CHAPTER 1

LEARNING FROM MR. TAIICHI OHNO

Lesson 1: No One Really Understood What I Was Saying, So I Had to Go to the Gemba and Give Detailed Instructions

This is an excerpt taken from Taiichi Ohno’s writings at the time of the commemoration of Toyota’s first automotive plant (the Honsha plant) in 1988:

In 1963, we deployed kanbans throughout the factory. There was no grand design for this, so even we didn’t know what would happen. The impact of the words “increase productivity,” “limit overproduction,” and “use fewer people” could not really be calculated. No one really understood what I was saying, so I had no choice but to go to the gemba and give detailed instructions.

The first time Ohno’s name appeared on the manager roster was in 1944, when he was a manager of the Toyota automotive arm. At that time, he was the manager of assembly; this department spent at least the first week of each month “gathering parts,” which meant that the assembly line was actually working only two weeks out of the month. In 1945, he became the leader of the machine shop, and that was the start of what the world now knows as the Toyota Production System (TPS).

Ohno spent the next 16 years in the Honsha plant establishing the many foundational elements of the TPS, such as multimachinery handling, multiprocess handling, visual management, one-piece flow, and kanban. This was not done overnight, and, as Ohno has reflected, it was done within his sphere of influence at the time, and as his influence expanded, he tried out new methods.

In 1963, Ohno finally became the factory manager of the Honsha plant, and he decided that areas such as the foundry and forging, which at the time were not connected to the machining plant and the assembly plant, would be joined to those plants by kanbans. In short, he decided that the whole plant needed to be joined together. When we look at Ohno’s quote, we can see that he was determined to see this through and make it a success.

People in the forging and foundry areas were a lot tougher than the people in the machine shop had been, and so Ohno himself took the lead in cutting waste, increasing value-added, and making good-quality parts.

When we see how Ohno implemented new things, we can note the following key success factors:

1. Hands-on role modeling by top management

2. No compromises, no fear of failure, and no giving up halfway with results that were “half good”

3. Recognizing that it would take time for things to become sustainable and for the people working there to be able to improve and do kaizen, so a long-term follow-up had to be part of the process

Because of this, top leaders must avoid short-term changes in policy. The reason that we have the Toyota Production System is that Ohno stayed in production for a long time and focused on making it work.

Getting Back to the Point

In these sections, which I have called “Getting Back to the Point,” I will be discussing experiences that I had based on Ohno’s teachings and the things he said.

I was deployed in Toyota’s machine department in the headquarters plant in 1968. I spent an hour listening to the department head give a speech about “the mindset and heart of the new employee,” and then I was told, “Your job will be at the Kamigo plant, so go there immediately.” I basically spent one hour at the headquarters plant and then was at the Kamigo plant the rest of the time.

When I arrived at the Kamigo plant, there were a lot of engineers from the headquarters plant who were there to get a new production line up and running. The factory where the first Mark II model was made was in the middle of a trial-and-error process of trying to get parts to flow on automatic lines, modifying and optimizing equipment, and deciding how to detect and show abnormalities—all under the direct supervision of Mr. Taiichi Ohno.

“Hey, newbie, get up on the gangway where you can see the andon”—I remember standing up there. The engineers were working very hard, and I could see the stress and pressure on their faces. This was because Ohno would soon be arriving.

When Ohno came, all of us went to the place where we could see the andon the best, and we went to each workstation where there had been a stoppage. An explanation of why the stoppage had occurred was given, and the maintenance staff was then called to make the necessary changes. It was right at the time of the Mark II debut, and we were having a hard time making the engine. Every day Ohno would come to the gemba and give commands. When he found areas where the handoffs and connections from one line to another were not good, we often had to spend weekends on the construction and modification tasks, sometimes even going without sleep. He would come on Monday to see what we had done. If we heard the words “I think it was better before,” we would have to spend another weekend putting it back the way it had been.

For Ohno, the way of working that was being developed was going to be the foundation for all other lines, so he was very engaged in the process. He would always come to the gemba, and if he didn’t like what he saw, we had to keep changing it until it was satisfactory. Even I, the new guy, could see that Ohno was also seriously thinking about the best way to do things. I thought to myself, “I now know the meaning of trial and error.”

I once asked my boss, “Why don’t we leave it as is, and then change it when the next project comes around?” He told me, “If we keep something that is not good, then someone will copy it. That’s why we have to fix it right away.” He then continued, “We mustn’t notice a problem, but pretend that we don’t see it and walk by. If you think of the person who has to work there and has to deal with that situation, you shouldn’t be able to let it go.” When I heard these words, I thought, “I have come to work with a really wonderful team.”

The seriousness of top management has been transferred to the people below, and I think that’s why I got this kind of answer from my boss. I wasn’t part of Toyota in 1963, but I am sure that the casting and forging areas also went through a period when Ohno visited them daily and worked with the local management to build up a workplace that could perform.

Lesson 2: Kaizen Equals Getting Closer to the Final Process

This phrase was hardly used by people at Toyota, which I think is why it stuck in my mind. It seems like a simple phrase, but many people have told me that it is confusing, so I will explain the background to the situation.

When I was around 30 years old, I was working as an engineer in the headquarters machining plant. I was on the number 3 truck undercarriage line. This line was synchronized with the main assembly line and supplied parts to it. The final process was to do a quality check, and sometimes that took a long time, delaying the delivery of parts to the main line and causing a stoppage. As a result, this problem was interfering with the stability of operations.

In order to keep the main line from being adversely affected by this, we moved the parts line to a shop next door and delivered finished parts to the main line. (Basically, we disconnected the subline from the main line and delivered finished parts to the main line.) As you might expect, we had solved the problem, as the line stoppages to the main line had ceased. When I reported this to Mr. Ohno on the gemba, he said, “Yes, we have fewer main-line stoppages, but the time required for adjusting on the subline has not changed. Kaizen equals getting closer to the final process. Now, however, it is further away.”

Our job as engineers was to find the reason for the variability in the process and change the conditions so that good product could be made in an environment that didn’t require lots of effort. Because production engineering couldn’t do this, we had just run away from the problem.

“If you lack skill, you can’t get closer to the final process.” How important it is to get the right conditions in place to make good parts. I felt that I had learned a fundamental truth about production at the time.

At the same time I thought how impressive it was that Mr. Ohno could see the real issue. I was also embarrassed. We had had no intention of running away from the problem, but the result of our actions had been to choose the easy way out, and to top it off, we thought that we had done a good job! “What are you guys doing?! Get it back to how it was before!” would have been an appropriate remark for Mr. Ohno to make, but his actual comment was more of a philosophical nature that really encouraged us to find the root cause of the variance and get the subline back to where it belonged—closer to the final process.

I think Ohno knew that our young group had done this without ill intent and felt that it was better to do something and fail than to do nothing at all. If you’re up at bat and you are going to strike out, it’s better to swing and miss than not to swing at all. Just by swinging the bat at the ball, you are going to gain some experience. We felt that we had learned a lot about kaizen when we struck out swinging.

When you think about it, the less inventory you have between processes, the closer you are getting to the end process. The same is true when you join processes. It was a simple sentence that had profound meaning for me. Ever since, I have asked myself and my staff to go check to see whether a change has gotten us “closer to the last process.”

Getting Back to the Point

There are a lot of so-called kaizen or improvement events out there that actually go further away from the last process. For example,

• Outsourcing the rework process to a vendor

• Outsourcing a deburring process

• Outsourcing the separation of good and bad parts to a third party

• Outsourcing the removing of the runner and gate (which are not required in the final part, but are necessary for the making of the molded plastic) from a molded plastic part

These are the worst offenders, but they actually happen quite frequently. Well, why is it wrong to give a process to a third party? Some people may say, “That party can do it at a lower cost than we can, so outsourcing it is a good thing.” The issue is that by keeping that process in-house, you keep the problem visible. If you see many people doing rework, it should motivate you to do something about it.

