Jeff Liker’s Lean Lessons (For Tesla Motors, Faraday Future, & Others…)
With Elon Musk announcing in the recent quarterly conference call that he’s hell-bent on seeing Tesla become the best manufacturer in the world, now seemed like a good time to interview some of the world’s leading manufacturing experts, listen to their wisdom, and think through implications for the feisty fossil-fighters in Fremont, California.
Tesla will definitely carve its own path; its engineering teams have no doubt developed their own systems and evolved their approaches. Unfortunately, those details are confidential, and for good reason: no company wants to tip off its competition. That means those of us on the outside can only speculate about the kinds of things Tesla might be up to.
Even so, we might as well try to speculate accurately, eh? If we ask manufacturing experts to share their key insights, we might get a glimpse of the kinds of challenges Tesla is working to overcome, to go from where it is now to where it soon wants to be.
We kicked this series off by interviewing Dr. Jeff Liker over email. A highly regarded authority on production methods, Dr. Liker’s biographical blurb reads as follows:
Dr. Jeffrey K. Liker is Professor of Industrial and Operations Engineering at the University of Michigan and President of Liker Lean Advisors. He is author of the international best-seller, The Toyota Way: 14 Management Principles from the World’s Greatest Manufacturer, 2004 (26 languages, over 900,000 copies sold), and has coauthored seven other books about Toyota, including Toyota Culture and The Toyota Product Development System. His most recent books are: The Toyota Way to Lean Leadership (2011) and Developing Lean Leaders at all Levels: A Practical Guide (2014). His articles and books have won twelve Shingo Prizes for Research Excellence. In 2012, he was inducted into the Association of Manufacturing Excellence Hall of Fame, and in 2016, inducted into the Shingo Academy.
Now, on to the content!
CleanTechnica (CT): how would you describe “lean production” to someone unfamiliar with the term?
Jeffrey Liker (JL): “Lean Production” is a term introduced in the book The Machine That Changed the World, to describe the superior systems in Japanese automakers compared to US and European automakers in the 1980s. These companies, especially Toyota, were superior in many ways by large amounts, including less inventory, higher productivity, higher quality, fewer engineering hours for developing new vehicles, and more. This was later modified in Lean Thinking to be mainly due to the superiority of Toyota Motor Company. Since then, Toyota has become a model of operational excellence for customers around the world. Toyota’s system streamlines processes and involves all employees in continuous improvement to better provide value to each customer.
[Matthew’s musings: readers might be surprised to know that Toyota’s fanboys in the “Lean” community are as passionately fanatical as Tesla’s fanboys in the electric vehicle sector. The parallel makes sense when one considers that just as Tesla revolutionized / reinvented what an electric vehicle could be, Toyota revolutionized / reinvented mass production (into “lean production”).
Given Toyota’s tepid-at-best enthusiasm for plug-in electric vehicles, readers might be tempted to write their wisdom off as passé. That would be a mistake, because the principles have proven successful in many manufacturing industries, and even the service sector: many hospitals have adopted “lean” techniques to reduce errors and improve patient outcomes.]
CT: what is the role of “kaizen,” or continuous improvement, in a lean production environment?
JL: Toyota defines their broader business model, the Toyota Way, as having two pillars — continuous improvement and respect for people. It applies to every part of the company from design to manufacturing to sales to purchasing and more. Continuous Improvement, or kaizen, is at the heart of Toyota’s excellence.
[Matthew’s musings: until Toyota came along, conventional mass manufacturing was very top-down. It was influenced by the concept of scientific management, or Taylorism, which essentially posited that if management studied and measured exactly what workers did, it could figure out how to squeeze more productivity out of a factory.
You may already see the Achilles’ heel in this approach: only a small minority (a few percent?) of employees actually study how to improve things, while everyone else coasts along, doing what they’ve always done.
The kaizen approach puts every employee on the front line of improving processes and eliminating waste (basically, any non-value-added activity). I hate to use the cliché of “empowerment,” but all employees are encouraged and expected to look for ways to reduce waste and improve quality. Of course, it’s a two-way street — the company is expected to invest in the employees to help them make the shift from unskilled labour which simply follows instructions to creative problem solvers who identify ways to improve them.]
CT: many products are manufactured in batches, but lean manufacturing systems encourage an approach known as “one-piece flow.” Could you explain why one-piece flow is preferred?
JL: One-piece flow is an ideal for all processes, manufacturing, and service. The ideal is to directly flow value to the customer without interruption by what lean calls waste. This is an ideal because it would require perfect quality, complete coordination among parts of the value stream, everyone keeping up with their part of the work, and perfection is never fully achievable. It is an ideal to strive toward.
[Matthew’s musings: it’s well known that Tesla batches its production, generally working on overseas builds in the first half of a quarter, before turning its attention to North American builds in the second half. Given where Tesla is at right now, this probably makes sense. Going forward, as it tries to tack on a zero to its annual production totals, it is likely to adopt more of a one-piece flow approach. This is difficult to do — one needs extraordinary coordination with suppliers — but is extremely worthwhile.
The biggest benefit from one-piece flow relates to cost savings from defects. If you (or your supplier) make a very complex component in batches of 100 and then discover a problem, you will have wasted 100x as much money building faulty components, than if you made them one at a time, in which case you could fix the problem and make 99 good ones. That adds up fast, especially at the $20,000 and $30,000 automotive price points!]
CT: many products are manufactured in batches, but lean manufacturing systems encourage an approach known as “one-piece flow.” Could you explain why one-piece flow is preferred?
JL: In order to work toward one-piece flow, we need a leveled system without big ups and downs or big batches of one product or another. For example, if Toyota is making the Camry and the Avalon on one line and the Avalon takes more work because of more high-end options, then a bunch of Avalons coming down the line would overwhelm the capacity of the people and processes. It would overburden team members, causing stress. And with a leveled flow the system can function with far less inventory.
[Matthew’s musings: levelizing production is kind of a cousin of one-piece flow. It’s the idea that you’ll do best when you have a steady pace, without resorting to heroic ramp-ups (which will stress your hundreds of suppliers — remember, only one of them has to have a problem, for you to have a problem!) to meet goals.
Tesla certainly seems to push hard at the end of each quarter, perhaps to beat its guidance and taunt its naysayers. The concept of heijunka, however, suggests that a slow-and-steady-wins-the-race approach will, well, win the race. At some point — certainly by the time Model 3 production begins in earnest — it will be more practical to produce at a relatively constant level every quarter, every month, every week, and indeed every day. Indeed, this kind of steady production would allow the Fremont factory to shift to daily deliveries of subcomponents, meaning it could carry less inventory, opening up more of the factory floor for actual production, helping Tesla scale its production without needing to expand the building’s footprint.]
CT: Could you describe the Shewhart / PDCA cycle? As a follow-up, do most organizations pay enough attention to the “P” in their problem-solving efforts?
JL: We can think of this as the scientific method. We identify the target we need to reach and develop a hypothesis for a countermeasure we believe might move us toward the target (Plan), we try it out (Do), we check what happened (Check), and we reflect on what we learned and its implications for further steps (Act). An important part of scientific thinking is healthy skepticism. We do not know that an idea will work until we test it.
[Matthew’s musings: when faced with problems you want to solve quickly, it’s human nature to want to skip “plan” and jump straight into “do.” You can sometimes get away with it when dealing with simple, self-contained products, but it’s unwise when dealing with complex, interdependent systems. Too many unintended consequences can arise.
I’ve been on a few high-urgency task forces which wasted weeks (hundreds upon hundreds of person-hours of work) because the team was so confident it knew what the problem was, it proceeded straight to “Do.” And boy, were we ever wrong. Repeatedly!
Learning from its mistakes, the company eventually formalized how its task forces worked. Among other things, it mandated that teams use a “5 Whys” approach so that instead of merely tackling symptoms, they attacked the ultimate root causes of problems. What’s a “5 Whys” approach, you may ask? We’ll let our expert talk us through it.]
CT: could you explain how the “5 Whys” technique helps people tunnel past the immediate cause of a problem, to the ultimate root cause?
JL: This is kind of a heuristic device to challenge us to think deeply about the problem. Why did that defect occur? Because of an assembly error. Why did the assembly error occur? Because someone failed to align the parts right. Why? Because they were able to misalign the parts. Why? Because there was no mistake-proofing device to prevent the mistake. Why? Because mistake proofing was not designed into the process.
This questioning, which can be 3, 4, 5, or more whys, drives us to get below the surface and think about the real cause.
[Matthew’s musings: Tesla may well have already deployed the 5 Whys technique to inoculate itself from any recurrence of the hubris that led it to be too aggressive in trying to pack technology into the Model X. We don’t have any insider knowledge, but could speculate the “lessons learned” meeting could have unfolded along these lines:
Why #1 — why was the Model X so late? Those darned-cool but darned-complex falcon-wing doors.
Why #2 — why did we include those doors? Because of hubris; we were overconfident that if we could design cool new features, we could deploy them.
Why #3 — why were we overconfident? Maybe, like many software people, we under-appreciated how hard it is to make hardware. (“It’s axiomatic in Silicon Valley’s tech companies that if the math and the coding can be done, the product can be made.”)
Why #4 — why did we under-appreciate manufacturing challenges? Maybe we didn’t have enough manufacturing engineers with automotive experience, and/or they didn’t have enough clout. (Matthew’s friends have told him that, at least in the past, Tesla preferred their engineers not to have automotive experience, as they didn’t want the company limited by conventional auto-sector thinking.)
Why #5 — why didn’t they have enough clout? Maybe we didn’t have enough grizzled auto manufacturing veterans at the senior decision-making levels, and/or we didn’t give them veto power.
Solution to the root cause of the problem? Going forward, maybe we could hire some grizzled auto manufacturing veterans to balance out our approach. Like, say, this guy!]
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Latest CleanTechnica.TV Video
CleanTechnica uses affiliate links. See our policy here.
