If you want to see the car factory of the future, you need to beg Toyota to let you into their Takaoka plant near the company’s headquarters in Toyota City, Japan. If they let you in, prepare for a revelation. Ever since the first cars rolled down Ford’s assembly line a little more than 100 years ago, a Gordian knot strangulated carmakers the world over: Assembly lines are fast, but inflexible. After pulling a few strings to get into the Takaoka plant, you will see the Gordian knot become untied.
Last November, I had cajoled Toyota into admitting the 12 world travelers of the Magical Mystery Plant Tour into the inner sanctum of car production efficiency. Today, I am here for a much closer look, and for a long talk with the plant’s grand master Osamu Akahane, Project General Manager of the Takaoka Plant. But before we do a deep dive, let’s ask the most important question: Why would you want a flexible car plant? Isn’t it enough that the damn thing spits out cars in ever increasing numbers?
Did we just hear “ever increasing numbers?”
Assembly lines are great at building cars in great numbers and at relatively high speed. At the same time, assembly lines abhor change. Assembly lines pretty much have two speeds: On, or off. They hate to go much faster, or much slower, than their rated speed. Try introducing a new car model to the assembly line of old, and you sometimes face months of retooling. When demand for the car increases, customers sometimes must wait months for the long-tailed assembly line beast to catch up. When demand slackens, plants often must be idled. Takaoka is a marvel of production engineering that solves all that, and then some.
But first, lets find out why assembly lines are so stubborn. For this exercise, let’s assume you have been gifted a car factory with room for two assembly lines. While learning the auto business, you used the space of one line for a mostly manual production of your first two high-end cars. Encouraged by their success, you develop a car for the masses. To make that car, you build a high-speed production line, good for 250,000 cars per year.
People love the car, and demand is off the scale. How do you double output to 500,000 cars a year?
“Easy,” you say, “just make that line run faster.” Let’s see what happens as you crank up the speed.
There is that guy who used to fasten ten bolts on every car. Now, the line runs twice as fast, and he can, even when pushed, only fasten 5 bolts before the car has moved out of reach. So what do we do?
“Hire more people.”
Great. Now we have two guys who fasten five bolts each. However, the guys take up twice the space. And so do all the other hundreds of guys we added to the line. What do we do?
“Make the line longer!” Good idea. For a line that runs twice as fast, we would end up with a line twice as long. But where do we put it? You realize that the factory that was given to you is surrounded by residential housing. Your parking lots already are flowing over with the cars of the new hires.
To make matters worse, our production line is dotted with fixed stations performing fixed tasks, like adding seats to the car, or dropping in engines, or dashboards. To allow for the new length of the line, those fixed stations would have to be ripped out of the ground, and moved. Along with that, the conveyor belts bringing the seats, engines, and dashboards, also would have to be moved. In the old style car factory that was given to us, car shells are hanging from the ceiling until they get mated with the suspension system. The station where the mating takes place would have to be ripped out and moved. And that’s just scratching the surface.
Your engineers are already imploring you to build a new 250,000 unit per year line elsewhere, which would be much easier and more efficient, they claim, but you are the boss, and you want more cars NOW.
“What about three shifts” you say? You will quickly find out that three shifts should be run only in a dire emergency, and only for a short while. With a third shift, your cost goes up as workers get 25% more, or sometimes time and a half. Your line breaks down often, because you don’t have the time for maintenance at night. And remember, we wanted to double our output. With a third shift, we’d only add 125,000 units for a theoretical total of 375,000 … until something breaks.
“Robots. Let’s buy lots of robots.” Before you run to ABB, or Kuka, and buy an army of bots, you first may want to talk to other automakers, like GM or Volkswagen, who had the same great idea. They would tell you that robots are great for repetitive tasks like welding, or painting, but they fail at picking and handling small parts. Robots take up more space than humans. And more often than not, robots slow down the line as opposed to speeding it up. Unbelievable, no?
Enough theory already. Let’s go to Toyota, and watch robots race against humans.
We are in Takaoka I, a fairly modern car plant put in in 2007. Factory chief Akahane-san pulls out a stopwatch, and we time a robot putting a spare tire into a Corolla. Between pulling the tire out of a rack, navigating between Corollas moving down the line, and dropping the tire into the Corolla, 57 seconds expire. “We make a new car every 60 seconds on that line, and the spare tire robot is one of the many reasons why the line can’t do it faster,” Akahane says.
Then we walk across the road to Takaoka II. The only thing this line has in common with Takaoka I is the name. This “simple and slim” factory is the antithesis of a gigantic Alien Dreadnought. Takaoka II is the Judo fighter of car plants: Quick on its feet, flexible, and smart.
Here, a worker pulls the spare tire with something that looks like a pushcart with long arms. The worker wheels the cart around, and he drops the tire into a Prius. Elapsed time: 17 seconds. I am looking for electric cables, or air hoses that power the lightning-fast spare tire machine. I don’t find any. “It’s all done with springs and counter-weights,” Akahane beams.
On top of it, the muscle-powered, spring-loaded, counter-weighted tire machine can travel up and down the line, to add spare tires where needed. Over at Takaoka I, the 57 second slow robot was bolted in place, and it stubbornly held up any attempts of adjusting the line.
“Alright,” you say, “you replaced a robot with a human, but how does this make the line flexible?”