Things like burr-free casting and gateless molding come from the need to improve. If you cannot see the process, no one will think about it. The danger is that if something is difficult, it is outsourced; this then makes the company’s engineering weaker and in the end reduces its competitiveness.

I once had a big surprise when I studied the supply chain for a certain part that we used. This particular part involved four processes. Two of these processes were done locally, another was done on the island of Kyushu (1,000 km /600 miles away), and the fourth was done on a different island. My goodness! How much work in process did we have? I was convinced that it was at least six months’ worth. I am certain that this situation had a long history, but I think it was the culmination of Purchasing asking over the years, “I wonder who can make it more cheaply?”

By having a shorter supply chain and bringing it closer to the last process, a company can be in better control in case problems happen.

The effect of Ohno’s “Get it closer to the last process,” or, in other words, “getting things closer to the assembly line is a good thing,” can be seen in Toyota’s history, where all processes that came under Ohno’s control as his responsibilities grew became closer to the last process.

The most important thing Ohno had in mind to reduce the production lead time was the concept of flow. When he was still the machine shop manager, “We initially focused on reducing the changeover times and reducing the batch sizes” (from Ohno’s memoirs). He then continued, “We had people handle two of the same machines to increase productivity, but the batch sizes didn’t change. We then changed the layout to put the machines in the process sequence and our batch sizes were significantly reduced.”

This did increase productivity, but in order to prevent overproduction, a downstream pull/replenishment system was incorporated. This was the start of the kanban method, which actualized the just-in-time philosophy. Let’s review Ohno’s role and the history of kaizen that accompanied it.

Section Manager for the Number 3 Machine Shop

• Increased the span of the layout according to the process sequence and laid out the machines based on what parts they made.

• Decided on a five-piece batch size up to the assembly line and based the pull system on a five-piece batch size.

• Developed the “water strider picking method,” so named because it looked like an insect called a water strider that was moving left and right on the water surface collecting parts. With this method, even though the different part lines are located in various locations, the water strider picks up the required set.

• The final process of each part line kept a minimum of five finished parts ready for replenishment, which reduced the finished parts inventory and which moved that process closer to the final process.

Assembly Factory Manager

• Decided on making five of the same cars as one unit and linked the assembly shop and machine shop with carts pulling a five-car set of parts at a time.

• The carts that carried the five sets of parts were treated the same as a kanban and were called “kanban carts.”

Headquarters Factory Manager

• Linked the machine shop with the forging and casting plants using kanban.

Thereafter

Put the skill into the machines and, by eliminating the need for skilled artisans, went from a part-based layout to a setup in which the part making and part assembly for different cars were in the same area. This pretty much completed the evolution of getting the process closer to the final assembly.

Collecting the parts that had the same takta time (pace) in the same area allowed the teams to eliminate a lot of the waiting time associated with making parts. This also allowed for an easier way to structure the combination of work and reduced the losses associated with combining different jobs.

Other

Innovated in many other ways, such as the tsurube or well bucket method, single-minute exchange of dies, kitting, and the trolley method, to get everything closer to the last process.

I will sound like a broken record soon, but Ohno made changes only in the areas that he directly controlled. He never touched areas that were not under his control. Even if there is a highly skilled technician with a great method, going beyond the bounds of one’s authority will not work. That is why the role and responsibility of top management is so important.

Lesson 3: You Need by the Line Only the Parts for the Car You Are Assembling Now

We were told that all you really need are the parts for the car that is being worked on now, and that we should deliver them one car at a time when they are needed. All the other parts should be put away in one location. The obvious meaning is to keep only what is needed right now, but the other reason for this is to free the workers from the difficult task of choosing the right part and instead have them focus on value-adding work. It allows for simplifying the job as well as being able to assure quality more easily. The reason for keeping all the parts in one location is so that one can see the entire inventory at a glance. Back then, we had to adjust the number of kanban cards every month. By keeping all the parts in one location and not all over the place, it was easy to see at a glance what inventory there was, which allowed us to minimize it more easily.

At the main assembly line in the headquarters plant, the line Ohno built didn’t have the big parts by the side of the assembly line; instead, they were moved to the main line one by one on a conveyor. This was called the trolley method.

In the same way, the machine shop kitted one set of parts for the engine, the transmission, and the chassis in a box that we called a pallet. It looked a lot like the job of delivering meals in a traditional Japanese inn, so we called it the haizen method (hai means “to deliver”; zen in this context means “food”).

The automotive assembly lines used to use the trolley method, but around the 1980s, because of the combining of picking and assembly work, the trolley disappeared. After that, when suppliers delivered parts, they delivered them to the area of the assembly line in which they would be used. The assembly workers picking the instructed parts by the line became the norm.

In 2000, Toyota Taiwan was trying to both increase first-pass yield (quality) and speed up the learning curve of its employees, and it developed a method for delivering one set of parts for the engine assembly to the Corolla assembly line. This was very successful, and Toyota headquarters took the idea, developed it more, and called it the SPS (set parts supply) system. We can probably say that Ohno’s haizen system and trolley system came to completion with this.

Ohno’s Teaching

If you make the work as simple as possible, you will have fewer mistakes and you won’t have to say “obey obey” because people will obey the rules. Also, if it is simple, a method for having less inventory will be born.

Getting Back to the Point

The SPS system was developed to make inventory control easier, speed up learning the job, increase quality, and simplify the job—all of which were realized, but there were a lot of other nice surprises.

• Transportation time was reduced. The biggest initial problem was the increase in transportation time needed to deliver one kit at a time. To reduce this, we standardized the box size, which allowed us to automate the delivery system. This drastically reduced the work hours spent on delivering parts compared to the old method.

• The length of the assembly line was shortened. By taking away the parts from the assembly line, our line became 30 percent shorter. Up to this point, we had had a mixed model line, which was used for a long time. Even if we were making a new car, the old assembly line made work hard. We started using the method of making a new low-cost assembly line that could take advantage of all the new technologies every time we had to make a new type of automobile.

• Availability of supervisors and engineers for problem solving was increased. A lot of our supervisory staff’s time had been taken up by the need to sort out parts and deal with assembly mistakes. With these tasks eliminated, supervisors were able to focus on the value-adding assembly tasks, as well as on attaching processes that drastically reduced our internal defects.

• Staff could focus on developing a new line concept. With management and supervisory staff increasing their abilities, they were able to focus their talents on making a new line, which was great for management motivation. SPS’s goal is to simplify the assembly and kanban management. Just being able to deliver one set of parts to the worker does not mean that we have arrived, however. Most assembly lines are multimodel lines, which means that different types of cars are built on the same line one after another. That means that in the current situation, the employee has to perform a mental changeover, and will have to keep concentrating so that he does not make a mistake. Also, because different cars are being made, the worker will not necessarily attach a part at the same place. If we are striving for “simple,” we need to challenge ourselves to do the following:

Make a dedicated assembly line for each car.

Dramatically reduce the assembly line cost structure.

For the new cars, use the latest technology on an assembly line that you build yourself.

The management can see the next stage, which allows it to chase its dreams more.

Lesson 4: Building in Batches Stunts the Growth of Your Operations (Don’t Combine Kanbans and Build a Group of Them)

The final year in which Ohno drilled the machine shop was 1977. I remember that this was the longest and also the toughest period when our improvement efforts were under Ohno’s direct supervision. He would come to the gemba day in and day out, and he never gave us a moment’s rest. In addition, there was a lot riding on this effort, so we were told things we had never heard before. The focus was on the shop cutting the gears. Ohno saw the production kanbans on the line and started in on us:

Ohno:        How come you have three of the same kanbans in a row?

Us:              We are producing in lots. Once we get three kanbans, we then make the parts.

Ohno:        Why don’t you just make them in sequence when you get a kanban?

Us:              It takes an hour to change over the gear-cutting machine.

Ohno:        How long have you been doing this?

Us:              For a long time, sir.

Ohno:        A long time. I see. OK, from now on, you need to make the gears in the same sequence that you receive the kanbans in.

