
What Is Bottleneck Analysis? Theory of Constraints
A bottleneck is the most constrained resource that determines the total output rate of a production flow. The Theory of Constraints says that a system’s performance is always limited by its weakest link and that improvement effort should focus on this link first. Speeding up other resources does not increase total output unless the bottleneck changes.
- Bottleneck: the most constrained resource that sets the pace of flow.
- Symptoms: WIP piling up in front of it, idle stations behind it, high utilisation.
- The Theory of Constraints proposes five steps: identify, exploit, subordinate, elevate, repeat.
- Time lost at the bottleneck cannot be recovered; at other resources it usually can.
- Key benefit: the improvement budget goes where it pays off most.
How Is the Bottleneck Found?
The most reliable method is to read the data from three angles together:
- WIP build-up: in front of which station is work always waiting?
- Resource utilisation: which machine is idle the least?
- Starvation: which stations stand still waiting for material?
The resource with a pile in front and starvation behind is the bottleneck. Looking only at utilisation is misleading; some machines look busy because of poor scheduling but are not the real constraint.
The Five Steps of the Theory of Constraints
- Identify the constraint. Which resource limits output?
- Exploit the constraint fully. Do not stop it during breaks, shorten setup times, catch quality defects before the constraint.
- Subordinate everything else to the constraint. Let other resources work to the constraint’s rhythm; overproduction creates piles.
- Elevate the constraint. Invest if necessary: a second machine, an extra shift, subcontracting.
- Repeat. When the constraint moves, go back to the start; do not let inertia become the constraint.
Practical Ways to Exploit the Bottleneck
| Method | Effect |
|---|---|
| Not stopping during breaks and shift handovers | A few hours of extra capacity per day |
| Shortening setup time (SMED) | Direct increase in output |
| Checking quality before the bottleneck | No constraint time is wasted on defective parts |
| Buffer stock in front of the bottleneck | The constraint does not stop waiting for material |
| Moving planned maintenance outside shifts | Less loss during production hours |
Common Misconceptions
- Trying to run every resource at 100%. Running non-bottleneck resources at full capacity only produces WIP.
- Assuming the bottleneck is permanent. When the product mix changes, the constraint moves.
- Looking only at machines. The constraint is sometimes a mould, a certified operator or a quality inspection step.
- Deciding without measuring. “I think that machine is the bottleneck” and the bottleneck found with data often turn out to be different.
- Improving the constraint and then stopping measurement. If the new constraint goes unnoticed, the gain is soon lost.
A Short Example
On a five-station line, let the station speeds be 120, 110, 80, 115 and 130 pieces/hour respectively. The line’s output is limited by the speed of the slowest station, the third one, that is, 80 pieces per hour. Raising the first station from 120 to 140 does not increase total output; it only grows the pile in front of the third station. In contrast, raising the third station’s speed to 90 by shortening its setup time directly lifts the whole line’s output by ten percent. This is the Theory of Constraints in practice: when improvement effort is not directed to the right place, the work put in does not turn into output.
Frequently Asked Questions
How is a bottleneck identified?
Does the bottleneck move?
Why is an hour at the bottleneck so valuable?
Bottleneck Analysis in EFFIX MES
EFFIX MES makes it possible to find the bottleneck with data instead of guesswork; the production monitoring module records machine states, stops and downtime reasons in real time, and cycle time screens show speeds per resource. When OEE and its components are compared resource by resource, constraint candidates become clear. On the planning side, capacity analysis evaluates machine, mould and human dimensions separately; with scenario simulation, the question “what happens if we give the constraint an extra shift” can be tested. WIP tracking shows in front of which station a pile is forming. For details, see the EFFIX MES page or request a demo.
