On Leveling (Heijunka)

Translation. Odier Tadashi

Today I would like to talk about one of the foundations of the Toyota Production System – "heijunka" – which means leveled production. I will discuss how Toyota is placing new emphasis on this "back to basics" element.

I will share some recent developments at Toyota that make heijunka (leveling), and our entire system, even more flexible and responsive to customers' wishes.

Leveled production is something that can deliver impressive benefits to any manufacturer, and I hope my remarks will be relevant to your businesses as well.

Before I start talking about heijunka, I would like to give you an update on Toyota's overall growth in North America and our plans for the coming years.

In North America, we now have ten manufacturing facilities making everything from parts and components to engines and transmissions to complete vehicles. In addition, we have three additional facilities being built right now: a truck plant in San Antonio, Texas, a pickup and truck bed plant in Baja California, Mexico, and a foundry for aluminum engine blocks in Jackson, Tennessee.

In 2003 we built approximately 1.3 million vehicles here in North America, and by around 2006 we will have the capacity to build 1.66 million.

At the moment, we build ten models here, including the Camry, Corolla, Tundra, Sienna and the Lexus RX 330.

We have also been fortunate that our sales continue to grow. In 2003, we passed the mark of two million vehicles sold in one year in North America for the first time in history.

And, most importantly, we continue to increase the number of American Toyota employees. We currently have more than 35,000 people in North America working in Manufacturing, Sales and R&D (Research & Development).

And that does not include our dealers and the people employed by our suppliers to work on our business.

To put all this in perspective, twenty years ago we did not build any vehicles in North America. Today, more than 60 percent of the vehicles sold here are built here!

Clearly, this rapid growth has been a challenge for leveled production.

Let me begin by giving you a brief overview of heijunka and how it fits into the Toyota Production System.

The simplest definition of heijunka is leveled production.

Here is an example that illustrates heijunka. Think about the last time you went to a football game. Before the opening whistle, people make their predictions, look for their seats and get ready to watch the game. But after the final whistle, everyone stands up and heads for the exits.

And what happens? A massive funnel effect and a huge bottleneck.

This shows the need for leveling (heijunka).

Like many aspects of the Toyota Production System, heijunka arose out of necessity.

In the early 1950s, Toyota needed to increase its production of trucks for the U.S. to use in the Korean War. Unfortunately, however, there was a shortage of everything, from raw materials to parts.

We could not get enough parts in the quantity or at the time needed. In fact, we literally could not do any production during the first half of the month!

Those two weeks were spent gathering the parts, which arrived in no particular order. Then we spent the second half of the month assembling the trucks.

We realized that this system could never work if production levels increased.

So, to meet this challenge, the heijunka system was developed.

Heijunka is one of the main foundations of TPS, along with standardized work and kaizen, or continuous improvement.

If production were exactly in line with customer orders, production on some days would be very high and on some days very low.

But you would always need to have machines, manpower and material to handle peak production. Of course, there would be significant idle time during the valleys. We call this waste "muda" (waste).

Friends, muda is definitely a dirty word at Toyota.

So the solution is to take all the orders during a certain period – a month, or a week, or a day, and put them into a "pool" (order bank) and then level production.

Although it does not make production a straight line, the peaks and valleys have less variation and are more predictable. This is heijunka.

To give an example, let us say that during one week we receive 700 orders for Camrys and 300 orders for Avalons. Now, of course, orders do not arrive at 100 Camrys a day. Some days it might be 175, other days 65. But we gather all the Camry and Avalon orders and then build them in a leveled way.

So the production line might look like this: Camry, Camry, Avalon. Camry, Camry, Avalon. And so on.

This leveling is used not only for the vehicles themselves, but also for various other options, such as engine size, sunroof, colors and much more. Each of these items is leveled in the "pool".

Heijunka requires the company to rethink how it should buy from suppliers, how to design machines and tools, how to develop work processes and how to plan its teams.

Heijunka drives the creation of flexible processes.

But without heijunka, a company ends up with surplus resources, idle time, high inventories, high costs and much more.

People often ask about the relationship between heijunka, just-in-time and quality. Well, heijunka is an idea for achieving just-in-time production. And, without building in quality, heijunka would be impossible. Defects, of course, make our production fluctuate.

So coordination with our suppliers is essential. We probably have some of our supplier partners here in the auditorium today, and they will be the first to tell you that we are far from perfect in our heijunka.

But we are continuously working to improve our heijunka, and we have made great achievements in recent years in process improvement. Today I will share a little about these recent initiatives.

The first subject I would like to address is our new "Global Body Line" implementation, or GBL.

The GBL gives us more flexibility to adapt quickly to customer demands, by mixing several different vehicles on the same line.

