Edge computing & the eye in the sky

During my discussions with IT and industry executives on the internet of things, analytics and artificial intelligence — digital transformation for short — what silver-bullet application he expects technologies to yield. He immediately answered: “Video.”

I was especially fascinated by the example he gave of a developer’s recent trade show demo. He — I’ll call him Mr. Smart — said it depicted a factory with cameras installed at each of roughly 2,000 employees’ work stations. The goal, he said, was “to figure out if folks were deviating from the agreed processes.”

It all sounds very straightforward, in that industrial firms, like businesses of all types, have forever sought to standardize and automate their work processes to reduce labor, boost productivity, cut costs, improve profits, and so on. Now, it’s just including better measurement of human work processes.

Organizations routinely monitor and measure machines and other inanimate factory assets using sensors, and video serves the same purpose, albeit for human assets. Per Mr. Smart, the goal of AI-assisted video analytics is “to effectively use a camera to sensorize a person.”

It’s a logical extension of what Henry Ford and Ray “McDonald’s” Kroc did to observe and improve upon the status quo to improve automaking and burger-slinging. There is, however, something of a qualitative departure as research into AI seeks to get inside people’s heads to discern how they feel. Using artificial emotional intelligence, smart machines — software bots — have already been used to identify anger and frustration in doctor-patient relationships and customer service chatbots. By analyzing audio as well as facial expressions, movements, even posture, anyone, one has to ask whether tomorrow’s employees will have a right to be paranoid.

The enabling IT underpinnings

My tech-guru Mr. Smart isn’t a video expert but a leader in AI and a key proponent of developments at the crossroads of IoT, AI and analytics — enabling technologies that he believes finally provide the bandwidth for the kind of video applications in the trade show demo.

In particular, edge computing enables real-time analysis of data (such as video feeds) at the source (the factory in this discussion). In contrast, remote cloud services are better suited to larger datasets and deeper analysis (among many other things). In a recent story in SmartIndustry, Intel’s Chet Hullum and his team used real-time edge analytics to improve semiconductor manufacturing, because the cloud would have been “much too expensive.” (See graphic below.)

Intel IoT social callout for SI edge coverage
A little log-roll on the blogroll: Intel tweets my ‘edge’ coverage.

With IT advances and applications proliferating from edge to cloud, the eye in the sky really isn’t the limit anymore.


FURTHER DISCUSSION: It’s important to consider the changing relationship between human beings and technology, and the social impact of the changes. As I covered in “The Robots Are Winning,” credible sources say automation has eliminated more jobs than it’s created and destroyed the U.S. middle class. Do you dis/agree? Do you have sources to help me revisit the topic in greater depth? Contact me.

The robots are winning

AI disruption seems to have thinned the workforce for good. Now what? 


Where would we be if wheels were still square? If the Luddites kept machines out of textile mills? If the cotton gin didn’t decimate the pre-war Confederate workforce? Today, one has to wonder how many truck drivers will be able to find new work once driverless rigs rule the roads. 

With regard to the the long-term impact of automation general, I’ve been ambivalent for decades; gung-ho for progress but concerned about a gutting of the workforce. In the mid ’90s, I spent a day with an engineering leader who gave me a tour of his workplace, a refinery that stretched for miles. He showed:

  • Pride in his control network, which bore early AI enhancements
  • Sorrow for the loss of most employees at the plant and industry-wide due to automation, and
  • Resentment of the coming “corpocracy” in which individuals lose their humanity.

I shared my ambivalence in a column I wrote for Control magazine at about that time after a conversation with a senior engineer at another Big Petro firm folding 17 subsidiaries into one. I started the column ranting about corporate “destructuring” and “dumb sizing,” and ended with: “It’s hard to argue with the bottom line.” 

ReThinking robots

Likewise, I’m mixed about the impact of industrial robots, worth about $40 billion and pegged to top $71.72 billion by 2023. Companies choose to replace people when the technology is available because they’re a better-faster-cheaper way to go. I was saddened at the demise of ReThink Robotics, maker of Baxter and Sawyer, the industry’s friendliest collaborative robots, or cobots. (collaborative robots). And I was glad the company was quickly bought-up by Germany’s Hahn Group. Rethink, a small player compared to leaders such as FANUC, ABB and Yaskawa, failed not for lack of demand, but in large part due to technical issues, if this RobotReport postmortem is accurate.

Baxter (pictured) and brother Sawyer have a new parent in Hahn.

In better times (2016), I spoke with ReThink’s Jim Lawton, chief product and marketing officer, who extolled his bots’ rapid ROI (as low as 1.5 years) and quick installation time (under a month) as well as speed and flexibility. For instance, he told me of a Tier One automotive supplier that replaced 20 hours of manual labor day: “$25,000 for the robot, and a little bit for the grippers, and we’ve saved them $180,000 a year.”

