Tag Archives: poka-yoke

Robot Day

I responded recently to a LinkedIn post regarding AI-assisted robotic recycling.  The sorting speed is so fast, we almost miss each sort in the blink of an eye.  Having observed this same activity attempted by humans —  and overlooking the upstream potential to avoid this kind of recycling mess at the source (the wasteful consumer) — I’m all over the potential to pass off these kinds of tasks to machines.  Humans doing this work must operate at a much slower pace, risk injury, and are not so precise as the AI robot.   And of course, humans must also deal with the stench of garbage; these robots, at least, have no sense of smell to distract them from their work. 

Today, smart robots are economically feasible for even small companies, and are increasingly deployed to work in concert with human counterparts.  Referred to as co-bots, collaborative robots, these machine counterparts share the work in factories, restaurants, warehouses, operating rooms and offices. Given the advancement of artificial intelligence, robotics and multi-sensing technologies, it seems that eventually, no job will be beyond the capability of robots; science fiction is becoming science fact.  With the recently apparent shortage of humans to fill jobs this Labor Day, the role of smart robots is on my mind and leading to some whimsical questions; not so much about doing the work, but improving the work:

Will smart machines, for example, become smart enough to solve problems?  Will their designers build into them what Frank and Lillian Gilbreth called “a motion mind,” i.e., will they examine their own motion, or perhaps the motion of other robot team members, to reduce waste?  Robots are subject to the same wastes as humans; e.g., in a poor floor layout a robot must travel the same distance to do the job as its human counterpart.   Will AI be imbued with algorithms to identify  better flow?  Will smart robots analyze and improve their standardized work?  Will they collaborate with humans and with each other to brainstorm (sort of) and test ideas?   Collaborative improvement presumes a harmonization of different perspectives to create a more robust solution than would be available from any single  contributor.  Referring to Masaaki Imai’s definition of Kaizen (my favorite), can it evolve into this?

In the absence of hands-on human experience of the people who do the work, what context will this definition have?  And, what facility can be given to smart robots to create the will to improve, the most basic condition,  according to Shigeo Shingo, for improvement.  So-called human qualities such as “constructive dissatisfaction with the status quo” are not necessary to do the job, but are fundamental to improving the job.  Joseph Weizenbaum, an early AI pioneer, argued that such qualities are not transferrable to machines.  While his AI psychoanalyst simulation, Eliza (Circa, 1966), was designed to mimic a therapist’s response to human discourse, Weizenbaum warned that “no computer, can be made to confront genuine human problems in human terms.”  This warning has been echoed ever since in dystopian science fiction literature and cinema.   How much longer will this be fiction? 

These whimsical questions may not be immediately relevant.  In a world where there is currently  a shortage of humans to fill jobs, robots offer a practical solution.   As I write this post, there are currently twice as many job openings in the U.S. as job applicants.  While much attention has been given to the acute shortage, the rate of U.S. population growth (Fig. 1)  has decreased steadily since 1990 and is nearly flatlined, with many states showing a net decrease in population.  No doubt, the people shortage has been exacerbated by COVID-19, but the issue is chronic.  

In view of this persistent need, what might we expect from the inexorable evolution of science fiction to science fact?  With the current pace of technical advancement and likely long-term labor shortage, what will dissuade employers from replacing most human labor with robots?  Rosie the Robot, after all is an investment, while Rosie the Riveter is a variable expense;  they each hit the bottom line very differently. The allure of a productive resource that is apparently tireless, reliable and requires no benefits may simply be overwhelming – almost too good to be true.  In the words of pioneer systems thinker Russell Ackoff,  

“Managers are incurably susceptible to panacea peddlers. They are rooted in the belief that there are simple, if not simple-minded, solutions to even the most complex of problems. And they do not learn from bad experiences. Managers fail to diagnose the failures of the fads they adopt; they do not understand them.”   

While AI and Robotics are surely not panaceas, applying them as a total solution may just be. I visited a company recently, for example, that utilizes smart robots to move inventory miles between multiple storage locations rather than improving layout to consolidate stockrooms.  An engineer bragged to me “We even programmed it to ride the elevator.” Dr. Shingo called this superficial improvement, automating waste rather than eliminating it. 

