Lean-Lite versus Lean-Deep: Interview with Michel Baudin

I’m excited to present this interview with Michel Baudin who has had a very interesting and long history with the Toyota Production System.

In this interview, among other things, you’ll learn:

Why Michel believes that Lean is to the Toyota Production System as American Chinese Food is to Chinese Food in China.
Why he prefers the terms Lean-lite and Lean-deep to describe Real Lean versus Fake Lean.
Why he feels Bob Emiliani is missing some critical aspects in his description of “Real Lean”.
Why he likes Eric Ries’ Lean Startup principles, but fears that using the word “Lean” may backfire on him for reasons that have nothing to do with the ideas in his book.

Enjoy this very informative interview. And check out our other interviews also with other Lean Leaders.

Thanks Michel for spending time with us today. Can you please tell my audience about yourself and your work?

Thanks for the opportunity. I was born and raised in France, but, from an early age, used all the opportunities I could get to immerse myself in other cultures, first Germany, then the US, and finally Japan, where I went as an exchange student in 1977 after finishing my engineering studies. Three years later, a job that included organizing tours of Japanese factories for foreign industrialists turned me on to manufacturing.

Back in the US in the early 80s, I made my bones in factories as a process engineer in semiconductors, and moved on to the development and implementation of Manufacturing Execution Systems (MES) for this industry, an activity which took me again to Japan five years later. There, I met consultant Kei Abe, who took me on as a junior partner in his group and I spent the following eight years on the road with him in factories in a broad range of industries, in the US, Europe, and Latin-America. In 1996, I started my own group in the US, the Manufacturing Management & Technology Institute, later renamed the Takt Times Group, after a newsletter we had published early on.

Kei Abe could have his hands in foundry sand in the morning when coaching a team through an experiment, and discuss company strategy with the board after lunch. In manufacturing, he felt, management and technology issues should be addressed jointly. You cannot improve an organization by just introducing technical tools, but you cannot do it either by just reorganizing. You must have an approach that changes the way people work at all levels, engages them, and grows their skills. The key to get started is finding the right opportunity within the organization. Years later, reading Robert Schaffer’s The Breakthrough Strategy, I found it consistent with what we had been doing.

Working with Kei, I became our group’s main presenter, and developed training modules that we used at the start of projects, on topics like cell design, SMED, Kanban, Quality, etc. Later, I collected these materials into short courses that I started teaching through UC Berkeley extension in 1995, and later through other institutions, as well as in house for clients. These courses were then the basis for my books.

My first book, Manufacturing Systems Analysis, came out in 1990, and was based on my early MES work. It was followed by Lean Assembly (2002), Lean Logistics (2005), and Working with Machines (2007). For more details, visit my blog or my LinkedIn profile.

How did you begin your journey in the Toyota Production System? Was there a specific event that sparked your lifelong interest?

A hallway conversation in Tokyo in 1980. A colleague told me that, if I went to any neighborhood bookstore in Tokyo, I would find at least two books about the Toyota Production System. It intrigued me because I had never heard of a “Toyota Production System” and who would want to read about this while standing in a commuter train? I assumed such a topic would only be of interest to specialists, not the general public. So I checked it out near my apartment, and this is where I bought the copy of Taiichi Ohno’s first book that I still have on my shelf.

After reading it, I took every opportunity to visit factories that my job afforded me, and it included not only Toyota but car plants from other automakers, heavy industry, consumer electronics, and even semiconductors. It was the spark. The following year, I moved to the US and went to work as an engineer in a manufacturing plant.

You’ve been vocal in the past between the distinctions between Lean and the Toyota Production System. Can you explain?

There are plenty of reasons not use an explicit reference to Toyota when applying the Toyota Production System (TPS) in other organizations. From a public relations perspective other car makers cannot openly copy a competitor’s methods. Outside of the auto industry, people are skeptical about the relevance of these methods. And then there are nationalistic responses to the effect that “Japanese management won’t work here.”

But what is a good name? Consultants tried several. “JIT” was used in the early 80s, but it is does not encompass the whole of TPS. “World-Class Manufacturing” is too generic, and so is “Operational Excellence.” Finally, in 1989, John Krafcik came up with “Lean.” As it caught on, however, it was gradually drained of its TPS content and replaced with VSMs and “Kaizen events,” while implementers continued to believe that it was fundamentally TPS, with improvements.

I think I wrote somewhere that Lean is to TPS as American Chinese cuisine is to the original. The most popular “Chinese” dish in the US is General Tso’s chicken, which is unknown in China. It is reasonably harmless for cuisine, but the problem with Lean is these watered-down and distorted implementations failed to deliver the expected improvements.

We interviewed Dr. Emiliani recently and he makes a distinction between Fake Lean and Real Lean. Would you agree with him or are both of his distinctions fall in the “Fake” category in your mind?

