What’s a bottleneck? Every production system consists in one way or another of a series of discrete steps, right? We noted earlier that the rhythm of the production process is The Drum in DBR. Exceed the capacity of one step in the process, and you have a bottleneck. That’s not bad, it’s just natural. But that bottleneck thereby defines the capacity of the system as a whole.
Each step has an input and an output component. Looked at that way, it’s usually pretty easy to spot the bottleneck. Is one step out-producing the ability of another step to absorb its output? There’s your bottleneck. Moreover, overall production only increases if you can increase the capacity of your bottleneck. Until then, all other processes are subservient to the bottleneck. It’s the old weak link in the chain concept.
A guy named Pete Abilla has summed up nicely the leading thoughts on managing bottlenecks here.
His thoughts come from multiple sources, but summed up, he points out five principles, which I quote directly (because someone has already invented this wheel, and as they say, a hack borrows, an artist steals…), as a perfect synopsis of how to treat your bottleneck. Knowledgable readers will recognize these as basically The Five Focusing Points of TOC theory…
- Bottlenecks should never be idle; to lose time on a bottleneck, is to lose throughput.
- Never let a bottleneck run out of work. It’s okay to build inventory in front of a bottleneck.
- Increase productivity rates (offline and online processes) by reducing down-time, change-over time, and off-task time.
- Reduce defects by having Quality Assurance and Quality Control in front of a bottleneck, not after.
- Focus all improvements on the bottleneck.
The key point: identify your constraint(s), then manage it (them). Every bottleneck solved creates a new bottleneck somewhere else. Hence, the notion of continuous improvement. But let’s not get ahead of ourselves.
Instead, having now looked at the Drum, let’s look next at the Buffer…