For those of us who subscribe to Taiichi Ohno’s idea that “95% of the elapsed time between paying and getting paid” is waste, the current labor shortage presents a huge impetus to develop an army of problem solvers to eliminate it – a workforce with the will and creativity to reap greater benefits from Lean. While the vision of TPS is human development, certainly this includes thoughtful use of IoT, just as it has included mechanization and automation.   But will Lean thinking inform the designers of AI?  Will they go to the Gemba to learn? Will they regard employees as the most valuable resource?  Will they be systemic thinkers?  Or will they, as Shigeo warned, just be table engineers or catalog engineers,  distant from the process and perhaps a little too enamored of the technology?    That story is not yet written.   Will it be as dystopian as the science fiction predicts, or a golden age for productivity and human development?   And how can Lean thinkers influence the latter?

What do you think?    Please share some thoughts.

O.L.D.

P.S. Speaking of human development,  I hope you’ll be able to join us on September 28-29 in Springfield, Massachusetts,  for the 18th Annual Northeast L.E.A.N. (Lead, Enable And Nurture) Conference.  Four great keynotes, four tracks of breakout sessions, a Lean Lounge with poster presentations, and a Lean After Dark networking session.  AND – if you can’t make it in person, all sessions will be live-streamed on Whova and recorded for later viewing.  You can check out the agenda here:  Conference Agenda

Everyday Collaboration

With GBMP’s 18th Annual Northeast Lean Conference on the horizon, I’m reflecting on our theme, “Amplifying Lean – The Collaboration Effect.”   The term ‘collaboration’ typically connotes an organized attempt by unrelated, even competitive, parties to work together on a common problem; for example, the NUMMI collaboration between GM and Toyota or the international space station.   In a sense, these types of organized collaboration are analogs to Kaizen events and significant organizational breakthrough improvement.   

Being a longtime proponent of ‘everybody everyday’ type Kaizen, however, I think the greater amplification to our continuous improvement efforts lies in our ability to work together in the moment to solve many small problems.  But, just as intermittent stoppages on a machine may be hidden from consideration, so too these on-the-fly opportunities for collaboration may pass without notice.   An example from my own career as a manufacturing manager sticks with me as I consider the importance of everyday collaboration:

Walking through my factory one morning, I overheard a heated discussion between John M. a product designer and Ann C.  a team lead from our subassembly department.  Both individuals had deep experience in their respective areas – perhaps 25 years each.  John was waving an assembly drawing for a particular part as they argued, and Ann was holding the component parts and an assembly fixture.  All the elements of production were present: man, method, material and machine (4M’s).  What was missing was collaboration. 

“If you’d just follow the assembly drawing, there’d be no problem,” John argued.  

“What?”, Ann shot back. “Do you think I’m stupid?  Why would I call you out here if that were true?” 

This was the general tenor of the discussion, each party defensively talking AT the other.  Specialization, necessary as it is, often creates invisible boundaries we commonly refer to as silos.  When any party ventures beyond those boundaries, it’s viewed as an invasion of turf.  As the argument continued, the resolve of each party only increased. 

I inserted myself into the discussion.  “Why don’t we observe the assembly process and drawing together?  I’d like get a better perspective on the problem.”   John and Ann reluctantly agreed.   What seemed to me like an obvious opportunity to understand was, for each of them, possible exposure that one of them would be wrong and lose face.  Philosopher James P. Carse refers to this interaction as a ‘finite game.’  Somebody wins and somebody loses.   I recall saying something trite like, “Aren’t we on the same team here?”   Truth be told, we weren’t.  At least, however, we were all in the same space observing the 4M’s together.

Ultimately, John and Ann began to attack the problem rather than each other and, in fact, pulled a parts buyer and a tool maker into the investigation.  Working together they uncovered a series of contributing factors involving each of the 4M’s.  No single perspective would have been nearly as effective.  The errant assembly problem was solved.    But more importantly, collaborative relationships were created.   James P. Carse would call that an infinite game; everyone wins. 

O.L.D.

