Always baffling how knowledge that is obvious in one domain is ignored in another one.
For you this is obvious, but I had to stumble upon ToC, study it and draw all connections myself before using it in practice. Where I come from - biometrics, clinical and now business, always medical devices industry - nobody knows it!
Thank you. You did a very good job of explaining it. So bravo!
One thing experience taught me over the years is that there is always another bottleneck. Once you “fix” the current one, it moves somewhere else. You never really eliminate all the problems. You eliminate one constraint and then go find the next one.
Would you care to expand on why trying to keep to the bottleneck in a specific location might turn against the whole system?
If I increase the capacity everywhere else and keep the bottleneck running, I am still maxing out the throughout regardless of whether the constraint position is artificially or naturally stable.
I agree that the bottleneck is the slowest part of the system, and it doesn't really matter whether it occurred naturally or you intentionally put it there. It is still the bottleneck, and you are right that you can keep that bottleneck running and maximize throughput at that constraint.
Where I see the potential issue is when you increase capacity everywhere else specifically to keep the constraint in that location. You may end up carrying capacity that the system doesn't otherwise need.
The added capacity has a cost. You may be trading efficiency for the ability to keep the constraint where you want it. That may be a good trade, but it is still a trade.
The very definition of bottleneck as where flows slows down the most implies that there will always be one piece of the process where the definition applies.
For how I see things today, the game is to A) know where the constraint is and subordinate the rest and B) keep it where you want it.
For example, maybe you want the process to be constrained by the upstream customer demand while you are committed to NOT having it in production, meaning that you are willing to keep hiring to keep the constraint upwards.
On B, my experience has been a little different. I look at the whole system and try not to maximize the individual parts. That means I follow the constraint to the next bottleneck rather than choosing where I want it to be. If you are not careful, trying to keep the bottleneck in a preferred location can lead to a suboptimal solution for overall throughput.
My B would be: never starve the bottleneck. Keep enough of a buffer in front of it to protect it from normal variation. If the bottleneck sits idle, you will never get that time back. I had a boss in the early 1990s who would drill me every time we missed a cycle. He was brutal, but he wasn't wrong.
Nice piece Luca. Very good explanation of the bottleneck and constraint thinking.
That solves it I’d say!
Always baffling how knowledge that is obvious in one domain is ignored in another one.
For you this is obvious, but I had to stumble upon ToC, study it and draw all connections myself before using it in practice. Where I come from - biometrics, clinical and now business, always medical devices industry - nobody knows it!
Thank you. You did a very good job of explaining it. So bravo!
One thing experience taught me over the years is that there is always another bottleneck. Once you “fix” the current one, it moves somewhere else. You never really eliminate all the problems. You eliminate one constraint and then go find the next one.
Would you care to expand on why trying to keep to the bottleneck in a specific location might turn against the whole system?
If I increase the capacity everywhere else and keep the bottleneck running, I am still maxing out the throughout regardless of whether the constraint position is artificially or naturally stable.
What am I missing?
I agree that the bottleneck is the slowest part of the system, and it doesn't really matter whether it occurred naturally or you intentionally put it there. It is still the bottleneck, and you are right that you can keep that bottleneck running and maximize throughput at that constraint.
Where I see the potential issue is when you increase capacity everywhere else specifically to keep the constraint in that location. You may end up carrying capacity that the system doesn't otherwise need.
The added capacity has a cost. You may be trading efficiency for the ability to keep the constraint where you want it. That may be a good trade, but it is still a trade.
That is what I meant by suboptimal.
Understood, thanks for expanding!
Yeah, that is structural.
The very definition of bottleneck as where flows slows down the most implies that there will always be one piece of the process where the definition applies.
For how I see things today, the game is to A) know where the constraint is and subordinate the rest and B) keep it where you want it.
For example, maybe you want the process to be constrained by the upstream customer demand while you are committed to NOT having it in production, meaning that you are willing to keep hiring to keep the constraint upwards.
As per your experience, how does this sound?
You are spot on with A.
On B, my experience has been a little different. I look at the whole system and try not to maximize the individual parts. That means I follow the constraint to the next bottleneck rather than choosing where I want it to be. If you are not careful, trying to keep the bottleneck in a preferred location can lead to a suboptimal solution for overall throughput.
My B would be: never starve the bottleneck. Keep enough of a buffer in front of it to protect it from normal variation. If the bottleneck sits idle, you will never get that time back. I had a boss in the early 1990s who would drill me every time we missed a cycle. He was brutal, but he wasn't wrong.
I retired a little bit ago and used to run production factories so I guess you could say I grew up on it :)
Thanks!
Was this your first time coming across Theory of Constraint or you already knew the topic?