What he meant was that we had to do a one-hour changeover on the machine every time we had to make a new type of gear. Just because Ohno said, “Do this” didn’t mean that our changeover time of one hour had been reduced. Furthermore, he would come over every day, so there was no way of fooling him. We were spending more time changing over the machine than we spent making parts, so we started running into part shortages on nearly all the parts. When the downstream pull transportation person showed up, he was about to put the kanbans in the “shortage” box. Ohno stopped him and had him call over the supervisor of the machine shop. “Don’t put the kanbans in a box. You need to let the supervisor know. Give the kanbans to him by hand. Let him know that there is a part shortage, and you stay here until the parts are made. Going home empty-handed is meaningless, right?”

I wrote this in very gentle language, but it was a lot more combative, and one day the “shortage box” disappeared from the factory. From that day on, none of us knew what Ohno was trying to do, and Ohno wasn’t clear with us. He just came, and, “You’re making parts in the same sequence as when you get the kanbans, right?” was all we heard. Someone finally said that maybe Ohno was saying that we needed to reduce the changeover time on the gear-cutting machine. All of us were adamant: “No way; that can’t be done.” Still, if we didn’t attempt it, we could not do what he was asking of us, so we got together the design engineers, production engineers, and all the other departments, explained our predicament to them, and discussed well into the night how we could have an easier changeover. After we tried out our ideas, in the end we had got the changeover down to less than 10 minutes. Actually, in the late 1970s, the stamping press machines were already doing changeovers in less than 10 minutes, and this episode was six years after that.

The gear-cutting machine was one of those processes that just had a long changeover time. Because of that, it was making parts in batches, and that method had been working out OK. This kept us from seeing the need for reducing the changeover time, which is something that I really regret. It was true that reducing the changeover time was going to be hard, but the mistake was that we had accepted the long changeover time as a fact of life.

Building in batches stunts the growth of your operations. Don’t combine kanbans and build a group of them. Produce in the same sequence you get the kanbans.

Again we had to wait until the last minute to learn something important. It just showed us that we really weren’t growing as much as we ought to have been growing.

Getting Back to the Point

I suspect that there are still a lot of areas where people collect kanbans (orders). “We’ve figured out the pattern, and we will build batches in this logical fashion” is not what real improvement looks like.

Here are some definitions of proposed solutions:

• Batch logic. Collecting the kanbans (replenishment signals) until you have a group of a certain size and then making the order

• Pattern production. Putting the kanbans given to you in a certain sequence that makes the changeovers easy to do

These solutions may seem like a great idea from the manager’s or supervisor’s perspective, but it’s important to remember that those who live by patterns will die by patterns. Collecting things in piles or collecting kanbans can make things easy, but anyone can do that. That is not real kaizen.

Unfortunately this idea of collecting kanbans and then starting work has become quite commonplace. Even if changeover times have been reduced, the batch sizes remain the same, so the time saved is just used for more production. I understand that we have to be competitive in the global marketplace and that each area wants to look good, but I would strongly advocate moving away from local efficiency to an overall efficiency in which stoppages are lessened and the overall flow is taking shape.

If you allow people to collect orders and work in batches, they start thinking only about what is easiest. Stick to the principle of making each order in the sequence in which you get it. Ohno told us that the minimum kanban size should be five. We kept doing kaizen until we could flow things in sets of five. That was tough, but it’s something we should not forget. Production and mechanical engineering are about fighting the constraints to flow. This incident reminded me never to forget that it’s a fight, and a fight isn’t easy.

Lesson 5: Nine Out of Ten, One Out of Ten

Unless there is a major issue, nine out of ten managers can increase their productivity numbers when volumes are increasing. However, there are very few managers who can increase their productivity numbers when volumes are going down; even keeping the productivity numbers constant in such a situation will be difficult. Maybe there will be one out of ten people who can do that. “Well, it just means that there really aren’t many people like that” is what Ohno used to say frequently.

When volumes are down, most managers will say, “When I adjusted my manpower to reflect the new volumes, even though the line was connected, there were a number of locations where we couldn’t divide up the work properly. That caused some waiting, which is why I have more overtime,” or something like that.

Actually, this is a problem with the layout of the line. The truth is that those problems always existed, but because their effects were minor, we just ignored them. But when the volumes went down, the problems surfaced and caused lots of waiting.

Ohno noticed this phenomenon and pushed through the concept of “productivity keeping lines even with low volume.” In Japanese, we called it making shojinka lines. The first model line at Toyota was the transmission line used in small trucks. It has machining and assembly processes. Use the process that has waiting as the end of the line, and lay out the work so that it is in a straight line. We used to call this line “an eel’s sleeping quarters” because of its straight and narrow shape, or the “one brush stroke line” because the movements of the parts could be drawn with one brush stroke (there was no need to go back or out of the area). We changed the number of parts that were being machined at one time, automated the setting of the parts, and even added auto-kick-out mechanisms on the machines to eliminate the waiting areas.

One person moved from right to left through all the processes. When she had gone through all the stations, all the necessary parts for one car had been made. What that means is that we modified the processes so that they would make only one car’s worth of parts.

Our engineering staff worked on the line, too. Its job was to find out what the issues were and how things could be improved. Because of this effort, when production volumes went up or down, we were able to adjust the number of people in the line and keep the productivity in the line constant. It was our first shojinka line.

In the automotive industry, production volumes are adjusted very often. First, you develop your productivity on low-volume lines and have all the waiting time at just one area of the line. You will be a lot more effective if you develop a strategy for kaizen—are you going to focus on reducing machine breakdowns or the labor time? I will keep repeating myself, but because all the “remainder” time is in one location, no matter where you improve the line, if the saved time equals one person’s worth of time, you can now take out one person from the equation. If there are many branches in the line, that means that there are many remainder points for labor time. You may have made certain processes more efficient, but this won’t result in a cost reduction. The first step is to take away all the branches and make it one continuous line. It will be a lot easier to work on reducing the remainder time and getting your cost savings in terms of people if you do this.

Any improvement you can make when production volume is down will be multiplied when the volumes go up again, so use the extra time you have to work on issues that you didn’t have time to look at when you were busy. The low-volume time won’t last long, so work hard and get the results. This is really the test of a good manager.

When production volumes go down, it’s not that we have more machine stoppages. It’s just that they affect us more. Productivity going down is something that we would expect. However, if you are constantly improving the work area by doing such things as reducing machine stoppages and reducing the number of hand tools needed, you will have a great bench strength in your area. When production volumes go down, I can see which managers are really doing their job. Also, depending on how much effort you put into your work area when volumes are down, you will be rewarded accordingly when volumes go up again. To get an environment in which you get the “one in ten managers,” you need to deal with the reasons why people and parts can’t move.

Getting Back to the Point

I wrote before that when volumes are down, you have a chance to work on things that you normally cannot, but it really is true. It’s actually a very welcome opportunity. Business success largely depends on whether top management notices the opportunity or throws it away and lets it pass by. Actually, most managers do notice and do take action, but they don’t realize that the action they are taking will affect things negatively. Examples of this are letting people take time off or making the production early and shutting down for a day.

Here is what you should do: first, collect all the extra people. Then look at the flow of money. How many orders do you have out there? What are you asking your suppliers and contractors to do for you? Look through them one by one. During the period when production is going down, few things are likely to be urgent, and for those that are, there will always be a reminder from someone. Can we do something a different way? Can we make it ourselves? Is the construction we are asking for doable by the extra people we have? A company with a lot of history will have lots of opportunities like this. Once you do that, your extra people are no longer extra. They are now delivering value to the company.

Also, this is a great opportunity to make some jigs and simple equipment, and maybe even automate some processes and try to do them yourself. When you become busy, the people will go back to the line, but the experience of making something themselves will stay with them. That experience will allow them to use those short downtimes to make something that will increase the speed of improvement in your company.

In Japan, we have a saying that goes, “A disaster can turn into a blessing.” I say, “A disaster can turn into growing your people.”

