This column is near and dear to my heart. It contains two rules, we all know and likely are paying the price for not following. 
First, it reminds me of one of my favorite encounters with my Toyota mentors. I will share their words of wisdom.

Second, the wounds are still fresh from yet again, learning the hard way. Read on and I will hopefully help you avoid this pitfall. 

Stop and fix
How many lean books have you read? If it’s more than one, you have been exposed to tried and true manufacturing principles. One simple question, did you listen? Or did you come up with 5,000 excuses why it won’t work for you?

Among the hundreds of principles that build the systems like the Toyota production system or lean manufacturing in general, there are two that if you start doing today, the results will be immediate, and staggering.

The first one is: “Stop the line to fix your problems.” Or as we commonly call it at my factory “Stop and fix.”

When an issue arises, our knee-jerk reaction is to dream up a solution that’s a workaround to make the product and be able to ship it. We woodworkers are expert fire fighters and because we’re all broke, we are super creative! So, this approach gets us by. However, the right thing to do, is “Stop and fix.”

Stop making product, get everyone involved in root causing that error, make the changes in engineering (I’m picking on engineers, because that’s where 90% of all the mistakes happen), re-issue the drawings or cut files, re-cut the parts, test if the fix worked and start the line again.

What if you can’t?
So, why don’t we do this? In a nutshell, it’s probably because everyone is too busy. Also, what if the fix involves a new custom router bit, and there is a two-week lead time? Then what? Leave the factory shut down for two weeks? I don’t think so.

This exact scenario happened at my factory, where the “Stop and fix” included a two-week lead time on some tooling. As per usual, we created a work around and suffered a bit until the new tools arrived. Then I found myself having lunch with Mr. Amazawa and Mr. Umimura both Japanese mentors and TPS experts. I asked, “How can we stop the line for two weeks, surely this isn’t what Toyota would do.” As usual, their answer was both simple yet profound. 

They just asked me, “Why is the tool two weeks away? Did you call the vendor?” And what they meant was did you call them like your life depended on it? What would have happened if you offered 10 times what the tool was worth? That’s still much cheaper than shutting down the line, and much more motivating for the vendor.
  
Did you call every vendor in the country? Again, no. Basically, after 15 minutes I realized that we just didn’t try hard enough. I feel like in North America our “try hard” is radically different from the Japanese “try hard.” They told me at Toyota, if a crane would break, they would deploy three different solutions all at the same time, knowing one would be faster, and one wouldn’t work, they stack the odds in their favor. Seems crazy until you calculate the cost of defects in your factory. Only then do you realize how cheap it is to “Stop and fix”. 

Subordinate to bottleneck
The second rule you absolutely need to follow is: Subordinate all activity to the bottleneck. I cannot overstate this one. We tend not to do it probably because we aren’t sure where the bottleneck is, and/or its counterintuitive enough to be confusing. 

No one likes slowing down any operation in their factory, particularly on purpose. However, it’s the #1 thing you can do to speed up your operations. Let me give you an example of how it manifested itself in my company. For anyone who doesn’t know, we manufacture MDF cabinet doors, under the trade name “My Door Factory.” We make a unique two-piece MDF door, and there are two categories based on time to manufacture. Let’s call them A & B doors. For this example we will simplify the factory into three areas: CNC department, Sanding, and Assembly.

We know our CNC department is the bottleneck. On average let’s say it takes 6 minutes to machine any door. We have three CNCs, so we get a door coming down the line about every two minutes. 

Sanding is the wild card. We know sanding sucks, so we make sure we do all of it for our customers before we ship, so they can focus on making cabinets. (No, not 95% sanded; 100% ready to finish.) This takes some effort. A-type doors take 1 minute to sand; B-type doors take 4 minutes to sand. 

Assembly is about 1 minute per door, and there are four stations involved. 

Confusing, I know, but I hope you got the picture, because here is where it gets complicated. We hadn’t subordinated any activity to the CNCs. We just had everyone “working hard, and doing their best,” which is a bad plan. 

On days with a ton of A doors, the whole place felt productive, doors were flying down the line at 1 minute each. But the CNC department was totally stressed because they couldn’t keep up. Eventually we cleared the buffer after CNCs, and the pressure of six people standing there doing nothing is all on them, and their hands are tied. Not good. 

Worst of all, at the end of the day, our door count wasn’t higher, because you will never get more than your bottleneck can produce. 

Slow down to speed up
So, what did we do? Slow down to speed up! We have no choice but to sand the profiles by hand, and we always felt the quality of the panels was never good enough right from the mill. So, we put a sanding robot in the cell to sand panels while simultaneously the operator could sand the profiles. And the best part, we made sure the robot takes 2 minutes per door.  

That balanced the sanding time exactly with the CNC. The robot paces the sanding bench; they couldn’t go faster if they wanted to. By default that also paced assembly. 

What happened, you ask?

• We immediately realized we had double the people we need in assembly.
• The CNC machines could comfortably keep up, no stress for operators.
• The start and stop of the assembly line vanished, slow and steady all day now.
• We were able to add sanding of the stiles and rails in the same time (improved quality).
• Everyone was working way slower.
• As a measure of productivity, we track labor minutes per door, and on day one, we saw a 30% reduction.
• Stress down, everyone working slower, quality through the roof and 30% more productive. That’s the power of subordinating to your bottleneck.

If you need help figuring this out for yoru factory, you’re in luck. Thats what we do. Text the word bottleneck to (226) 971-2144 and we will take it from there.