Thread by @DellAnnaLuca

THE TOYOTA PRODUCTION SYSTEM (TPS)

In this thread, I will explain the most relevant practices that are part of the TPS, with a huge focus on their problem solving approach and on their incentive systems.

[1/N]

2/ First, a disclaimer. I am an automotive engineer, I worked for years in automotive manufacturing and visited tens of plants, but I've never set foot in a Toyota plant myself. All I know about TPS is second-hand. So this thread might not be exact, but I trust it will be useful.

3/ Toyota is famous for its Just In Time (JIT) approach.

Most management schools define it as: ensuring that the supplies arrive only at the moment they are needed, so that they can be directly delivered to the plant, without unnecessary transit in a warehouse.

IT IS INCOMPLETE

4/ Just-In-Time is about removing buffers (the warehouse) so that problems can surface (in the supply line) so that they can be solved before a black swan hits.

As we will see, everything in the TPS is about facilitating problem solving.

5/ Some consultants would say that buffers are "inefficient" or a "waste", but that's missing the point.

Buffers are great when the unexpected happens. They are competitive advantages.

The problem with buffers is that they cover problems, preventing them from being addressed.

6/ Removing buffers because they are inefficient is a solution that works till, suddenly, doesn't.

Removing buffers to surface problems so that they HAVE TO be addressed is instead a solution that works.

One is hit by the bus he doesn't see coming. Buffers hide buses.

7/ Toyota used to have a red rope hanging over its production lines. Any worker (any!) was supposed to pull on the red rope when he noticed a problem. That would stop production.

What a way to ensure problems are solved immediately, rather than being lost in a suggestion form!

8/ The previous example highlights how Toyota internalized (decentralized) problem solving as one of its core values, and how it demonstrated it by allowing its blue collars to spend millions (in production lost when the red rope is pulled & the lines stop) to get problems solved

9/ (Before I continue the thread, let me repeat the disclaimer. I have years of experience in automotive manufacturing but I've never set foot in Toyota plants myself. Everything I know about them is 2nd hand. This thread might not be exact, but for sure it will be useful.)

10/ Toyota is known also for applying some form of Six Sigma. 6σ is a decentralized problem solving techniques, whose goal is to give most blue-collar workers a set of heuristics & tools so that they can solve problems & improve the value chain themselves.

11/ I am very critical of the overall validity of 6σ when applied to any field which is not Mediocristan (as defined by Taleb), but when used in, say, manufacturing lines, it can be very useful.

12/ Giving tools to workers to solve problems is much better than giving them suggestion forms to mail into boxes because:
- It gets problems solved faster
- By the workers who know the machines best
- With them retaining ownership & motivation

13/ Every manager has, in addiction to his own department-based targets, a set of problem-solving targets, on both
- himself
- his direct reports

If his direct reports don't improve their problem-solving skills & apply them, the manager cannot get promoted, no matter his results

14/ This is another great example of aligning (costly) incentives to values.

As an example of that, let me tell you the story of Mr. DuPont (from the DuPont company). He flew the guillottine in France to open a gunpowder company in Delaware.

(continues below)

15/ The problem of gunpowder factories is that, well, they tend to explode.

Mr. DuPont moved in with his family, in the middle of the plant.

And he ruled that all new machines have to be operated by management for the first hour.

Skin in the game works.

No more explosions.

16/ Which bring us to the next point of the Toyota way.

Problems have to be solved on the floor. Where they happen. By people who know them.

This is called Genchi Genbutsu. To truly understand a situation one needs to go to genba (現場), the "real place" - where work is done.

17/ Managers are required to regularly spend times on the lines.

I've seen this when I consult manufacturing companies. The simple act of forcing the directors & CEO to regularly walk on the factory floor & talk with workers asking the right Qs does wonders for their operations.

18/ There are many other elements of the Toyota Production System, but those I listed above are the fundamental ones.

I will not list more, because if the fundamental ones are not taken care of, no amount of "additional practices" will work.

However, a few more words on values.

19/ Your company can have values but, unless it demonstrates them with costly tradeoffs, they will not be internalized by your workers

You want your workers to behave rightly when things go wrong, when they are rushed, fatigued and the boss is not there. Not when things go right

20/ For company values to be internalized, 2 conditions are necessary:

- Your company must take *costly*, *visible* tradeoffs.

- Values are standards; standards have no grey area (luca-dellanna.com/grey-area-comp…). They must be *crystal clear* and *obsessively* hold people accountable.

21/ This ends the thread, for now.

If you have questions on TPS or on operations management, please ask!

22/ The reason why I'm not talking about other TPS practices is because the above are fundamental, and adding a few more practices without getting the fundamental rights will produce nothing good.

Focus on the fundamentals, then focus on them a bit more.

23/ A few more words on Toyota’s problem solving.

They do have systems to reduce variation in their processes, but this does not make them fragile (as it generally happens).

Why?

24/ Toyota reduces variation in their operations without becoming fragile because:
- it reduces volatility without using shortcuts (ie real reduction in volatility, not only a skewedness)
- it still keeps redundancy & flexibility (Toyota’s new plant is not bolted to the ground)

25/ Doing *both* of the above is a rare example in which conditions are created so that a removal in volatility is true, and not just a postponement of it (eg Fence Paradox, etc).

(Of course, not true for all contexts, but true for most of those in manufacturing)