Shiny new things and improvement programs

Michel Baudin

Mistake-proofing election ballots

Kevin Hop

How do I hear the voice of the customer?

Preston Smith

The lean supply chain – sizing up the opportunity

Jim Ayers

MMTI/NWLEAN workshop and seminars in Portland, OR, 3/5-9/2001


Shiny new things and improvement programs

by Michel Baudin,

“I want shiny new things,” said my 7-year old son who was tired of hand-me-downs from his big brother. It didn’t matter that the item in question was in perfect condition and met his needs; he just wanted something new. When it comes to improvement programs on the shop floor or in the office, adults behave the same way, and the change agents who promote, facilitate, steer or simply root for success have better take heed of this fact. Regardless of how successful they are, improvement programs should never be viewed as permanent features of the plant organization, or as a career path for support personnel. After five years, whatever enthusiasm they were able to generate is gone, and it is time to move on.

In the early 1980’s, many American companies started QC circle programs. At that time, QC circles were to quality management as Kaizen events are to lean manufacturing today. As the only element of a comprehensive approach to be sufficiently new, different and spectacular to command the interest of the mass media, QC circles were embraced by American executives as the panacea to solve their quality problems.

While falling short of these lofty expectations, QC circles worked. Operators volunteered to participate, were trained in the “7 tools of QC,” teamwork and brainstorming, and were able to solve some problems. Yet, by 1985, QC circles had become passé. Nobody launched these programs anymore, and those that existed were allowed to wither.  The consultants who had been facilitating them moved on to other things. Now, the concept is so dead that an executive who would propose it would sound like a character in “Blast from the past.”

The kind of problems they were set up to address still exist, and there is no reason to believe that, technically, QC circles would be any less able to solve them now than 20 years ago. But we’ll never know. QC circles are  no longer new and shiny. They are stale, and no one wants to get involved with them anymore. In this movement, the faddishness of American managers and the consulting industry's need to constantly "shift paradigms" feed off and reinforce each other. 

Not everybody, however, plays that game. Some Japanese companies, such as Toyota and Honda, still operate programs respectively called "Problem Solving Circles" and "New Honda Circles" in their American plants. Not being slaves to fashion, and having programs that worked, they just kept them in place. NUMMI, the Toyota-GM joint venture in Fremont, CA, today has about 100 circles in activity, for a total workforce of 4,900, which means that between 1 in 5 and 1 in 10 employees is participating. And NUMMI management still views this system as the main vehicle for continuous improvement. 

These circles have much common with the old QC circles: they are groups of five to ten volunteers who meet once or twice weekly to solve problems identified by management and receive technical training. The difference with QC circles is in the content of this training, which, instead of the old "7 tools of QC," covers a variety of aspects of lean manufacturing.  Yet, there is anecdotal evidence that these circles do not work as well as they used to, simply because, after 15 or 20 years, fatigue set in, and the facilitators complain of "participation burnout." 

This suggests that, while many American companies pulled the plug too soon, Toyota, NUMMI, or Honda may have stuck with the program too long. At this point, tweaking the process is unlikely to rekindle operator enthusiasm: something radically new is needed. It also suggests that, regardless of their intrinsic value, currently fashionable approaches like Kaizen events or Six-Sigma black belts will also have a finite life. 

If we forget about fashion, the rational approach is to start from the problem at hand:

  • Is it continuous improvement or a breakthrough? 

  • Does it affect one aspect of the life of the entire plant, or multiple aspects of one small section of it?

  • Does it require a team of 5 volunteers on a part-time basis, or a full-time task force of 50?

  • What skills, knowledge and resources are necessary to solve it?

  • To whom can we delegate the responsibility and be confident that we can live with whatever they come up with?

The appropriate team and structure should then fall out naturally from the answers, But that will only be a technical solution, which won't matter if the "right" people have been there and done that too often already. Like my 7-year-old, they want shiny new things.


Takt Times

Technical bulletin from MMTI

396 Shasta Drive, Palo Alto, CA 94306-4541

Tel: (650)856-8928 FAX:(650)858-1873

Michel Baudin Editor

Copyright 2000



















































Mistake-proofing US elections             

by Kevin Hop,

What a surprise it has been to US citizens as well as our friends around the world how messy our whole election process has turned out to be! It was less of a surprise, however, for practitioners of lean manufacturing, who know what happens when you don't mistake-proof your processes. The bigger surprise is that is has been in this current state for at least the last 4 decades if not longer. It was clearly not designed to have a quality “reject “ rate in the parts per million reject levels but rather in the percentage reject levels. Between 1% and 5% of votes are not being counted in many cases.

First let's look at the infamous Palm Beach County butterfly ballot and wonder what kind of quality we might expect out of a manufacturing process with instructions at this level of clarity (See Figure 1).


Figure1. Top view of the Palm Beach County ballot

The layout is even more confusing when seen at an angle by a voter in a booth, as shown in Figure 2, with the arrows pointing to spaces between holes. 