At the very least, the factory should stop the unnecessary, nonurgent jobs that are going to outside contractors and try to do those jobs itself. That will also tell you whether the price you are paying for those services is a fair one. Making the “one in ten” managers is not just about gaining experience in kaizen, but about allowing managers to think about the risks associated with their work areas and preparing for the future. This will include making a “model line” that limits the current risk. There is a need for a bit of experience to estimate the risks that arise from internal and external factors, but at the very least, managers should be given an opportunity to think about and plan what they will do when the environment they are working in changes.

For example, there are many companies that make their goods overseas in low-cost countries and then ship them back. What they should be thinking about is, “What will we do when the labor rate goes up?” which should include making a new line or developing new equipment. “Don’t worry, we’ll just find another low-cost country” is not a way to grow good engineers. The companies that think of ways to make things happen even with high labor costs and that build equipment and lines to make those things happen are the ones that will keep growing.

Lesson 6: The Foreman or Leader Is the One Who “Breaks” the Standard (When You Make an Improvement and You Can Take Out One Person, Give Up Your Best Person)

In the 1940s and 1950s, when Ohno was the manager of the machine shop and was starting kaizen, there were no such positions as team lead or indirect staff members under each supervisor. Rather, each supervisor just had his direct reports.

These supervisors were called “shop daddies” (oyaji). They controlled people’s pay and wielded a lot of authority. They were quite a force to be reckoned with, and even Ohno was not able to tell them what to do. Because of that, Ohno would work with one of the “skill men” (workers) on the line who he felt was a high-potential person.

After the improvements started taking effect and one less person was needed on the line, Ohno would take the best person—the person he had been working with—and remove him from the line. I think we can see how well Ohno was able to understand the needs of the future, his ability to develop and motivate people, and above all his ability to put a structure in place that would ensure continuous improvement.

The concept of the “outside line man” started here. Ohno used his “outside men” to increase the speed of kaizen, which also increased his staff of “outside men.” After a while, Ohno designated some of these people as “team leads.” (They were called Ohno leads back then.) At that time (1950–1951), Toyota didn’t have any position in its HR structure called “lead.” It took a few years (until 1954, to be exact) for Toyota to officially recognize the position of lead. From this time on, the outside line man was officially recognized as a person who was to do kaizen. On the one hand, the supervisor developed the standard work and taught it to the workers. He showed it, had them do it, and made sure that they did it properly. In other words, he set the standard and enforced it. The lead learned the standard taught by the supervisor, thought of better ways to do the work, and worked hard to get permission to try his new methods. In other words, he was there to break the standard.

At Toyota, the standard is there to be changed. The supervisor ensures that it is followed; the lead is there to improve it.

Getting Back to the Point

The outside men didn’t stop at just being leads. Ohno took a group of them who were good at making equipment and created a maintenance department whose job was to do kaizen. These people were there to make the changes that those on the shop floor asked for. They had worked on the line before, so they understood the needs of the shop floor very well. They made simple automation devices, equipment to shorten changeover time, and machines that finished the job within takt time and allowed the worker to keep going to the next process (we call them nagara equipment), and a whole host of other things that greatly advanced the implementation of flow in Ohno’s Toyota Production System.

Making an improvement that can take one person out results in just one person’s cost being saved. If you take that person and have her make improvements, you start getting savings of two, three, four, and five people and so forth. Taking out the best person and making her improve the rest is really effective.

When a new line had to be made, those people became the supervisors, leads, and workers. They made the equipment that was to be used as well as the racks, part storage devices, and covers, and they did a lot of other work to prepare the new line for production. Once the line was up and running, many of them would stay in the line as workers.

When I joined the machining department, there were about 30 people in the group who made many useful things such as simple automation devices or a machine that could do work that was difficult to do or that required heavy lifting. The parts needed for these machines had drawings, so the workers knew what to make, but how to put all the parts together was just in their heads, so they were really fast at getting things done. I am sure that even now, many of the things the current workers are using were made by this group. During busy times, they would join the production line and help out with the extra load. This kept them from forgetting how to do the work and at the same time gave them an opportunity to find more hard-to-do tasks that they could fix when volumes dropped again. While they were working, they had their eyes open for what else needed to be fixed. This greatly helped with improving flow and improving the standard.

Lesson 7: Multiskilling Means Learning the Next Process—Keep It Flowing Until You Reach the Last Process

When Ohno came to the gemba, he would look at the one repetition of the workers’ tasks. Often this would result in the creation of a “relay zone” where the worker would hand over the part he was working on to the next worker. One day when Ohno was looking, the worker doing the first process had finished his work and was in the relay zone. As it happened, the next person was not in the relay zone, so the first worker was unable to give it to her. Instead, he placed it in the area and went back to the first step of his cycle. Ohno then said to us, “The worker after this process is delayed and so wasn’t there when she needed to be. If the process after you is delayed, why do you place more work into it and just leave? You haven’t done proper multiskill training.” He continued, “Multiskilling means that you teach the person the next process so that you don’t stop the flow of work. Make sure people are helping out those who are delayed.”

Up to then, all we had been taught was that you should have a relay zone so that you could hand over the baton to the next worker without losing speed. This was to buffer the normal variances we see in work. Ohno’s instruction to us was to keep going until the final process, even if there was a problem downstream. If you stopped halfway, work in process would start piling up. “Don’t pile up work; make it flow” and “do your best to keep the flow” were the lessons we took home.

Getting Back to the Point

Normally we are told to give the piece of work we are working on to the next person by hand in a zone much like the one in a relay race where each athlete passes the baton to the next runner to keep the flow going. In this case, Ohno said that if there was some issue preventing the next person from being in the relay zone, we should continue to work until we bumped into that person. When we had handed over the work, then we should go back to our own area and wait. This type of operation is very difficult to achieve, however.

Real multiskilling involves teaching and training the person to do all the work in the sequence of processing in his line. That way, we are preventing stoppages in the flow. The reason why we insist on transferring the piece being worked on by hand is that it clearly shows the differences between the speeds of the two workers, and the relay zone acts like a buffer to limit the damage caused by the speed variance. If we put the work pieces on a conveyor belt or if we put some racks between the workers, it becomes hard to see the differences between the speeds of the two workers. That is why we use relay zones.

This method is mainly used in machining or processing, where the work mostly involves carrying parts and placing them in the equipment, taking parts out of the equipment, and inspecting the parts; in other words, it is mostly nonprocessing work such as you would find in machining lines.

On the other hand, in assembly or fabrication, where the worker is doing most of the processing, we would not do this, since the risk of using wrong parts, missed processes, or missed parts would increase. If there was a delay, an extra person (an outside line person) would come into the line to help out, and if that wasn’t enough, the line would just stop.

Going back to multiskilling in the sequence of processing: historically, when Toyota had its workers work on multiple machines of the same type, the productivity increased, but the amount of work in process didn’t change much. When the company relaid out the equipment in the process sequence, however, the work in process was reduced to one. This was a huge change in the operations.

This change forced other changes, such as auto feed and auto stop of the machines, which up to now had been done manually. That freed the workers from having to babysit the machines. It also simplified the work sequence and allowed people to run many more machines.

After some time, the number of cars being made increased. The production lines then morphed into automobile type–based lines, which further reduced work in process. During this evolution, a change in our management and the multiskilling efforts for the workers took place. Up to that point, it had been OK for a worker to know how to operate one machine or make one part. Now the workers had to learn how to operate all the machines and be able to make all the parts for the car. People now understood how everything fit together. Once they were promoted to leadership and management positions, they were able to make fast decisions and take on leadership because of this knowledge.

Processes that had previously needed special skills became fewer and fewer, creating a line where “everyone was welcome.” This allowed us to hire from a much larger base of the population.

Just-in-time was obviously a factor, but I also think that multiskilling and the product-based lines are historic in that they largely contributed to “making things in a set.”

