Production & Lean
Bottleneck Analysis in Manufacturing
Every production line has one step that limits the output of the whole line. Improving anything else changes nothing. This guide shows how to find the bottleneck and what to do about it.
By Career Plus Academy — manufacturing career development for Production and Quality professionals in India.
What it means
A bottleneck is the process step with the least capacity — the longest effective cycle time or the worst availability. Because stations are linked, the line cannot produce faster than its bottleneck, whatever the other stations can do.
Why it matters in manufacturing
Engineers often improve the station in front of them and wonder why line output did not move. Only bottleneck improvement increases output. Finding the real constraint focuses time and money where it pays back.
How to analyse and relieve a bottleneck
- 1
Measure cycle times
Time every station over several cycles, including loading, unloading and walking — not just machine time.
- 2
Account for losses
A station with a good cycle time but frequent stops can still be the bottleneck. Use effective capacity, not nameplate speed.
- 3
Watch the WIP
Work-in-progress piles up before the bottleneck and starves after it. Walk the line and look.
- 4
Relieve the constraint
Reduce its cycle time (method, tooling, changeover), improve its uptime, or offload work to another station or shift.
- 5
Re-check the line
After improving, the bottleneck usually moves to the next station. Repeat the analysis.
- 6
Protect the bottleneck
Keep it running: buffer material before it, prioritise its maintenance, never starve it.
Practical shop-floor example
Illustrative learning example — not a client result
Example: welding cell limits the line
A fabrication line's stations run at 40–55 seconds, but the welding cell runs at 75 seconds with frequent tip-dressing stops. WIP piles up before it. The team adds tip-dressing during scheduled micro-stops and moves one weld to an underused fixture. The cell reaches 58 seconds and line output rises by about 20% — no new machine purchased.
Common mistakes
- Improving non-bottleneck stations and expecting more output.
- Using nameplate cycle time instead of measured, loss-adjusted capacity.
- Letting the bottleneck starve during material shortages or breaks.
- Stopping after one improvement — the bottleneck always moves.
Practical checklist
- Cycle times measured at every station
- Availability losses included in capacity
- WIP pattern checked on the floor
- Improvement aimed at the constraint
- Bottleneck protected with buffers and maintenance
- Analysis repeated after each change
Frequently asked questions
Can the bottleneck change over time?
Yes. Product mix, tool wear, absenteeism and seasonality all shift the constraint. Bottleneck analysis is a repeating exercise, not a one-time study.
Is the slowest machine always the bottleneck?
Not always. A fast machine with poor availability or long changeovers can be the real constraint. Measure effective capacity, not speed.
Want to check how well you know this topic?
Take the Skill AssessmentPreparing for interviews too?
Check Interview Guide & Kit Availability