PS This year’s Northeast Lean Conference will examine collaboration from every angle.  Top-down, bottom-up, horizontal, networked, virtual, intercompany, governmental and societal.  The “Collaboration Effect” touches every system and every interpersonal relationship.  I hope you can join us on September 28-29 (face-to-face or live-streamed) as we explore better ways to work together.  It’s just six weeks away – sign up today

Uncommon Sense

“Never let a good crisis go to waste.”  This advice, attributed to Nicolo Machiavelli, and later cited by Winston Churchill at the conclusion of WWII, resonates once again in the aftermath of the COVID-19 pandemic.   Widespread shortages of products, services, raw materials, fuel, equipment, transportation and people have shocked the system in our land of plenty, creating an almost universal burning platform. From manufacturing to healthcare to service and even the public sector, providers can’t deliver – this while costs are rising, unemployment is at record lows and customer demand is through the roof.   Could it be that this tsunami of challenges is driving providers to experiment with ideas that just a few short months ago were light years from top of mind?

One factory manager noted recently for example, “It’s just a no-brainer that when you can’t find workers, you can increase productivity by removing waste from the job.” Funny that this should be considered a no-brainer now, when the supply of employees is limited.  But I’ll take it.   The pandemic, it seems, has elevated the role of the frontline while casting a brighter light on the shortcomings of off-shoring.  In 2011, when the theme of GBMP’s Northeast Lean Conference was “Made Lean in America,”  several supply chain pundits shared with me in no uncertain terms that this was wishful thinking. It would never happen, they said.  Times change. 

For organizations that have traditionally responded to growth by acquiring more equipment,  telescoping purchase times for equipment have turned our attention also to making the best use of the machines we already have.   Practices like preventative maintenance and quick changeover address the scarcity of equipment and equipment operators. 

And, then there are the omnipresent material shortages.  At the start of the pandemic, many providers took the commonsense step to actually reduce supply, even as consumers horded whatever was available.  The resulting deficits have now placed keen attention on what’s made and when.  “Certain alloys are super scarce right now,” a manager of a machine shop related to me, “so we can ill-afford to produce any material that we cannot ship.”  

Referring to the window analysis in figure 1, in times of plenty, Lean improvement efforts fell more into the “known-but-not-practiced” category.  Let’s call it the mediocrity zone.   As a VP of Operations confessed recently, “Yes, we attended the Lean 101 workshops in the past, and dabbled with some experiments; but in reality, these methods are not a part of what we do.”  I hear the same message in every industry.  The props may change, but the half-hearted practice is the same.  

My teacher, Hajime Oba, used the word “commitment” to describe the difference between Lean dabblers and organizations that seize the opportunity to gain significant benefit from TPS.   With cautious optimism, I believe that commitment potential may never be greater than today. I have observed in the last six months the greatest resurgence of my career in interest to learn and practice problem solving and continuous improvement methods. This crosses every industry, but shares a similar burning platform, one that no one would have predicted before 2020.   Writing in 1988 about the Toyota Production System, Taiichi Ohno stated “The oil crisis opened our eyes . . . “   Or as Shigeo Shingo noted around the same time, “the biggest obstacle to improvement is the will to improve.”   Here we are more than three decades later.  Let’s not let a good crisis go to waste. 

O.L.D. 

P.S. Speaking of… GBMP’s 18th Annual Northeast Lean Conference is coming up in less than two months. Have you checked out the agenda yet? We are so excited about our theme – “Amplifying Lean: The Collaboration Effect” and sincerely hope you will join us for the 2-day event which will feature four terrific keynote presentations plus more than 30 breakout sessions, a dozen benchmarking organizations in The Community of Lean Lounge and the chance to network with hundreds of Lean practitioners just like you! Read all about it.

By the way – if you can’t make it to Massachusetts, don’t worry. All of the sessions will be streamed – LIVE – so you and your team members can attend from anywhere in the world! Register Today.

You’re My #1 Customer

I sat on the phone on hold this morning,  serenaded by Christmas music, interrupted periodically by a recorded message, “Your call is very important to us . . . “  As I waited, I mused on that scene from the Christmas classic, “Jingle All The Way,” where Howard Langdon (Arnold Schwarzenegger) frantically tries to power through the queue of waiting customers.  At the end of each call, he reflexively concludes with the expression “You’re my number 1 customer.”    The scene makes me chuckle because I’ve been on both ends of that queue many times.  I do believe that most of us really want every one of our customers to feel like #1, just as we would like to feel that way when we are on the receiving end.  We want perfect quality and zero hassles; and in the information age, we order today because we want it today.