I have many opportunities to visit different companies, and lately I have been seeing a lot of multiskill boards. The people get a circle for each of the processes they have mastered, and at a glance you can see who has what skill. The most important thing, however, is flow. I think you will get a lot more use out of your multiskill boards if you can identify where flow is disrupted for whatever reason and link that with the skills training.

Just to be perfectly clear, multiskilling is an activity that we do in order to increase flow. Keep that in mind when you are using those boards.

Lesson 8: What’s That Red Circle on the Top Right of That Graph?

We used to have a monthly meeting with Ohno with all production departments present. In the early 1970s, the department whose turn it was to present would show a six-month productivity increase plan, and the goal was shown as a red circle (see Figure 1-1). This was the conversation that took place between Ohno and the department manager over the red circle:

Ohno:        What’s that red circle on your graph?

Manager:   It’s our goal for productivity increase this year.

Ohno:        Can you do it?

Manager:   We will work hard to achieve it.

Ohno:        Do you know what you need to do to achieve it? Is it possible?

Manager:   We will work hard and do whatever we need to do to achieve it.

Ohno:        You know what’s going to happen four months from now and you say you can do it?

Manager:   We will do kaizen every month and work hard to achieve it.

Ohno:        We don’t do monthly kaizen here. Every day, every day, fix a problem or eliminate a barrier to flow. The results are what you put into it, and we are not here to do kaizen to achieve a number. If you can make a monthly plan for kaizen, why don’t you just do it now?

The lesson was that achieving goals is good, but don’t forget the process because you have both eyes on the numeric goal. I think Ohno was telling us not to fall into the trap of “target figures.”

It’s not as if you can do anything to increase productivity. If numbers are very important, we may be tempted to do whatever makes the numbers better. However, if we do that, we may be just delaying addressing the real issues. The more the managers say “targets” and “numbers,” the more likely we are to get more people on the gemba becoming “number matchers.” As Ohno said, productivity is just a result of our kaizen efforts. In other words, “Limit overproduction, do it with less people, and increase productivity.”

If there is a department that has not met its goals, then the manager needs to go there and work with the supervisors to eliminate bad conditions and stoppages to flow. The manager’s job is not to manage with numbers but to understand the reality and fix issues.

I think Ohno gave us a really wise lesson that day. It was: “Your work as managers is on the gemba. If you can’t reach your targets, go to the gemba yourself and make the necessary improvements. Find the issues that are giving your people a hard time and fix them. If you can do that and work on flow, the results will come naturally.” Or, to take it one step further, I think Ohno saw that target management kept managers away from the gemba and directed their attention only to numbers, and he wanted to put a stop to that.

Getting Back to the Point

I am convinced that the spirit of Ohno’s way of achieving higher productivity was, “Profit is a result of focusing on flow.” If you just keep focusing on the flow of materials, of course profit will follow. There are a lot of situations in which people are discussing things like “Let’s change that machine to a better one and get an extra hour a day, an extra 20 hours a month” or “Let’s automate that machine and reduce the manpower.” These discussions take place in an office, and people are adding up hours in their heads. I feel that my biggest lesson from Ohno was not to add numbers in my head, but to go to the gemba and make the changes happen.

When you are on the gemba, you see many places where things are stagnating. Eliminating these stoppages to flow will increase our technical skill, of course get us more productivity, and make us more competitive as a company. I think many companies make their operations weaker by saying, “Just focus on making money,” or, “Just give me results,” with a short-term profit-minded view. When you make a profit number the target, there are many ways of achieving that. For example, maybe we can outsource the rework of a product to a cheaper company. There are many examples like this. Yes, we got the extra profit, but now we have lost the opportunity to fix this problem.

Focusing on making money is not a way to long-term success. It also doesn’t make you stronger. However, we can always do kaizen. Happily, in my company we had the philosophy of the Toyota Production System. This is heavily inculcated into all employees’ hearts. Making things flow and managing overproduction are what has to happen in order for the process of meeting the goals to be acceptable.

Profit comes while pursuing flow. In Japan, we have a saying, “Profit comes after good business (with the customer).” I think the same thinking permeates Ohno’s Toyota Production System. In fact, I think that in Ohno’s view, the “customer” is none other than the worker who is adding value. There is a very famous saying of Ohno’s that is known beyond Japan. It says:

Once you do it, once you know the way, you can’t help it; you have to do it. That is the kaizen spirit.

When I read this line and I look back at everything I learned from Ohno I feel him saying:

The only person adding value around here is the person on the line. Don’t you feel sorry that you are making them do non-value-added work? Hurry up! Free them from meaningless tasks, get rid of things that hurt them, and make them more value-added. That is the way you live in kaizen!

Lesson 9: Are You as the Manager Having Them Do It, or Are They Just Doing It Their Way? Which Is It, Man?

Compared to the average worker or engineer on the shop floor, Ohno was a lot stricter with the shop floor management, especially the first-line managers (people who were not working on the line). Once he noticed that there were two pieces of work in process in an area where the standard work said that there should only be one. He called the manager there and said, “Why are there two pieces here? Did you tell the worker to work this way, or is she doing it by herself? Which is it?” That is where this quote came from. I was trying as hard as I could to be invisible while watching the poor manager get grilled by Ohno.

The manager was toast whichever way he answered the question, so he remained silent while getting hammered with, “Which one is it, man? Which one?”

The reason why he was being grilled was that the standard work (takt time, work sequence, and standard work in process) was decided because that way anyone would be able to do the work safely and correctly. That was clear, but if the person was not doing it according to standard, all that effort to establish a standard was useless.

The role of the supervisor is to establish the standard work, teach people how to do it, and ensure that they are doing it to standard. If the work is not being done to standard, then retraining is needed. In the same way, the manager was supposed to look at the standard work that was displayed and make sure that people were doing it that way. If not, correction was needed.

When there was an extra piece of work in process where there was supposed to be only one, it meant that the supervisor was not playing his role. If this minor infraction was let go, it would send the message to all the lines that it was OK not to keep exactly to the standard, and we would then have a workplace where standard work was not followed.

I remember that Ohno was especially strict in making sure that the shop floor standards were being followed. I feel it was because he wanted his managers to display strong leadership in making good products, safer, faster, and at a lower cost.

Getting Back to the Point

Ohno got terribly upset when standard work, which the gemba was supposed to develop and then manage, was not followed. He was extremely hard on the managers of the shop floor, as I witnessed numerous times. Quite often, when we were giving a presentation on one line and he saw something wrong on another line, he would ignore the presentation, go to the other line, and have the people there fix the problem.

Once the people had mastered the standard work, the next step was to take out a person or reduce inventory and then figure out the whole process again. Because standard work was the basis of improvement, trouble was sure to follow if the standards were just posted but not followed. When you visit the shop floor, you can immediately see how well the manager is managing her area. In fact, going to the gemba is a great way to see how good a job the manager is doing.

There was another reason why Ohno was so adamant about standard work. The standard work told us the takt time, which was the pace at which we had to make things. Different cars had different takts (paces) to which we had to make them. With everyone working to takt time, we could level out the production and the whole factory would have a sure foundation for making things smoothly and efficiently.

Lesson 10: Standard Work for the Andon Is, “Go There When It Flashes”

This was the answer Ohno gave when someone posed the question, “How should we define and show the standard work for the outside line people?” during one of the meetings with the production staff.

For people on the line who are working in a cyclical fashion, showing one cycle of work on the standard work sheet is enough, but for areas that are heavily automated or for people who are there to solve problems on the assembly line, defining their work was a lot more difficult, and hence the question. These people’s main job is to go where the problem area on the line is (as shown by the andon) and fix it.

What is the standard work for things when you don’t know when they will occur or where they will occur? The answer was quite simple: “Once the andon flashes, start moving; go fix the problem.” By the way, we said the same thing to our maintenance staff—”Go to the gemba that calls you”—and to the transportation staff—”Be at the display board and deliver parts in the sequence in which people call you. If there are no requests, just stay there and wait.”

I know many people who would be extremely uncomfortable with the idea of people just waiting—and there were many in Toyota at the time who felt that way, too. I felt that what Ohno was trying to teach us was, “Don’t make things complicated.”