I have talked more about Lean-deep versus Lean-lite. Lean-lite comes in many flavors, the most common one being the combination of VSMs and Kaizen events, but there are others, like doing 5S and nothing else. What all these flavors have in common is that they are simplistic and ineffective.

Lean-deep is not necessarily comfortable. It is an “Eat your vegetables!” message. It starts from an in-depth study of TPS. You have to drill down from specific tools to underlying principles that you can then deploy as needed, possibly with new tools, in a different context. The list of principles I have found the most useful for this purpose is as follows:

Focus on people as the main driver of productivity.

Search for profits on the shop floor.

Make it easy to do what you do the most often.

Flow, flow, and flow.

Improve, don’t optimize.

Bob Emiliani’s Real Lean boils down to “continuous improvement” and “respect for people.” But there is more to TPS than continuous improvement. It contains a great deal on, for example, designing a production line, which is not continuous improvement. And “respect for people” is a mistranslation by Toyota people, of a principle that literally means “respect for humanity,” and it is taking advantage of the specifically human abilities of employees. Showing respect for people is being polite; showing respect for their humanity means putting their cognitive, sensory, analytical, and creative abilities to work. It is quite different.

If Lean boils down to “continuous improvement” and “respect for people,” then you can’t tell the difference with the now defunct “HP way,” with which Bill Hewlett and Dave Packard grew their company for four decades. When you read what Dave Packard wrote, and talk with HP veterans who experienced it, you can make the case that the company worked continuously on improving its products and its operations, and showed great respect for its employees, for example by subsidizing their PhD studies. In its day, it made HP a spectacularly successful company, but which in no way resembled Toyota.

I’m glad you bring that up. I’ve always used the phrase Respect for the Human. That’s what I learned. It has a much deeper meaning than just being polite as you say. It has more to do with the absolute reverence for other sentient beings. It actually has a very Eastern flavor to it that I think escapes most Americans.

Let me switch to another topic.

John Shook has said that “Lean is a Big Tent”, meaning that Lean is broad and is inclusive enough to include many different instantiations. To this end, we see Lean Branding, Lean Startup, Lean UX, Lean for IT, and many others. What do you think of all of this?

Remembering that TPS is the best known way to build cars, you have to ask yourself, on a case-by-case basis, how much of it is relevant and applicable to a new domain. I find it arrogant to assume that any business activity can always be improved by applying tools or even principles abstracted from TPS. In each case, it may be true, but you have to be open to the idea that it may be irrelevant.

I remember being impressed with steel makers in North America. These people and their organization had survived an intense shakeout that had bankrupted most of their competitors. They were the ones still standing. Listening to them left me in awe of their talent and experience, and I was not absolutely sure I could help them. It is an industry that is, perhaps, too mature to respond to Lean/TPS. You have similar issues at the opposite end of the maturity spectrum, in high technology, where you are chronically dealing with processes that are not ready for concepts like one-piece flow to work.

About Lean Startup, I have read Eric Ries’s book with interest. Having been involved with startups, I find that many of his ideas make sense, but I don’t see any connection with Lean. I think he attached the “Lean” label for marketing purposes, and that, while it has helped him so far, it may backfire in the future when the label loses its popularity, for reasons that have nothing to do with his ideas or their effectiveness.

IT is also an area that I don’t see as needing to learn from Toyota, as opposed to learning from companies that have been good at IT. People have been thinking about how to organize software development and implementation since Alan Turing. In his writings from the 1940s, you see discussions of the professional behavior of programmers that are astonishingly prescient, considering the profession did not exist when he wrote. In the 1970s, you had Fred Brooks’s “Mythical Man-Month” collection of essays about the management of software projects. Later on, Tom DeMarco’s remarkable insights. And today, you have scrum, with its pigs, chickens, sprints, and assorted mixed metaphors… All along, you have had people from the industry reflecting on its idiosyncrasies, and struggling to find ways to make it work better. It is not clear at all that they need to borrow from the car industry.

Healthcare, I think, is a promising area. 100 years ago, Frank Gilbreth successfully applied his analytical methods from production to the design of operating rooms, resulting in the organization in place to this day, with nurses making sure the surgeon always had the needed tools available. It means that there is a history of the health care world accepting ideas from manufacturing. Of course, it has to be packaged carefully, as the notion of a hospital run like a car factory appeals neither to health professionals nor to patients.

I am not familiar with Lean Branding or Lean UX, and have no opinion about them. In general, consultants and authors are free to attach the “Lean” label to whatever they want. It can be an attempt to create a perception that their offerings have something to do with TPS when they really don’t. After 25 years of being used in this fashion, the “Lean” label is long in the tooth, its meaning is diluted to almost nothing, and I am not sure using it is even a wise marketing move at this point.