Alas, while most organizations aspire to create that level of customer experience, their systems and policies make it very difficult.  Like factory inventory, customers must be placed in queues when they cannot be served immediately.   Lines at supermarkets, traffic jams, waiting rooms, and, yes, phone queues.  The invention in 1989 (not so long ago) of the auto-attendant was intended to improve efficiency by automatically directing calls; a job that older folks like me will recall was once done by a person.  Where the desk of receptionist once stood, there is now just a phone with a sign above it:  “If you know your party’s extension, please dial it now.”  

If you are calling from outside there are  further enhancements to improve the waiting experience:

  • A clarifying greeting. (“Please listen carefully to the following options, as our menu has recently changed.”)
  • An explanation. (“All representatives are busy serving other customers.” Or, “ Due to high call volumes. . . .”)
  • An apology. (“We’re sorry.  Someone will be with you shortly.”)
  • Music. (Who chooses that?  Improvement idea: Give the caller the option to choose.)
  • A marketing pitch. (“Rated #1 in service by . . . .” )
  • A message to let you know where you are in the queue.  (“There are 14 callers ahead of you.”)
  • An offer to call you back. (“Dial 1 if you’d like us to call you back . . .” )
  • Or the old standard. (“Please leave a message . . . )

An advanced auto-attendant may, in fact, intermix all of these responses – or you may be encouraged to use an app (“For faster service please contact us at www. . . .”)  Of course, the nano-second capabilities of the Internet do not guarantee an immediate response.  Here’s a screen capture of an online inquiry I made in February 2021 🙂

Shigeo Shingo referred to these enhancements as “superficial improvements” because they automate the waste of waiting rather than eliminating it.   All of the embellishments exist only because the connection is not available.  Ultimately, if the proper party does not pick up, as Eli Goldratt might have noted, we have just moved the bottleneck.    

The original auto-attendant concept was intended to improve the flow of the customer’s inquiry by quickly directing it to the proper party.  If we were to consider only the operational time, that might be true.   But, for a customer faced with a nested process of choices based on 10 phone digits, there are plenty of opportunities for mistakes, rework, and frustration.  For me, there is nothing more surprisingly delightful than to reach a real person like Howard Langdon immediately.  But I will confess, if you try to reach me by phone, you may hear: “That mailbox is full.”  (A little 5S problem.)

In any event, have a Merry Christmas and remember: YOU’RE MY #1 CUSTOMER.  🙂

O.L.D.

Looking for a last minute gift for the passionate Lean practitioner in your life? Look no further than ShopGBMP and our annual Holiday Sale!

Profitless Part Proliferation

leadwireI wrote a post a little more than five years ago about Variety Reduction Program (VRP), an amazing but little known product design optimization tool.  At the time I referred to VRP as an idea whose “time had not yet come.”  Last week, as I gave a short presentation on VRP, I realized that five years later its time apparently still has not come.  In the interest of creating more interest around this significant technique, the following post expands on my epistle from 2011 and provides a couple of tangible examples of that significance from my own experience.

First, I think the technique deserves a new, mnemonic and alliterative moniker:  Profitless Part Proliferation.   I suggest this clarification because the word “variety” has an unfortunate positive connotation in the sense of greater customer selection, and therefore turns off sales and marketing folks before you can explain that VRP is not about product line trimming.  That was my initial experience in my own company many years ago.  “Just another anti-customer maneuver by operations,” I heard.  In fact, VRP aka P3 is about trimming needless part variety and all of its associated costs (e.g. drawings, inspection, purchase orders, stocking locations, etc.)