“When it flashes, go” is actually the same instruction that the kanban has, which is, “When the kanban comes off, go and fetch some parts,” or, “When the kanban comes off, make some parts.” “When one kanban comes off, make it,” and, “when the andon flashes, go” were both important teachings of Ohno’s spirit to do things in a standard size.

Getting Back to the Point

“Don’t waste your time making standards when no matter what you write, you will always be wrong, as things won’t work out that way. Don’t make the tools on the gemba hard to use.” Keep things simple and clear, and think of how you can reduce the number of times people call you. If you do that, you will improve the flow.

In fact, the outside line person started out as a person whose main job was to make improvements. If you give these people some work to do when there are no issues for them to deal with, they will not be able to address problems when they occur. While it may look good on paper to have the team leader do actual work when there are no issues, this puts too much pressure on the system and causes dysfunction in real life.

Even as it is, there are a lot of calls for help, so if the team leader is doing actual work, she will just make more people wait. We are trying to make flow, but now our people are the biggest obstacle to it. Why did we put an andon up in the first place? The whole logic of the andon system starts to collapse.

When the andon is not flashing, the team leader’s job is to find the root cause of the last andon flash and figure out a way to keep it from happening again. That’s it.

Let me give you the historical background of the andon system and what the thinking behind it was. The first andon was put on the engine machining line in Toyota’s Honsha plant. After more and more andons were put on lines, the company standardized the light colors. When you felt that you were going to be late, you pushed the call button, which lit up the andon yellow light. If you could not catch up during the allocated time, then the light turned red and the line stopped.

We also added andons to machining lines, which were automated. We divided up the line into sections, and when a part didn’t go through a section, that andon lit up yellow. You may think that it should have been red, but we felt that this meant, “The machine is calling us for help,” so we kept it yellow. After that, we differentiated between abnormal (machine breakdown) and normal (changing of cutting tools, quality checks) issues and made the normal stoppages light up white.

Based on this, the standard became the following:

Red. Red is for fire extinguishers and fire hydrants. As much as possible, don’t use red, as we want to keep it special for emergencies. For us, red was used only when there was a stoppage on the final assembly line. That meant that product was not leaving the factory. Red is the most important color. Drop whatever you are doing and go fix it.

Yellow. This was the color that meant a “request to come over.” Whether it was a machine or a worker that was calling for some help, the color was yellow. Yellow is the color that tells you to communicate with the worker or the machine and think about how to prevent the issue from happening again.

White. Do things properly, make an improvement to decrease the frequency with which it calls you.

I really wanted you to see how much thought went into just deciding on the colors of the andon. There are many companies using andons. It is important for things to change with the changing environment, but I wonder if the same amount of thinking is going into what the colors used should mean. Red means drop whatever you are doing and run; yellow means how can you make it not flash again; white means do it properly and improve it so that you don’t have to do it so often (for example, reduce the wear and tear on the tool so that you don’t have to change it so often).

I hope you understand the thinking behind this. The more people you have who understand this thinking, the more people you will have who are driving improvement in the company. Ohno was very interested in the andons. If the andon lit up while he was visiting a factory, Ohno wanted to know why the machine had stopped and had people find out the root cause. If he saw a machine stoppage with no andon light going off, he would wait there until the machine had been fixed. He called those instances “lying andons” because when the andon didn’t light up, people thought everything was OK.

One of the worst situations in a production facility is not being able to tell whether things are standard or out of standard (normal or abnormal). When Ohno started on this journey, it was hard to tell whether the situation was under control or out of control. I think Ohno’s strictness came from a strong resolve to never go back to that kind of situation ever again.

Lesson 11: Standard Work Is the Foundation of Kanban

This is an excerpt from Ohno’s writings to commemorate the fiftieth anniversary of the Honsha plant in 1988:

The supervisor must be able to correct and improve on the worker’s incorrect work method. That’s why I had them put up the standard work above the line. This is the root of the kanban.

Ohno is saying a number of very important things here.

The standard work can be useful to the worker if he is confused about his job, but it is more of a visual management tool than anything else. The manager will check to see whether the work is being done according to the posted standard and, if there are mistakes, to correct them. That will keep good-quality production. At first, people posted the standard work on the line and called it kanban. (Kanban means “sign” in Japanese.) It is a tool to keep the facility on the path toward quality, safety, and lower cost by continuing to update and “shape up” the standard work.

“Standard work is the root of the kanban”: actually there is no record in either Ohno’s writings or things that he said that tells us exactly what he meant by this. “When we received the Deming Prize in 1965, I thought of kanban as our catchy phrase” and “I made up the word kanban to make it harder for foreigners to understand what we were doing” are documented, but neither of these really tells us why standard work was the root of kanban. However, I think that the fact that Ohno wrote this on the fiftieth anniversary of the Honsha plant (10 years after his retirement) means that there is an important message there.

When Ohno built the just-in-time system using “downstream pull,” he changed the label on the parts to “kanban.” And he changed the original kanban, which was the standard work chart that was put up on the line, to the standard work vote (both pronounced the same way in Japanese) to show that it came from the supervisor’s intention. By doing this, he eliminated the need for a ledger to manage the shop floor. Full managerial control was now at the shop floor level, and I am sure Ohno was very satisfied with that result.

When one continues to read what Ohno wrote in the same document, he ends by saying, “I feel that in the future, we need to value the system that allows people to work without mistakes even if done slowly. Also, the supervisors need to value their employees and create their own managerial structure.”

Getting Back to the Point

What does “standard work charts are the root of the kanban system” really mean? I’ll share my thoughts on this. When we look at Ohno’s work, we see that he takes work that is normally done in a management department and moves it to the shop floor. Even today, in many companies, something like a standard work chart is drawn up by engineers who are in management positions. The andon, part shelf labeling, management, and fixed locations are all in some way moving management from the managers’ office to the gemba. I think we would find more examples of this kind of thinking if we looked more.

My personal experience was one in which we moved the part making and assembly line of a unit to another company. I worked with the other company’s engineers on the move plan, the training plan, the layout, the production structure, and so on, and things pretty much went according to plan.

What we changed was how we ordered parts. Instead of having orders come from an office to the supplier, we installed the kanban system with them. They just made parts to replenish their finished goods. This was good for us in that it allowed us to change the sequence farther downstream than before.

A number of days later, someone told me that Ohno came by and said, “That is good. We eliminated a controlling office.”

Basically, things do not go according to plan. If they did go according to plan, you wouldn’t need managers. The kanban system is based on this logic, and it allows for fine tuning based on changes to the plan. Because of this, the “management infrastructure” needs to be located on the gemba, which is where those changes happen.

“I would like to see whatever the gemba can do itself moved to the gemba, and if there are changes, they should not come via some management office, but rather be done on the gemba in a fast and effective way.” I think that is what Ohno wanted to share on the fiftieth anniversary of the main plant. Here is his entire column:

Around 1963 the kanban system was spread throughout the entire factory. It didn’t start with a master plan, so even we didn’t know how it would develop. Concepts like “Increase productivity while preventing overproduction and do it with less people” were ideas that could not be calculated. No one was able to understand it from the explanation, so I had to go to each area and give detailed instructions.

In the Toyota Production System, when one goes to help out another department, everyone must be able to do the work the right way, and if they are making a mistake, the supervisor needs to be able to instruct and improve the work. Because of that, we put up the standard work charts above the line. This was the root of the kanban system. Lately words take on a whole new meaning by themselves and get before the real meaning of things, but the Toyota Production System actually started out quite simply.

Placing importance on creating a structure of being able to “do the job correctly, even if slowly”; having supervisors praise and recognize their workers; enabling supervisors to follow through on their managerial goals will become more and more important. I am hopeful that the many changes that will come in the Honsha plant will be met with the appropriate response by its people.

Lesson 12: When the Worker Pushes the Start Button, He Has Stopped Moving. Can’t You Guys Figure Out a Way for Him to Push Start While Still Moving?