Figure 2. Palm Beach County ballot as seen in a voting booth

Obviously the ballot was not tested before being used in the election. This would have revealed its flaws. Have a million people use it, and the results are predictable, Figure 3 shows the differences  in the results between Palm Beach and the other counties in Florida.                                                                                                                                         

Figure 3. Buchanan votes in Palm Beach County versus the rest of Florida




But this situation pales in comparison to William Rouverol's Votomatic punch card system used in the US since 1963. It has been known for a long time to be inaccurate, and some states have even banned it. Incompletely detached “chads” are the most common problem. This comes from voter technique, variability in the card chad die cuts and chads piling up under the punch area from not being cleaned out properly. The incomplete chad then is read improperly by a punch card reading machine, which has little tolerance for the non-perfect chad removal. These votes then never get counted, unless it is close. Then we get into the whole abnormal manual vote process which then introduces human error into the system. Also new questions about voter intent and how to judge it pop up - to count the “hanging chad” or “dimple” or not? 

The answer is to mistake-proof both the voting process and the tabulation of the results. The voting process must be simple and easy to understand, whether you are 18 or 90, rich or poor, with a primary school education or a PhD. It has to be tested to make sure it has errors in the low PPM area,  not the current 1%  to 5% error range. The system that tabulates the vote must then  be reliable and able to cope with any variability introduced by the voting process.

The optical scanning systems currently being used in many states are a  short-term improvement but they still have their own problems, such as smudges, dirt, light marks, wrong pen, wrong type mark.  The optical scanner systems were actually used in about 42 of the 67 counties in Florida. Unfortunately the punch card or “Votomatic” system was used in the remaining 25 counties. Expect to see the punch card system disappear quickly in the US.

The next step will most likely be an interactive system that can provide a quick feedback to confirm that you actually chose the candidate you thought you did - similar to purchasing something on-line with Amazon. Below is an example of an interactive system offered by Hart Interactive Another system called the Patriot can be seen at These systems are already seeing some limited usage. They were actually used nationwide in the most recent Brazil elections. This type of system could provide near perfect voting if done correctly, with a good visual layout, multi-language selection, multilingual audio back up and the quick feedback to confirm the choice with the ability to easily change the selection before making it official.

Tabulation can also be near perfect as it involves no machinery – purely a digital transaction. Obviously, digital has problems too. A virus and hacker-proof system that also has good backup systems would have to be there to build confidence with the public and ensure accuracy. The biggest psychological hurdle would be the lack of a physical ballot trail. This seems like a big risk but remember that we trust our ATM’s and our on-line purchasing now - but it took a few years to build up that trust.

Shall we have a more accurate election system by 2002 and 2004? Technology is needed but it is not a panacea. We can use the latest hardware and software and still produce ballots as confusing as VCR programming today. As lean manufacturing specialists, we know how to provide ergonomic workstations, clear and unambiguous instructions, and mistake-proofing, both in conventional and high-technology processes. We can contribute this know-how to bringing the promise of democracy in the electoral system closer to reality.


How do I hear the voice of the customer as I develop a new product?

by Preston Smith,

Product developers have learned over the past decade that they should listen to the "voice of the customer" as they design their products. But this just begs the more difficult question: "How do I 'hear' this 'voice'?"

The good news is that we all have the voice of a customer inside of us, based on our life experiences over the years. In computer lingo, we come with a "default" voice "installed" inside of us.

The not-so-good news is that this default customer can be highly untypical, so it is a dangerous design reference. Alan Cooper, in his book, "The Inmates Are Running the Asylum," explains why so many computer products are maddening to use. (See our column: ) Many computer and software products are designed by engineers living in Silicon Valley, which is a highly atypical place. Everywhere you turn in Silicon Valley, you run into engineers, so the engineers who work there soon think that all customers are as capable of using their product as the people they encounter every day.

To design products for typical users and customers, take some time and effort to get out into their world. The best ways of doing this are creative and highly industry specific. For instance, Black & Decker sends their designers out with service technicians in vans that visit construction sites and do-it-yourself stores. Hewlett-Packard stations them in shopping malls to watch customers fumble with HP products. This isn't hard to do, but you have to make the effort to identify typical customers and then take the time to "hear" them regularly.

This article is adapted for Takt Times from Information Strategy: the Executive’s Journal Winter 2001 published by Auerbach Publications and The Handbook of Supply Chain Management published in September 2000 by CRC Press LLC and APICS.  

The Lean Supply Chain – Sizing Up the Opportunity

James B. Ayers,

  “Leanness” in manufacturing is now an accepted virtue undertaken as an “act of faith.” Few CEO’s fail to drop the term “lean” when listing their initiatives for improvement. Mostly, lean efforts take place on the manufacturing floor rather than in marketing and administrative functions – the “hidden factory” where paper, not product, moves.

Increasingly, those pursuing lean are not limited to manufacturing companies; the goal of achieving leanness has spread to banks, healthcare, and other service businesses. The “enterprise” approach rather than the manufacturing-only functional approach is increasingly common – with practitioners applying lean methodologies to paperwork as well as production processes.