Secondly, I would like to call attention to the false sense of profitability that is often created through the addition of new parts and assemblies.   Minimizing the functional cost of material (the one that shows up on variance reports) for a single product looks good on paper, but almost always creates huge overhead costs arising from complexity.  Engineers and cost accountants typically focus on the apparent profit from product X, but ignore the resulting system costs.   They can’t see the forest for the trees, so to speak.   The following two examples for common part commodities, one a purchased part and the other a sub-assembly, speak to this problem:

O-rings.  A project was initiated to examine O-ring specifications and dimensions – things like durometer, chemical resistance, temperature range, ID and OD.   The first thing we realized was that there was no single repository for this information.  Our computer part master record contained dozens of fields to support ordering and costing, but most important design information was squished unintelligibly into a description field.  After cataloging specs and dimensions for O-rings, we realized that twenty-nine different O-rings were stocked.  Our discoveries:

  • Our information system made it difficult for designers see what was already available when they were choosing parts. It was just faster and easier to go to a supplier catalog. An alarming amount of part variety arose simply from poor design tools.
  • Once we were able to view O-rings as a part type from a design standpoint, we realized there was considerable overlap in specs and dimensions. Of the twenty-nine O-rings we cataloged, we determined that all production needs could be handled by only five O-rings.
  • Of the five remaining O-rings, one had metric dimensions because of unanticipated tolerances with mating parts. Rather than deal with correcting the mating parts, a unique O-ring was selected as a “bushing.”  Incidentally, that particular new part required the addition of a new supplier.

The rub was that the most robust O-rings cost a few cents more than marginally acceptable specifications.  Cost accountants argued that using the most robust  O-rings would increase product cost, ignoring the additional costs of maintaining two-dozen unneeded parts.  In fact, as we were a low-volume high-variety producer, we pretty much had to order months of supply for every one of the different O-rings anyway.  Finally, engineers argued that the cost of an engineering change – particularly a drawing change – was too great.  “We have better things to do” I heard.   Fact is, engineers are typically not rewarded for fixing up old parts; they are recognized for designing something new. Ultimately, however, some concessions were made in the interest of experimentation and the O-ring variety was reduced.

Lead wires.  A more egregious example of Profitless Part Proliferation was the variety of lead-wire assemblies. As a manufacturer of electro-mechanical products, my company built thousands of different lead-wire assemblies to support perhaps three dozen product families. At one point we dedicated a full bay of ASRS storage to lead-wires.  Still, lead-wire assembly stock-outs represented a major cause of late customer deliveries. Lead-wires were cut and terminated in large batches owing to the long set-ups on the machine.  While working on set-up reduction of the lead-wire machine, a production team lead astutely wondered why many lead-wires differed by insignificant lengths, as little as 1/32”.  During a project launched to catalog the variety in gauges, stranded or solid, terminations, insulation color and material – and many other specs – we did in fact identify an important opportunity just in lead-wire length variety.  This variety, we suddenly realized, stemmed from a single statement regarding the length of the connection leads outside the end item enclosure.  Sales and technical literature read something like this “Lead-wire length:  12” outside enclosure.”  In fact, our customers would have been happy with “at least 12” outside enclosure.”   Twelve and one-half inches would have been fine, as would twelve and one-thirty second inches, and so on.  The authors of VRP advised us to be clearer regarding which dimensions should be fixed and which could be variable within a range.   Once the product specification was changed to reflect “at least 12 inches outside,” the number and type of lead-wire assemblies plummeted!  So did the stock-outs.

These are just two of many specific examples where parts proliferation was pointless and profitless.  Now, before you say to yourself, “Oh that would never happen in my factory,” I’d encourage you to choose a common commodity of a purchased or manufactured part, and investigate the variety.   Please share a story for our readers about your discoveries. (One lucky commenter will be selected to attend GBMP’s 12th annual Northeast L.E.A.N. Conference coming in October to Worcester, MA. I am delighted to reveal our four exceptional Keynote presenters will be: Art Byrne, John Shook, Steven Spear & Dr. Eric Dickson (not to mention the forty other educational, informative, motivational and fun breakout sessions).

Shigeo Shingo was quoted as saying “The worst waste is the waste we cannot see.”   Help us to see by sharing an example from your experience.   I’d hate to think that I’ll be reflecting again in another five years on an idea whose time still has not yet come.

O.L.D.

BTW: GBMP’s calendar of Shingo Institute workshops is jam packed through October. Check it out here and join us for a workshop (or two) soon.

lfxAlso, I’m happy to share that GBMP’s online streaming video subscription service which we launched in March and call Leanflix  is receiving terrific reviews. We are so glad that we have been able to provide convenient, low-cost, on-demand video training content to meet the varied and ongoing training needs of so many in our Lean community. If you haven’t checked it out, I hope you will set aside a little time this week to do so.

– Bruce