When I was in the house (the office for the production staff), the manager in charge of kaizen received a phone call from the shop floor saying, “Ohno wants you to come.” Our thoughts were, “Oh, no; did we do something wrong again?” as we ran to the place where Ohno was. He was looking at the transmission machining line, and when we got there, he said, “Look at the movement of the worker. He is stopped when he pushes the start switch.” We were not too sure how to respond, as that was such an obvious thing and we had no idea where he was going with this thought. Ohno then said, “He attaches the part to the machine, walks a bit, stops, and then starts the machine. Can’t you guys figure a way out where once he starts walking, he doesn’t have to keep stopping to start the machine?” He continued, “Hmmmm, elevator buttons work when you just lightly touch them. I guess you have to stop if you have to push the button on the machine that hard, huh-uh-huh?”—meaning that the worker had to stop because he had to use a lot of force to push the start button.

Ohno knew that we had been using this same switch for the last few decades. The maintenance manager and I had both been preparing ourselves for a tongue-lashing, and we were relieved to find that he was only talking about switches. Those on the kaizen staff are always making things for the shop floor, and during busy times they go and work in the line. We were told to put this experience to full use and make something that is easy to use.

In the end, we didn’t use the touch switch on elevators that Ohno was talking about, but instead used a limit switch. In the past, we had developed a cam to stop the switches from breaking, so we put this experience to use and in two days had a model to show Ohno. He saw it and gave us the OK.

When we actually used this, the worker tired a lot less and the walking time went down by around 10 percent, which was a big surprise for us. After this, this limit start switch concept was used throughout Japan.

Maybe Ohno was thinking, “Elevator switch—touch only—we can use that somewhere—oh, yes, start switches,” which I think is why he said, “Can’t you guys figure out something huh-uh-huh?” Whenever Ohno thought of something, he said, “Come to the gemba,” and he used his own pool of people to squeeze their brains until they had a good idea of how to make things flow. Parts were flowing and workers were flowing. I think he kept the ideal in his mind all the time, which is why he was able to give these kinds of directives.

Getting Back to the Point

There was a kaizen group in each production department. They were made up of people who had a special talent for making things, and their job was to make things that the shop floor requested that would make the job easier. They really got excited when the request was for something that had never existed anywhere else. There was a difference in the talent level in different departments, but that difference created a healthy competition, and every year these groups would bring new contraptions into the world. We had high expectations of the kaizen group, and it always seemed to deliver beyond our expectations.

In the organizational structure, the group was between the direct and indirect departments, so during peak times it just maintained a skeleton crew and the rest of its members helped out with production. We encouraged them during those times, telling them that it was training to find more useful things they could make for the people. We also took particular care that they weren’t kept too long in production so that their equipment-building skills didn’t decrease.

We felt really bad that until Ohno told us, “The worker has to stop to push the button,” we hadn’t noticed anything wrong. Because of that experience, however, we were determined to make improvements before Ohno could say to us, “Hey, this is a similar situation—fix it!” One day we noticed a worker holding the impact wrench in place, and so we made a switch that allowed him to walk away once he had set the tool on the nut.

Automation in which something is set manually is called “one-touch automation.” It is much cheaper than full automation, and it can be done a lot more simply as well. Because of Ohno’s setting us on fire, we were able to see a lot more opportunities for this one-touch automation and were able to make many similar improvements.

Lesson 13: You Bought an Expensive Machine, and Now You Want an Expensive Foreman or Engineer to Run It? Are You Mad?

It was at the chassis factory and we had just installed our first transfer machine. It was an expensive and important machine, so to learn how to use it quickly, we had put our best supervisor on it. We thought that by doing that, we could get to a safe and stable production level quickly. I suspect most companies would do the same thing.

However, when Ohno came around to see how things were going, he saw the leader working very hard. That scene told him, “We are working very hard, and I am the only one who can use it,” and we heard the words that are the title of this lesson.

The people who were working on the installation—the production managers and the engineers—were all at a loss as to why he would say such a thing and didn’t know what to do. At that time, one of the manufacturing engineers, who was kind of like an interpreter for Ohno and had taken a lot of concrete actions based on Ohno’s verbiage, understood what Ohno meant and explained it.

We started out with a machine-based layout; at that time you needed a special machinist to operate each machine. Then we added intelligence to the machines, separating the manual and the machine work so that you didn’t have to be a specialist to run the machine. Now, you’ve spent even more money and you’ve put your best person on the machine! Besides, if we look at what this supervisor is doing, he is just trying to restart the machine when he runs into a problem instead of solving the root cause. Right now, if he takes a day off, production will stop. Let’s clarify the problems and fix them so that anyone can run this machine.

By convincing everyone that this was what we had to do, and by setting the standard at where a worker on the line could run the machine, we were able to clarify the issues and fix them one by one. The supervisor was now able to be offline again and able to support the direct people.

When we first purchased the transfer machine, it was labeled as a “one-person-control machine,” meaning that one person was needed to operate it. Because of this experience, the standard was changed, and the next transfer machines we purchased were all no-person-control machines.

Getting Back to the Point

I think most people understand the kind of thinking mentioned before, but all of us are worried about whether things will go well once we put this thinking into practice—even more so when we have some production equipment that has a deadline to be met. I think Ohno was thinking the following:

I know they are just throwing a supervisor in there and making it his problem. That’s not doing us any good for the future. This is a great opportunity for me to show them what kind of mindset we need to have.

This is why he said what he did to us. When I heard this story, I remembered another one of Ohno’s hard teachings. It was:

When you make an improvement and you can take one person out, you must take out the best person from the line.

There are three good reasons to do this.

• The person will now be able to look at her work from a different perspective, which will make it easier for her to fix it. Quite often, a good employee is subconsciously compensating for problems, so she doesn’t even know that those issues exist.

• You get the most powerful force from the line and focus it on improvement.

• The people who are left are motivated to be taken out of the line, so it helps to create a positive attitude as well as seeing change as a reward.

This is what we had been doing, but then suddenly our thinking changed when we bought a new machine. It is really important to keep repeating the principles and thinking to the people who work there; otherwise, they will be forgotten. I think there are many people who are very thankful for the people who came before them for creating the kind of work atmosphere they now enjoy.

Lesson 14: Engineers in Production Become the Horizontal Threads in the Cloth

In 1960, Ohno brought together the engineers who were assigned to the production department and gave them the following lesson:

The important departments required to make a car, such as design (function and quality of the car), procurement, and production engineering (equipment function, procurement price, duration) are like the vertical threads in a cloth. Yes, we can build a car with just these vertical threads, but no matter how well we organize these departments, with each one working toward its own goal and toward what is best for it, things will not work out, and we will not have a good production system. As a result, we will be making things in a sea of muda. The processes to make a car are very long and hard to see. There is muda happening everywhere to such an extent that no one can really measure it.

You engineers in production need to look at the entire process flow and look for the overall efficiency. That is the important job given to you. More specifically, look at the four elements (processing, inspection, transportation, and waiting) that directly affect quality, safety, cost, and delivery. Grab the vertical threads strongly and push them to work together to make good products at a lower cost and faster. That will make us a strong cloth. Then take the results of your work and give them as feedback to the vertical departments so that they can learn from it and use those lessons in their next project.

With this talk, Ohno showed the engineers what was the ideal state of production. It was that the gemba (where things were being made) took leadership and, without fear of failure, through trial and error made a better working environment. He also forcefully pushed this point with the engineers in production. Because Ohno took his machinists, made them the horizontal threads, and was able to get significant results, the whole company took notice, and in 1958 all five production departments and the related production engineering and inspection departments got a horizontal thread group added to them.

Engineers in production become the horizontal threads in the cloth—I think this is a great saying, and it still holds a lot of significance even today.

Getting Back to the Point

Right now, we think that things like layout by process sequence, low manual labor content equipment, downstream pull, and set load transport are just logical and make sense. But all these concepts were the result of a lot of working through a lot of difficulties and hardships. To bring that point home, I want to share, using a couple of examples, just how un-just-in-time and how un-jidoka-like Toyota was between 1945 and 1965.

