notes, “is a more imminent threat than an army of androids.”


Read Full Post »

chip fabIn a recent article from Bloomberg (“Focus On/Manufacturing” August 2015) the editors note that Intel and Texas Instruments have pretty much by now perfected the sci-fi form of manufacturing known as chip fabs – pristine, windowless clean rooms where some of the world’s most sophisticated chips are fabricated.  They run 24 hours a day, knowing that these multi-billion dollar plants could be made obsolete in as little as five years or so, as new technologies and capabilities leapfrog the old.

Now these firms want to show the rest of the world how it’s done.  The goal: an estimated (by IHS) $185 billion global market for the gear to automate industrial production.  To do so, firms like Intel and others are prodding companies to bring the IoT (Internet of Things) – physical objects embedded with electronics that talk to one another – into factories.

According to Bloomberg, on the assembly line of tomorrow, “industrial robots now caged off to prevent them from accidentally injuring human workers will move about more freely.  A machine outfitted with optical and motion sensors would be able to detect a hand that is delivering a tray of parts and adjust its movements so as not to inflict damage.”

Intel is also working to make technology for humans on the shop floor less error prone, including gloves that use chips to power a display on the wrist.  If an assembly worker correctly completes a task, a large green check mark appears; if not, a red crosses flashes on the screen – a useful accessory first conceived by a group of ex-BWM employees that could become a useful accessory in auto and electronics plants.

While autonomous robots may be years away, Ethernet connections are now making a real entrance onto the floor, while Wi-Fi, per Bloomberg, has hardly made a dent, meaning most plants don’t have the communications infrastructure – yet – to support the Internet of Things.  But it’s only matter of time.

This is partly by design however, it must be noted: Hackers cannot penetrate systems that aren’t connected to the outside, as the IoT, by definition, would.  As the head of embedded processing for Texas Instruments wisely noted, “The best way to protect your system is to disconnect it from the rest of the world… while the very idea of IoT is to connect it to the rest of the world.”

To allay these concerns, TI is pushing the development of multiple networks, so that a wireless link that transmits information on the internal workings of a machine can’t be hijacked to take control of the machine itself.  This is similar to the recent staged hack of a Jeep Cherokee that made news in July.

Still, as Intel is showing, advances in shop floor IoT have demonstrated benefits.  At one unnamed Intel facility, sensors and software correctly identified that pumps used to manufacture silicon wafers were about to fail.  The clue was found in irregularities in the pumps’ normal pattern of vibrations, detected by this sophisticated application of IoT.

For the near future, it’s thought that selling companies on the use of electronics for somewhat “discrete” functions, such as maintenance, will be a much easier sell than overhauling entire factories so every machine’s data can be parsed by computers.  But some do see a day when “factories will be able to talk directly to warehouses, which will be in communication with stores, which will allow companies to tailor production more carefully to demand,” as the article’s editors note.

The grand goal of this smart manufacturing is to create the ultimate supply chain – and if technology and computers have proven anything to us, it’s that it is only a matter of when, not if.



Read Full Post »