“Have you ever seen a process like this before?” asked the company president.  This was 90 minutes into a call scheduled for 30, but he was fascinated by how we had a Causal Explanation in only a few days when others had stumbled for months.

“No, I haven’t,” I said.  “But it doesn’t matter.  In such cases, we discover which function is failing, not guessing at what’s wrong.  And there are only a few functions machines perform. We know how to find and characterize function.  David, Tobias and I have been doing this for a long time, and often work on machines and processes that are new to us, but no machine can perform a function that is not on our 7×3 diagram.”

Screenshot 2020-05-25 09.41.11This was a conference call, of course, as most reviews are these days. It was Thursday and I had finished the project, which started on Monday. The engineer I worked with was on one side, the director who hired me, the other.  The president and vice president were in another country. The engineer who was the team leader did most of the talking.  He was fun to work with, a good engineer, and excited about having figured this out so fast. I was mostly silent, wanting the executive team to see his passion, confidence, and enthusiasm.

Before contacting me, the Director had spoken to another company.

“How long do you think it will take to figure it out?” he asked them.

“Maybe a year.”

Of course it would.  The consultants would come in with a bunch of training, pass out belts and certifications as markers of success, and pile on a bunch of structure that flies in the face of rapid problem-solving. They would make a list of variables using some standard tools, test to see if they are stable, likely plot them based on time, not in a way that matters. They will look for common or special causes, thinking a factory operates on the same principles as a casino, then submit an invoice, and leave.  They might try to see the differences between two machines, because one has less scrap, but would get them nowhere without understanding and characterizing function.

These were good and clever engineers on this project, who had been working to figure this out for six months. If they were frustrated, perhaps the strategy was no good. It clearly isn’t the engineers!  With the same team, we figured this out in four days.  And I had not seen such a process before.

Scrap was well in excess of 50%.  The company was not able to keep up with demand, but not in trouble with the customer yet, as one other supplier was worse.

Before I arrived, which was a three-day drive from the southern tip of Florida, I had spoken to the team.  Two called from the safety of their garages, as the plant was closed for a week because of the virus.  I asked them why they were in the garage, and they said something about kids, dogs, lockdowns, noise and homeschooling.

They sent a fishbone diagram and a fault tree, which I refused to look at. (I looked after we were finished.  The answer was not there.)  I looked at a sketch of how this large and complex part is constructed from several components and then cooked and squeezed.  The problem was the inability to squeeze one of the components into the proper shape.  I asked if we could shape the part better if we left out the last several components, which would have impeded the effort to squeeze.

“The shape will be perfect.”

They made one, which is always the best sample size.  The shape was not perfect.  In fact, it was horrible. The team was not disappointed, but surprised! This was an insight into function, as it was not as they thought.

When I agreed to work with these good folks, I said I would only help if we were having fun.  As always, they think I’m crazy.  (I am, but for other reasons.)  We have a couple more guidelines.  One is, the job is never to solve the problem but to learn one thing every day about the physics of function and failed function. That leads to daily progress through a process of elimination and learning.  Learning is fun.

When the single part failed to form the shape as expected, we knew which function failed, and went about characterizing the energy source from the manufacturing process that was to needed to create it, and the storage component in the assembly that was needed to sustain that function during the balance of the manufacturing process. Once we converged on the important function, we had eliminated every function that didn’t matter.  Problem-solving not based on convergence is a fool’s game. Problem-solving machine behavior based on an understanding of function speeds the game.  The insight and understanding gained from such an approach provides quality product performance, reliability, and competitive advantage.

These good engineers found out how fun it is to solve tough problems fast, when you know full well the chase is on and the fox is cornered! No guessing, and no lists.

At the end of the conference call, I asked the team leader if he had fun.

“I sure did! I learned more about how to solve problems in the last few days than I have in all the other training I have ever had.

What fun!

‘Happy the man who could search out the causes of things.’  Virgil, Georgics

Picture1

John Allen

John.allen@tnsft.com

www.tnsft.com

+1 603-969-0563