The trick of Takaoka II is that there is no production line per se. Instead, the cars trundle along on foot-high motorized dollies that run between platforms on both sides of the line. As long as you have the floor space, you can extend the line, and ramp your production over the weekend. The longer line now has room for more people, and with the no longer fixed stations moved, the line can run faster, and make more cars. The ramp won’t follow an S-curve, production won’t go up exponentially. On Monday, output will jump vertically, straight up to where you want it.
If you don’t need the capacity, you shorten the line, re-arrange the movable stations, partition off the hall to save on air conditioning, and re-assign workers to other parts of Toyota. Should the line not be needed anymore at all someday, it can be put into containers, and to better use elsewhere.
The latter is unlikely, because where Takaoka II really shines is in combination with its older sibling at Takaoka I. The older brother is most efficient when running at 200,000 cars per year. If there is more demand for the Harrier and Corolla cars made on Takaoka I, some production can be shifted to the flexible line at Takaoka II. If demand drops, production is shifted back to Takaoka I to keep the line running at its optimal one car per minute speed. Takaoka II is much more forgiving as far as cost per vehicle goes. “Takaoka II can increase and decrease capacity, and production cost stays the same” says Akahane.
For other automakers, substantial increases or decreases in capacity often mean building new factories, or idling current ones. At Toyota, it can be done over the weekend. Workers lay down cable trays on a flat shop floor. Instead of a mess of cables, only one fat cable is connected. Where cables must cross the line, a gate shaped cable tray on casters is rolled into place. Platforms are set down left and right to guide motorized dollies for cars in nascent state. Formerly fixed stations are rolled in place down the line, and on Monday, the plant has a completely different capacity than what it was on Friday. In the course of a year, this magical capacity conversion can happen several times.
Robots working along the line are a bit less flexible. Over at Takaoka I, the job of gluing in a windshield was the job of two robots, housed in a 33ft long cage to protect workers from their rapid movements. In Takaoka II, one bot was fired, the whole process is handled by a solitary robot in a cage one third the size. The 10ft long robot station sits on one base, and come capacity change, the caged bot is moved down the line with a simple forklift.
Like all Toyota factories, Takaoka II is built on the on-going “kaizen” improvements developed by its workers who want to make their worker go faster, smoother, smarter. Asked what kind of people he’s looking for to work at Takaoka II, Akahane answers: “We want people who can think for themselves, not robots who do as they are told.” The plant can’t and doesn’t want to do totally without the bots. They mostly work in a separate hall, where metal is stamped and welded before the car bodies are fed to separate paint shops of Takaoka I and Takaoka II, and from there to assembly.
Instead of trying to max out every cubic meter of the hall, Takaoka II more or less ignores the 3rd dimension. Everything happens on one flat plane. There are no overhead gantries, and because nothing happens above, there are no height restrictions for the cars made on the shop floor. There is a lot of those two dimensions in the back of the giant, but simple hall Takaoka II occupies: Half of its space sits empty, breathing space for the flexible lines.
The super-flexible “Takaoka II could theoretically build any number of models on the same line,” tells me Akahane, “but it probably would stop making sense at six.” Currently, Takaoka I makes the Corolla and the Harrier, while Takaoka II builds the Prius V and the RAV4. A few months ago, the mix was different, and in a few months, the mix will be different again.
Takaoka II also is capable of miracles. In older plants, re-tooling a production line between model changes can sometimes mean months of downtime. When Volkswagen introduced its MQB modular kit, it was proud of having shaved the changeover times down to a few weeks. How long does Akahane have to stop the line when a new model is introduced?
He shapes thumb and index finger to a zero. “There is no stoppage of the line at all. We spend less than an hour exchanging parts and tools, but we do that while the line makes other models.” As production of the new model starts at the beginning of the line, the last of the older models exit at the other end while the line keeps moving.
What would happen if the line stops? Akahane’s hand slices across his throat. “My boss would cut my head off if the line would be down.”
And how much did this miracle line cost? “The CAPEX of Takaoka II was half of Takaoka I,” says a smiling Akahane.
The simple, slim and flexible concept was born during the carmageddon of 2008. Across the world, the longtailed assembly-line beasts started devouring their owners. Fewer cars or not, the beasts demanded to be fed a steady diet of money. GM and Chrysler went bankrupt. Factories were closed. Brands and dealers were culled. Smaller carmakers like Volvo, Saab, Jaguar, Mazda, were turned into sorely needed cash at fire sale prices. Frugally-run Toyota survived 2008 with its first loss ever, and they decided never to be caught again with unneeded capacity. By now, the simple and slim concept has been rolled out to Toyota factories in Guangzhou and Tianjin, China, and it will be in the soon to be opened Toyota plant in Guanajuato that “should be able to make nearly anything Toyota executives ask for.”
When we were theorizing about a gifted car plant, the example may have vaguely sounded like Toyota’s NUMMI plant, that was, for practical purposes, given to Tesla in 2010 in exchange for shares. A year later, Toyota opened its first simple, slim and flexible plant as a pilot north of Sendai, Japan, and a few years thereafter, Toyota sold its shares in Tesla. If the relationship would have lasted a little longer, and if Elon Musk’s hubris would have been a little less pronounced, Tesla could have learned something.
Before landing in production hell, and pitching a tent to make his Model 3, Musk promised that he would out-Toyota Toyota when it comes to lean manufacturing. He had a catchy name for Tesla’s non-existent miracle plant. He called it the Alien Dreadnought. I ask Akahane what Toyota calls its super-flexible line.
“We don’t really have a name for it,” Akahane says, and he doesn’t seem to think it needs one.