First, I will explain the concept. Our old system was called the "Flexible Body Line". Similar to other automotive companies, our old system was structured from the outside in.

The best way to explain this process is to imagine a piece of paper being folded up on all four sides to make a box. Although very flexible and advanced for its time, this system was very large and expensive to maintain.

One of the aspects of this system that we wanted to improve was the use of pallets. For those who may not be familiar, pallets hold the vehicle's body structure together while the parts are welded.

On average, our system needed 50 pallets for each single type of model built on that line. The system was also not adaptable to low volume or manual welding, such as we have in countries like Vietnam.

The new body welding process is more in line with Toyota's production concept – flexible, smaller and simpler to operate. It accommodates manual welding for low-volume production and automatic welding for high volumes.

But here is the unique part of this process – the body framing station can now build from the inside out. This completely new concept makes it possible to have just one master pallet instead of 50 pallets.

Now access to the body by the welding robots is easier, improving overall accuracy while reducing the number of welding stations needed for body assembly.

The end result is improved body quality, smaller robotic welding lines, lower capital investment and shorter lead time for launching new vehicles.

It is now possible to produce up to eight different vehicles on the same line through a simple change of a master pallet.

The benefits are nothing short of impressive.

What is our initial investment when we install a new body welding line? Fifty percent less.
Space savings? Fifty percent with the flexible line concept.
Maintenance cost? Fifty percent less.
Cost of adding another car type to the line? Seventy percent less.

We are currently rolling out versions of the Global Body Line at our manufacturing facilities here in North America and increasing the number of plants around the world.

Now I would like to give an example of how this new body welding process has been changing our production in North America.

In 2003, we moved production of the Solara Coupe and Convertible from our plant in Cambridge, Ontario to Georgetown, Kentucky.

In the case of the Convertible, we not only changed the production location but also changed the entire manufacturing process.

Previously, the Solara Convertible was built by creating the same body structure as the Solara Coupe. Then the roof was cut off, the convertible top was put on and the manufacturing process continued. Along the way, we needed to add reinforcements to maintain the convertible's rigidity.

And the biggest question for North American production was: can it be profitable to produce a vehicle that sells around 10,000 units, when we normally expect to sell around 200,000 vehicles a year?

Well, with GBL, this would not only be possible but also an improvement to the entire process.

The all-new Solara Convertible was launched right here in Chicago a few weeks ago, and it was designed from the floor up as a convertible.

There is no cutting off of the roof and no adding of reinforcements. And it is a more efficient way to produce a vehicle.

So how did we do it?

First, the decision was made that the convertible would have a different body assembly line from the coupe.

Then an entire section of the plant was built in less than six weeks!

As I mentioned earlier, our new body assembly line needs fifty percent less space. So we were able to use the space saved to build a line dedicated to the convertible.

The reason for creating a separate line was quite simple – the Convertible would have much more manual welding than any other vehicle produced at the Kentucky plant.

As mentioned earlier, one of the reasons for creating the Global Body Line was to support welding processes both in plants with high levels of manual welding – such as in Vietnam – and in plants with high automated volume, such as Georgetown.

For example, we produce five different vehicles in Vietnam – including the Camry, Corolla and Land Cruiser. But the total annual volume is approximately seven thousand vehicles. Compare this with our Kentucky plant, which produced more than 359,000 Camrys alone last year.

But we learned that the new system can also be used to standardize these processes within the same plant!

The Solara Convertible may prove to be a glimpse of the future of Toyota manufacturing – when we will be able to produce smaller volumes of vehicles to better adapt to what customers want.

Here is another example of how we are improving our operations in support of heijunka.

When our plant in Princeton, Indiana, began expansion plans with the installation of a second line, they wanted to have enough flexibility to be able to "crossover" the vehicle, producing it on both lines. This would allow them to be more flexible and responsive to customers' needs.

However, there is a clear and obvious problem.

The new vehicle coming from the Indiana plant – the Sienna – is a car. The two vehicles already being produced there – the Sequoia and the Tundra – are pickups.

For those who are not part of the automotive industry, producing cars and pickups on the same line is very challenging and is almost never done in high-volume production. The assembly process for these two types of vehicles is very different in several key respects.

Well, we decided that the Sequoia would do the crossover between these two plants.

Now, if we operated in an environment that said "do not push the limits", this idea would never have come up! People would think we were crazy.

Since I have only a little time to spend here with you today, I will not be able to go into depth on how we achieved this. But let me talk about the body/frame marriage. This is the part of the process where the engine and steering are attached to the vehicle – and this process is quite different between cars and trucks.

Our engineers worked hard to design something that would work for both the Sienna and the Sequoia, with minimal capital investment. And they succeeded.

So what do we do with this exciting new method of building vehicles? Nothing!