Robots don’t call in sick, need healthcare, come in late or take days off. They do work when workers don’t want to, or can’t be found. And companies are bound by the need for competitive advantage and higher profits to use them if they’ll keep the shareholders happy. But again, the industrial workforce, once an engine of the middle class, is shrinking.   

The U.S. middle class: Doomed? 

Technology both eliminates and creates =jobs, but the former appears to be winning in the U.S. marketplace. 

As one Clorox exec said during a panel discussion a couple of years back, “every” manufacturing company is busy automating and “leaning-out” its lines. A Kraft Foods alum added that it’s commonplace for companies to replace upward of “100 people on a Lunchables line [with robots] picking up stacks of pre-sliced meat and pre-sliced cheese.” 

“Employment trends have polarized the workforce and hollowed out the middle class,” David Rotman, editor of MIT Technology Review wrote in the article, “How Technology is Destroying Jobs.” Since then, TechCrunch promulgated the paradox that “Technology is killing jobs, and only technology can save them” (2016),” The New York Times shared in 2017 “Evidence That Robots Are Winning the Race for American Jobs.

What to do? Thought leaders across business, politics and industry have since given credence to the movement for Universal Basic Income, a flat payment to every citizen, to address poverty and job losses largely incurred by technological advancement.

UBI has been advocated by the Brookings Institution, given credence by Fortune (no leftist-socialist totem), and promoted by Elon Musk, Richard Branson, and many in Silicon Valley, including Mark Zuckerberg. More recently, UBI generated headlines in 2018 in Chicago with a petition by city Alderman Ameya Pawar (@Ameya_Pawar_IL) to take the city Universal:

Chicago, future home of Universal Basic Income?  

There ain’t no “i” in TEAM

Henry Ford did it, McDonald’s did it, and now the collective, global technology hive mind is doing it: Standardizing business processes to advance the competitive mandate for greater productivity and profitability. I love to hear stories of happy employees, but below photo mesmerized me to the point of distraction. So I’m using it to illustrate a point:

The point — no offense intended to the editor who penned the caption! — is that companies can’t afford for their employees to be “themselves” in that we can’t have guitar players, poets or sewing circle meetings on the factory floor. The employees pictured are essentially identical, down to their garb (as required by sanitary food handling rules). Their jobs are the same, too: to comply with uniform standards procedures.

But if there’s no ‘i’ in TEAM,” there’s still something of a “we.”  The ranks will thin, but there will remain a critical need for creative human minds to solve problems, even at the line level.

Almost a decade ago I helped a manufacturing exec, Greg Flickinger, document a cultural transformation at food firm Snyder’s-Lance in Charlotte, N.C. His team reduced scrap 40 percent; reduced customer complaints 41 percent; and, among other good stuff, slashed production changeover time to save more than $300,000 annually.

The human-machine interface, writ large over time, points to great gains and equally daunting challenges. Let’s face it: As a species, we’ve got a troubling historic myopia, and an immediate need to reconcile longstanding issues relating to our technology and economy, or techonomy.



Retail analytics: Here’s lookin’ at you, kids

I recently reunited with FoodOnline.com to write a story on Big Data analytics in the retail food supply chain. The first bylines I had for that site were in 1999, when I was Editorial Director for that related sites — before the big bursting of the Internet Bubble. Accenture’s Stages of Analytic Capabilities

When the Internet was new, there were no iPods, let alone iOS, Android or Bluetooth-powered beacons; Big Data was just a gleam in its young Business Intelligence mother’s eye. And retailers had no idea what to really do with their Business Intelligence systems. Those who finally do, today, see BI as old news as analytics — predictive and now prescriptive — come into their own, powered by Big Data and cloud computing.

Today, as an exec from SAS told me, a shopper who stands in front of a Nescafé display for more than 10 seconds might just get a virtual tap on the shoulder, or rather a bzzzz in the pocket, with a coupon to get that package of coffee off the shelf and in to the cart.

My favorite interview in this story just might have been Nick Hodson, former head of strategy at Safeway Stores and current leader of the North American consumer and retail business practice of Strategy& PwC, who reminded me that the technology isn’t at all the point of progress so much as creative minds who come-up with new things to do with it. Yes, major retail marketers from Walmart and Nestlé may well fave a backlash over privacy concerns if opt-in/out issues aren’t handled correctly, but these times sure are interesting. Read all about it in my story, ” Predictive Analytics Helping CPGs Reach Individual Consumers.”