Key takeaways
Short answer: A bottleneck is the physical resource with the least capacity versus demand — the slowest step. A constraint is anything limiting the system, which may be a policy, a market, or a supplier, not just a machine. Every bottleneck is a constraint, but not every constraint is a bottleneck. Improving a non-constraint adds cost and zero throughput. See also oee for manufacturing.
The Theory of Constraints says you improve the constraint and subordinate everything else. If your real constraint is a scheduling policy, buying a faster machine changes nothing. Identify the true limiter first.
Look for the queue (bottleneck) but also question policies and external limits. Sometimes the floor is fine and the constraint is order release or a single-source supplier.
OEE on the bottleneck is the OEE that matters most — an hour lost there is an hour lost for the whole plant. OEE on a non-constraint is largely free time.
See how Fabrico captures this automatically on your lines — explore OEE for manufacturing or book a demo.
Overlapping — every bottleneck is a constraint, not vice versa.
On the bottleneck — its losses cost the whole plant.
Yes — policy, market, or supplier constraints are common.
You add cost and no throughput.
Connexion
Blog