Right after the new line opened, we realized that demand for the restyled Sienna was far exceeding our expectations. So we ended up producing the two vehicles together on the same line for a few months. Now we only build Siennas there.

But we are ready to change our production mix as quickly as necessary.

We do not dictate what flexibility is, because the customer does! Our business is really changing rapidly. Customers want more niche vehicles, greater variety, and they want everything even faster.

Does that sound familiar to your business?

In a hypothetically perfect situation, everything would happen like this: customers would walk into a Toyota dealer and order the exact, specific vehicle they wanted. And they would have it immediately.

This would be a huge challenge for heijunka!

Well, we are still far from achieving this, but we have made tremendous progress in the last year, which I would like to share with you.

Six years ago, we launched an initiative called "NED" – or New Era Distribution.

The NED vision was simple: "Right Car, Right Place, Right Time 100 percent of the time".

Let me start by explaining what NED is not:

NED is not build-to-order for the customer
NED is not a "5-day car"
NED is not a way for the customer to get any kind of combination of options on the vehicle. We are very sorry, but you simply cannot have a V8 in a Corolla!

Now that I have pointed out the downside of the method, let me talk about the upside, because there is a lot.

First of all, the concept of build-to-order vehicles ("build-to-order") goes completely against heijunka, because it would disrupt leveled production.

If the total demand  for, say, green Camrys represents 20 percent of our orders, we want to produce 20 percent per month, 20 percent per day and even 20 percent per hour.

An unexpected increase in the popularity of the green Camry could really challenge heijunka, and the entire downstream process in the supply chain.

However, we in production realized that we could serve our customers better.

Now, in the Toyota Production System, the customer is the next person or group in the downstream process. So our customer is not someone who walks into a Toyota dealership.

Without a doubt, our customer is the sales organization. So could we help our fellow sales people be more responsive to their customers? The answer is yes.

What was needed was to adjust our entire supply system to allow a certain amount of fluctuation in our production.

I mentioned earlier that heijunka is achieved by putting orders into a "pool" and then leveling production. Well, to achieve the goals of this initiative, we needed to reduce the "pool". Instead of gathering orders from, say, a month – we are trying to reduce this order bank for certain options to less than a week.

So if sales wants more green Camrys and fewer white Camrys, and this fits within the permissible range, we can change the production mix up to six days before the Camry is produced.

There is another element of this system that I will explain briefly, because it does not really focus on heijunka. Even before we consider changing our production plans, our sales associates are able, in an online system, to place an order for a "virtual" transaction and change the options on vehicles that have not yet been sold.

This project is still being developed and is available only on our vehicles and options built in North America, such as color, sunroof and sound system, but the recent results are encouraging.

So far, we have been able to fill 85 percent of customers' change requests, because they fall within our heijunka limits.

So, noticing that there was a way to serve our customers better, we made the entire heijunka system more flexible.

Yes, heijunka is challenging, but it is also rewarding. And transferable to any kind of business.

I would like to conclude my speech by talking about how heijunka helped transform the Herman Miller company into an even more efficient manufacturer.

Herman Miller, as you know, is the second-largest office furniture company in this country, and has been in business for 80 years.

Toyota has a division that works to help companies implement versions of the Toyota Production System in their respective businesses.

When Herman Miller approached Toyota, they were facing the same challenges as most office furniture manufacturers.

They were struggling with long lead times, high inventory and delivery commitments with no credibility. Most products are built specifically for each customer order, and with so many products, each with its own possible features and finish options, this generates millions of product combinations.

Now, I will explain. The heijunka concept was just one part of the Toyota Production System that Herman Miller adopted as the basis for reducing costs while improving quality and delivery time.

By changing from a batch manufacturer to leveled production, Herman Miller was able to:

Improve efficiency by balancing a variety of models on the same line...

Reduce the shop floor space used by decreasing inventory...

And improve reliability by eliminating batch orders.

A good example of their improvement comes from panels. Herman Miller used to produce up to two weeks' worth of 60-inch colored panels, store them and eventually use them as orders came in.

Now they produce only panels in the sizes customers order, as they order them.

It used to take the company three days and three miles of assembly line to produce a panel. Now it takes less than eight hours and approximately 200 feet of line.

I do not think I need to tell you what a big difference this has made!

I hope I have given you a good overview of the importance of heijunka for all manufacturing operations.

I have been with Toyota for almost 34 years, and this is still a process I am learning every day!
Leveled production saves a company time, inventory, space and money. Leveling uses both manpower and machines more efficiently.

Where is the road of manufacturing heading?

If anyone gives an answer with conviction, take that tip with a grain of salt!

But those who are flexible are not bound by pre-established maps; they simply follow their imagination and the wishes of their customers.