There is also a dawning awareness that “lean” goals won’t be achieved by limiting the focus to the four walls of a single company. Many believe they compete on a supply chain to supply chain rather than a company-to-company basis.

This has raised the visibility of “supply chain management,” or SCM, disciplines. A sports analogy is found in the relay events in track and field. One overweight, under performing “athlete” on a relay team makes winning impossible -- even if teammates are world-class athletes. Despite this obvious conclusion, many lean improvement efforts still seek to squeeze blood from the proverbial turnip – pursuing internal shop floor cost cutting when the real problems lie in basically ineffective supply chains ill suited to the markets they serve.

Therefore, the two trendy movements – lean and SCM -- should be consolidated in any assessment of “Where should we go to be world-class in both lean and SCM?” The reason should be clear. Many overweight and under performing companies evolve to functional organizations with “one size fits all” supply chains. These are seldom efficient in terms of cost, service, or speed in meeting the competitive demands of customers. Newcomers, with little investment in outmoded processes, take over markets with superior supply chains. Building lean supply chains focused on customers is insurance against this threat.

The important first phase is to assess the opportunity for applying lean to the whole supply chain, not just the single company. We recommend eight beginning steps as described in the following section.

Eight steps for kicking off a lean supply chain effort

1.      Define market segments. List end-user groups. These are ultimate customers, not just the next company down the supply chain. Groupings may be by application of the product, location, volume, the supply chain to reach the user, or other characteristics. Next, form segments from the customer groups. One or more groups may comprise a segment. For supply chain design, combinations of groups into segments are particularly valid when they share supply chains.

2.      Map products to segments. Next, list products or product lines. A line may be a group of products produced in the same facilities. Or it may include different products that serve common markets. A common supply chain could also define a product line. 

3.      Identify supply chains. Identify supply chains supporting customer segments. We might find there are three basic variations. These include:

    • One supply chain for all customer segments. This is most common in smaller companies.
    • A different supply chain for each product line. This is often the case when each product is made by a different profit center. This is the most common in larger companies.
    • Supply chains organized by customer segment. (Seldom do we see supply chains organized around customer segments.)

4.      Describe the supply chain(s). With the supply chains identified, their configuration should be documented with flow charts and supporting data. Accompanying the flow charts should be metrics and other information that characterize current performance.

5.      Document physical flow. Fundamental to most supply chain representations is physical flow. Flow charts should show the following:

·        Echelons of the supply chain including major suppliers, manufacturing centers, distribution centers, warehouses, and customer segments.

·        Methods of transport between echelons.

·        Volumes of product flow in dollars, units or volume.

·        Cycle time for moving material in the chain, preferably broken out by echelon.

·        Inventory levels and bottlenecks along the chain, including those at suppliers and customers. This step will highlight candidates for applying lean technology.

6.      Document information flow. Information flow is an increasingly important component of the supply chain. Competitive advantage is possible by improving information flow. Among the elements needed to describe information flow are the following:

·        An information flow chart that shows where sales information is generated.

·        Decision points along the chain, including the people responsible for the decisions.

·        An inventory of information exchange tools used to plan and control the process. These can include the information technology domain of computerized systems or lean approaches like kanbans.

·        A listing of formal and informal contacts between supply chain participants.

7.      Document financial flow. A similar process should include financial flows. This is particularly true if there are any innovations in the way this flow is handled. Among the elements to include are the following:

·        Cash-to-cash cycle. This shows how long it takes from the first expenditures to the time money is collected from the end user.

·        Balance sheet figures including inventories, accounts payable, and account receivable.

·        Estimated activity costs across the chain. Use end user purchase price to “allocate” funds to supply chain activities. 

8.      Document new product flow. New products often upset existing supply chains. To the extent new products are contemplated, the processes should be examined. This should include:

  • Expectations for new product introductions and the supply chain implications.
  • An understanding of how the new product process incorporates supply chain participation, if it does.
  • Special supply chain requirements for supporting new product development. An example is finding reliable sources for components requiring special features.

Before the Redesign Begins…

Before launching the lean improvement effort, managers should consider the following from the information they have assembled:

·        Effectiveness of current supply chains in terms of serving customers.

·        Proposals that would more effectively serve a customer segment. How lean methodologies would support customer requirements. Good examples address cost, flexibility in the face of changes in requirements, lead-time, and quality.

·        Recommendations for dealing with barriers to implementation.

·        Needs for systems upgrades including automated systems and lean technology links.

·        Opportunities for improving profits and cash flow in the supply chain. This includes both the company and its supply chain partners. Map the right lean technique to the opportunity.

·        Questionable product lines and market segments in terms of profit and the supply chain capability to serve them. Don’t waste time applying lean initiatives to these.

McMenamins Edgefield

MMTI/NWLEAN workshop and seminars in Portland, OR, 3/5-9/2001

The information in this article is now obsolete. For up-to-date information on the workshop and seminar, click here.