Every assembly line has a constraint. Sometimes it is obvious: a station with a visible queue and an operator working flat out. Sometimes it is less obvious because product mix, buffers, or operator movement hide the real limiting factor.
Start with queues
Work in progress usually tells a story. If units consistently build up before a station, that station may be the constraint. If a station is regularly starved, the constraint is probably upstream.
Do not look at one snapshot. Watch the pattern over time, especially across different product mixes.
Look at utilisation
A highly utilised operator or machine may be the bottleneck, especially if they have little idle time while others wait. But be careful: 100% utilisation everywhere is usually a warning sign, not a goal.
- High utilisation at the constraint is expected.
- High utilisation at non-constraints may create fragility.
- Low utilisation after a queue may mean the real issue is upstream flow.
Check product mix
In mixed-model assembly, the bottleneck may move. One product family may overload workstation two, while another overloads workstation four. A line that looks balanced on average can still fail under a realistic sequence.
Do not optimise the wrong thing
A common mistake is improving the easiest station to improve, not the constraint. If station one is at 60% utilisation and station three is limiting output, making station one faster will not increase throughput.
Use simulation before physical trials
A simple simulation can help compare likely bottlenecks before moving people or equipment. Try changing work allocation, adding buffers, changing operator assignments, and running different product sequences.
Find the constraint before changing the floor
Flowcell shows queues, lead times, operator utilisation, and throughput so you can compare scenarios before committing to a physical change.
Try the live demoRelated: buffers in assembly lines and assembly line simulator.