Back then, people got monthly production schedules, and in order to be efficient in their own processes (there is quite a big difference in the definition of efficiency nowadays), they would do the following:

• Try to have the smallest number of changeovers—big batch production.

• Avoid problems caused by parts shortages by having lots of inventory and overproduction waste.

FIFO was impossible with this kind of thinking; no one knew what kind of analysis to do when quality issues were found, and the labor hours required to “get back to normal” were very expensive.

On top of that, people’s pay was determined by how much their team produced, so the more production they did, the more the employees made—an obvious incentive to overproduce. Of course the fewer the number of changeovers required, the more they could produce. I am sure that in such an environment, the concept of making only what was needed today, then changing over to another product fell on deaf ears.

The overproduced parts weren’t even put on shelves in an orderly way so that one could access them easily, so just trying to find the right parts was extremely painful. Today, the parts marshaling areas (Toyota does not use the term warehouse) are not there to collect parts; they are areas designed to ensure the transportation of the “correct parts in the correct volume at the correct time.” Back then, however, the marshaling areas looked like anything but this. They were not organized, and things were shoved into any small space that opened up. Because of that, a great deal of time was spent searching for, finding, and trying to gain access to the parts. Ohno started out by putting parts in order on three-level shelves.

It was in this kind of environment that Ohno forcefully pushed through the concept of flow. By incorporating jidoka into the equipment, laying it out in process sequence, and moving processes that were traditionally outside the normal flow, such as plating, heat treating, and painting, into the overall flow, he eliminated transportation and work in process and eliminated the concept of teams of people who did only one process. In such ways, he kept removing barriers to flow. These were all stories I heard from the people in the machine shop.

After this, the concept of flow kept advancing. Parts that required the same kind of skill to make were brought together to be managed by one team. Afterward, a waterstrider (mizusumashi) was used to join the assembly line and the parts-making line by pulling one set of parts at a time. In another upgrade, the parts required for a certain model of automobile flowed continuously into the main assembly line. This then became the foundation of the modern unit line.

As the concept of flow started incorporating a wider area, the area to be managed also increased. The Japanese concept of “promote from within,” which everyone believed in, helped people use their talents and experience, which in turn produced good business results. This is why we often say, “Making things means making people.”

From here on, jobs will get more specialized and individuals will be held more accountable, which I am sure will lead to a lot more silo mentalities (in Japan we call this sectionalism). In a factory, the work done upstream all flows down to the factory. If one looks upstream, he can see how those departments are thinking. If people aren’t working as a team or cooperating, that will at least cause problems at the gemba level. The job of the engineer will become even more difficult, but it’s at these times especially that we need those horizontal threads to go between the vertical threads so that at the gemba, all aspects of production feel that they are pointing the same way.

Ohno put a great deal of emphasis on wording. In this book, I have used words like horizontal thread and vertical thread, but Ohno would use the technical terms. He would write out the characters (spell out the word) for people and explain them. Ohno was from Toyoda Loomworks, so he had a keen understanding of how to make cloth, but he was still a man who placed a high importance on words. He had people think about the wording, and he repeated those words often.

Let’s take a look at these words again: “Engineers in production become horizontal threads.” Why production? Well, all the departments and sections of the company come in as vertical threads in the production gemba.

Lesson 15: The Lowest Kanban Quantity Should Be Five

This was the answer given to a person who asked, “Is one the lowest kanban quantity?”

Sure, you could pull one, but why would you want to do such a useless thing?

When sharing Ohno’s thoughts, I must mention that the kanban quantity had to be at least five, or else risk Ohno’s ire. He was “Five lot Ohno.” I kept wondering how to talk about this concept. I wrote and rewrote the section, then postponed it to the final lesson in this chapter, and now I must deal with it. If I didn’t include this, I could not claim to have written a book about Ohno’s thinking, so I will just write down the facts and the words that I heard from Ohno directly. It will be up to the reader to try to get some meaning out of it. I hope it proves useful.

Ohno was quite adamant about the five-piece kanban. Five was the smallest lot size, and parts were kept in sections five at a time. When Ohno was in charge of the assembly plant, the welding plant, and the machining plant of the Honsha factory, the engines, frames, and transmissions were made in sets of five and the same car was made five in a row. After he became the head of the whole factory, he had the casting and forging departments think in terms of multiples of five as well and put in the kanban system to reflect that. That joined the raw materials to finished automobiles in this daisy chain of five units at a time. Ohno determined that this structure was the most efficient for the Honsha plant, and the machines had a changeover every five parts so that they could keep up with the kanban instructions. The saying was, “Get your changeovers down so that you can do one every five parts,” which was the goal for all equipment.

During one of Ohno’s training sessions where all the managers were present, I remember very clearly when one of them asked Ohno what the minimal kanban size ought to be. Ohno responded by saying, “I don’t think we would do a useless thing like picking one or two. You need at least five, which is why I determined that five is the minimal size.”

Just to add some more explanation to this, what it means is that if one took a part that had a kanban for one part on it, that was a useless thing. In fact, if there was a kanban that said, “Make one more,” there would be no real need for a kanban. As an aside, in the Honsha plant, we leveled out our production in multiples of five. Sometimes we would make 10 cars of the same type if a lot of them were being sold.

Digression

The five-piece kanban that Ohno put all his effort into making slowly spread until all parts of the factory, including part machining, forging, and casting, were finally connected by multiples of five. After assembly (in terms of information flow), everything was on the pull system, so that if one knew when the assembly plant would finish work, correct decisions could be made at the shop floor level as to how much more should be made and when to stop work. This was a very efficient factory management method, and at the time only the Honsha plant was putting kanban-pulled parts into the main assembly line.

At the same time, this five-piece kanban was a way to make production more efficient. The customer was ordering in multiples of one, so making another four (which we did not have orders for yet) seemed like overproduction, which was what Ohno hated the most. For many people, this seemed like a logical conflict.

If one goes to a Toyota plant today, each car that the plant is making is different, meaning that it is making cars in lot sizes of one. The ability and techniques to make one car at a time were developed after the age of the batch of five. Even at the time, if we had made the effort to get to a batch of one, I think we could have done it, but Ohno was absolute in his opposition to that idea. Ohno was not someone who would use lack of space or lack of equipment as a reason not to do something, so obviously he had a good reason for this. Let’s use our why-why analysis to see if we can understand what he was thinking.

Ohno used to say, “The profitable production is a forecasted production.” If we think this way, then the pull system or the replenishment system was a “forecast”; we don’t know when the part will be used next, but we will make it so that it is available. The kanban served as a governor on how far ahead we could forecast. The “forecast” that Ohno talked about was not a slippery slope or an open-ended thing, but something that had strict limitations placed on it by the kanban size.

The four extra cars made on the assembly line were also cars that had not yet been sold. They were cars that our sales department had to sell. I think this method dovetailed into the obvious function of sales, which is to “sell what is as yet unsold.” In this method, the one car sold became five cars sold and became “the profitable forecast production.” The four extra cars stopped becoming overproduction but turned into four profit-generating cars.

There was also no worry that these five cars would become long-term obsolete stock, as the only time we would make more would be when those five cars had been sold and we got the next order. Inventory was not growing on its own but was controlled. It’s a really well-thought-out system. This method was used for quite a long while, and the salespeople played their role and ended up selling the extra cars in a manner similar to what I described.

We no longer have a car assembly line in the Honsha plant, so this issue is now one of the past, but I think it is unwise for us to write off the logic of making five at a time too easily. Doing five at a time is easy to remember and makes it harder to make mistakes; when you have too much workload, your team leader helps out. It’s a very simple system.

The five at a time method accommodates both making the one for the customer and efficiency in production. It takes a complicated assembly process and makes it simple, self-contained, and clear. Methods like these are hard to come by, and I look forward to the next generation coming up with a new way to accomplish this.