Improving a stage that is not the constraint changes nothing. This measures waiting time at every stage so the real bottleneck is a fact rather than an opinion.
What's Included
- Time at each stage separated into working and waiting.
- Queue length per stage which is what reveals the constraint.
- The constraint identified rather than argued about.
- An assistant that says what would improve throughput most.
How To Use
- Separate working time from waiting time. Waiting is where the answer is.
- Measure queue length, not just duration. Queues point at the constraint.
- Improve only the constraint. Everything else changes nothing.
- Re-measure after any change. The constraint moves.
Feed it from process cycle time, and act with continuous improvement.