Please tell us about Takt Times Group and some of the results that you’ve helped your clients achieve.

The Takt Times Group is a small network of independent consultants rather than a firm. I also learned this mode of operation from Kei Abe, who had worked in a large Japanese consulting firm and felt that having overhead and “mouths to feed” created a conflict for consultants. Having associates that you need to keep billable inevitably influences the advice you give, and not necessarily in the best interest of your clients. When he explained it to me, it resonated, because I had heard it before from an international marketing consultant in Tokyo, who had grown a firm to 40 associates and left it to go out on his own for that very reason. My partners and I have known each other for many years, work under the same brand, and share the proceeds of joint projects. For international work, we also partner with local firms, like Asenta in Spain, OrgProm in Russia, or China-Inno in China.

I think we’re in agreement regarding the utility, history, and role of value stream mapping. I know this is a hot topic for you. Suppose you were hired by an organization that had a value stream manager for every available value stream in their organization. What would you do if hired by this organization?

When introduced to a new tool, John Seddon recommends asking (1) who invented it, (2) to solve what problem, and (3) whether you have that problem, and I tend to agree. To find the answers for VSM, first you need to go back to the original name of the tool, which translates to “Materials and Information Flow Analysis” (MIFA). The closest I could come to an inventor is Toyota’s supplier support organization in Japan. And it was developed for the purpose of solving delivery problems with suppliers. So, if you have a supplier who delivers the wrong items, the wrong quantities or at the wrong times, you map the flow of materials and production control information between you and this supplier to pinpoint the source of the problem.

Can it be useful in other contexts? Absolutely. Should it be a mandatory first step in every Lean implementation? Absolutely not. It is not a big deal in Japan, and the only publications about it in Japanese are actually translations from American documents. For more details see “Where do VSMs come from?“.

Related to my earlier question on the Big Tent, what aspects of TPS which make sense for manufacturing don’t make any sense for non-manufacturing environments? Do you have a specific example you could share?

Facilitating simulation games in a manufacturing organization is an opportunity to observe behavior. In our games, teams have 50 minutes between production rounds to reflect on what worked and didn’t and the last round, design changes to improve for the next one, and implement these changes. Teams of managers and engineers rarely have a problem following this simple, formal process, especially if they are colleagues who work together in real life. Teams of production operators, on the other hand, struggle with it. You often see them rush to make changes on the production line without making any drawings or calculations.

This tells me that, while these operators may be great organizers in their private lives, at work they are not used to solving problems in teams. The will to do it is there, but you need a formal structure, a process to help them work through it. That is where tools like PDCA, Why-Why analysis, and a variety of forms are useful. Try to train a group of PhDs in an R&D lab on “problem-solving” in this manner, and they will throw you out.

Given all your experience and knowledge of TPS, what else are you working on?

I have had an interest in data science for a long time. In manufacturing, I have found clients and colleagues interested in the information I could wrangle from data, but their eyes glazed over whenever I tried to explain my methods.

At first, I restricted myself to using software everybody had — like Excel — so that others could easily reproduce my results. But it is too restrictive.

Manufacturing data is usually quite dirty, with, for example, the same product given different names over time, or Engineering, Manufacturing, Marketing, and Accounting each having different ways of grouping products into families, not to mention products with no sales but positive revenue,… You really need other tools than Excel to sort this out, especially considering that its overuse is the cause of many of these errors and inconsistencies. I have found the query capabilities of a database management tool like Access to be effective for data cleaning. Once you have clean data, I also found that R, a tool that is free to install but takes time to learn, opens the door to a whole world of techniques that are not available through Excel.

About Michel Baudin

In 1980, tours of Japanese factories made Michel Baudin switch his career focus from mathematics to manufacturing, and move to Silicon Valley in 1981, to work in the semiconductor industry, first as a process engineer, and then in the development and implementation of Manufacturing Execution Systems (MES), an experience that provided the material for his first book, Manufacturing Systems Analysis.

Starting in 1987, he became a consultant as a partner to Kei Abe in Management & Technology Japan, learning the art of Lean manufacturing implementation in a variety of countries and industries. He started his own consulting group in Palo Alto, CA in 1996, known first as the Manufacturing Management & Technology Institute (MMTI) and now as the Takt Times Group. He has distilled his consulting experience in his “nuts and bolts” series of books, which has so far three titles, Lean Assembly, Lean Logistics, and Working with Machines.

Michel Baudin was born in France and experienced different cultures from an early age, first in Germany and the US, and later in Japan. He received his master’s degree in Engineering in 1977 at the Ecole des Mines de Paris, now Mines ParisTech, and then went to Japan to pursue research in seismology at the University of Tokyo, where he learned the language and conducted research on the statistical modeling of earthquake occurrences that led to four refereed publications applied math.

Other Interviews You Might Enjoy: