Mark Graban has been editing a book titled Practicing Lean: Learning How to Learn to Get Better…Better. I recently wrote a chapter which I have included below for everyone’s reading, but I recommended you pick up the book. The proceeds benefit the Louise H. Batz Patient Safety Foundation. Here is my chapter:
What’s wrong with your pull system?
My first formal experience with lean was in the early 1990s. I was managing the materials process at Harley-Davidson, where we had one of the first comprehensive, end-to-end pull systems. It was called MAN, for Material-As-Needed. The manufacturing plant was fairly fully integrated to make motorcycles. We had rolls of steel to cut, stamp, and weld gas tanks. We had our own (and very old) chrome-plating operation and zinc-plating operation. We had what we thought was a state-of-the-art paint system until someone left a bay door open overnight and the place filled with bugs, causing system problems for several weeks. Everything was connected with the MAN system. And I ran around with my radio trying to hold most of it together.
The MAN system was part of the what could be called the Harley-Davidson Operating System. There were only three components, represented in a triangle: Material-As-Needed, Statistical Process Control, and Employee Involvement. Statistical Process Control wasn’t groundbreaking, but focused on building some stability in the process which was much needed. The success of Employee Involvement wasn’t as obvious to me. There were spots where it was going well, and other spots where it hardly existed.
Our operation wasn’t doing terribly well at the time. Our “float,” meaning the bikes that had been built, but needed some repair before being shipped, was the equivalent of a whole day’s production by the end of the week, despite working to repair them as quickly as we could. Unlike my time later on in an automotive plant, we couldn’t store new bikes outside. So we had a find a place. We clearly didn’t have room for that many bikes. So, bikes were put everywhere, but the only flexible space was normally used to store material. By the end of the week, my biggest problem was where to put material. A typical call on the radio was “the shocks aisle is filled with bikes, where do you want me to put them?” I’d walk over, and we’d end up putting them where we could, including the aisles, which would generally cause my next radio call to get them moved.
The real circle-of-life is that the primary culprit of bike repairs was missing components. As the bike went down the assembly line, a component would be unavailable, and we’d keep building the bike as well as we could. As long as there were wheels, and there usually were, we could at least put it on the ground and push it out of the way. The materials department wasn’t blamed for all of them, as many times the parts had not yet been produced by another department in the plant, but I was always the first call.
While my boss will remain nameless, he was not the most engaged manager. He had come out of accounting with no operational experience, and also managed shipping and receiving, in addition to my area. We had a long and narrow plant and his office was at one end, with the administrative building at the other end. Instead of walking down the main aisle, where someone might stop him to ask a question or bring up an issue, he would walk outside down the length of the plant, and not just because it was a nice day outside.
He would complain often that the material handlers that had fork trucks were driving recklessly. Honestly, they despised him so much that I believe they often made mock runs at running him down when they could. They would never admit it, but it was fairly obvious. This resulted in a story that is a bit off topic, but worth sharing under the heading “bad boss.” Convinced they were driving too fast, he wanted me to prove it for them. I resisted, mostly because I thought it was stupid. Never mind that there was no reasonable way to prove it. Finally, he simply ordered me to do the following. I put down two pieces of tape on the floor. I then stood between them with a stopwatch. I was a speed trap, and a rather obvious one at that. They knew it wasn’t my idea, but certainly had their fun with me. They would speed up to the line, and then crawl through the zone at 2 mph. This certainly put my credibility back a bit, but completely destroyed the credibility of my boss.
Then came his next assignment. Knowing that our MAN system wasn’t working terribly well, he wanted to scrap the whole system. He believed that manufacturing was a precise and predictable machine. He wanted a schedule that stated we would pick up this part at this time. It would be a detailed, precise schedule to manage all of material handling. Again, I resisted, despite knowing that our current system wasn’t working. I already knew that resistance was not going to work, so I finally took on the task.
The pivot and the lesson
I set off under the guise of building his system, but with the true intention of proving him wrong. The idea that pull systems work was far from a common belief at that point. But, I was sure that a precision-scheduled system would fail. If our system relied on picking up the exact number of parts at the precise time on a certain day, then both the demand (schedule) and supply (quality and so on) must have zero variation. Now, I’m all in favor of putting challenges into a manufacturing system in order to drive improvement, but I have yet to meet the manufacturing system with zero variation. So I set out to meet two objective. First, gather the data needed by my boss for his magic system. Second, learn why our pull system wasn’t working.
I put down my radio and committed to observe every employee for an entire shift, spanning over a few weeks. It turns out I was practicing a fundamental lean method, of going to the gemba and directly observing the work, although I didn’t quite understand this at that time.
I did nothing but observe the work, the people, and the system. If I would have done it for just a day, I never would have learned what I ended up learning. In fact, it took well over a week until the patterns became clear. I think they only became clear because I kept an inquisitive mind the entire time (to which I credit whatever success I have experienced).
The pattern was this: everyone violated the system frequently, but always with good intentions. On the line, people would see the material handler coming and would grab a couple extra pull cards from their bins to give to them. In the supplying areas, people would make just a few extra parts while they were already set up. The material handlers, while picking up parts, would grab just a couple extra bins in addition to the number of cards, since the line would need them eventually anyway. And, the parts and accessories group, not large enough at the time for their own inventory, would come to the line to grab some parts that customers needed and, while they were there, would grab a couple months’ worth instead of just what they needed.
Everyone thought they were being efficient and helpful. But, the net result of all of their behaviors was that the system didn’t do anything that it was designed to do. For example, early card pulls triggered early production of the wrong parts, consuming valuable capacity. Every small deviation had a cascade effect on the system, which was now unreliable, which in turn drove even greater behavior distortion and work arounds.
I’m not sure if I ever convinced my boss that he was wrong, or if he just didn’t care anymore, but we never worked on his scheduling system. We instead focused on the behaviors. With a combination of educating people on the purpose of system rules and accountability for following them, we had a modest but measurable improvement in behaviors. But, that was enough to see real changes in the stability and effectiveness of the system. And, because some of the behaviors were workarounds, even slight improvement in the results caused further improvement in behaviors surrounding the system. For example, only pulling Kanban cards when you’ve actually consumed the material instead of when you see the material handler coming. Trust that the material handler will come back later, and know that violating that process will cause work to start pre-maturely. When just a few more people follow rules such as this just a bit more often, the stability of the system improves and so do the results.
Much more work was needed before we saw a substantial improvement in the motorcycle repair rate, but this, combined with other improvements, eventually improved our first-time right rate and, ultimately, the factory’s output.
I was extremely fortunate to have this experience, because I learned very early in my lean learning journey that the tools do not make the system. Its not the process. Its not even how all the tools fit together. It is the behaviors that matter most. If you don’t get the behaviors right, the tools, processes, systems, and the rest will fail. Lean is about how we think, with an emphasis on the “we” because it is the shared behaviors, consistently applied across the organization, that matters most.
This shaped most of my lean journey, although by no means was I able to articulate how at the time. It led me into the realm of system thinking and organization learning and I was among the very first to see the connection between organizational learning and lean. Organizational learning focused on important methods such as reflection and system dynamics. Reflection later became recognized as hansei, a useful method in the lean toolbox. The organizational learning community, in particular its standard bearer Peter Senge, recognized Toyota as the ultimate learning organization.
This experience affected my journey at Chrysler and, later, at DTE Energy, making increasingly more effective efforts at making people, behaviors, and learning central to a lean journey. Ultimately, that path led me to co-found a consulting and education firm with Andy Carlino, and write, with him, The Hitchhiker’s Guide to Lean. Our work and our book helped shift the lean community as a whole to be less about the deployment of tools and more about building a culture. At the time, we were considered contrarians. Lean was all about tools, especially highly promoted tools such as value stream mapping. One person told me that an organization wasn’t lean unless they had a value stream map on their wall. A university professor who was deeply involved in lean suggested that if you design the system, people will operate effectively. These sound wrong today, but representing the prevailing opinions 15 years ago.
I would claim that we were the first to make lean about behaviors, but I believe that Kiyoshi Suzaki who wrote The New Manufacturing Challenge and The New Shop Floor Management was certainly articulating a people and learning-centric model of lean, although too few learned from his teaching.
I believe I also learned a bit about Suzaki’s quote: “lean is common sense…after the fact.” People with good intentions wouldn’t be violating the system so much if it was intuitive. But many aspects of lean are counter-intuitive, which makes education and coaching so vital.
I learned a lot from my initial experience. Some of it impacted me immediately. Other aspects would take another decade to fully take root. For example, I didn’t have clear language to articulate any of this, but then I didn’t have an audience to share that language with. But, I was very fortunate to have the experience I had, and was mindful enough to learn everything I could from it. I’ve always put learning first, ahead of career, money, location, or working hours. That served me then, and has served me since.
For the most part, more practicing. I want to bring lean into new frontiers, such as the board of directors, and continue to find ways to make it less mystical and more accessible to everyone. I also intend to be a better practitioner in working on my own systems. I am refining my use of Kanban management systems with myself and various partners, I am building more deliberate daily experimentation, and I am continuing to refine my personal standard work. For example, as I plan my week, I determine my primary customer for 3 days of the week, meaning who am I trying to deliver value for. I then determine how I will deliver that value, and how I will evaluate that value, creating my hypothesis. And then I test.
And overall, I simply hope to keep learning through the deliberate practice of lean.