Getting more from what you have.
WHAT TO MEASURE
Output against capacity Downtime by cause Yield Defect and rework rate Cost per unit
WHY THOSE FIVE
Together they explain performance entirely.
WHAT TO IMPROVE FIRST
The bottleneck.
WHY
Improvement elsewhere produces nothing.
WHAT TO ATTACK IN ORDER
Downtime at the bottleneck Defects consuming bottleneck time Setup time at the bottleneck Everything else
WHAT DOWNTIME ANALYSIS REQUIRES
Recording every stoppage and its cause.
WHY RECORDING
Impressions of why the line stops are consistently wrong.
WHAT THE RECORD USUALLY REVEALS
A small number of causes producing most downtime.
WHAT TO DO
Address those.
WHAT TO INVOLVE
The operators.
WHY
They see the problems and they know the workarounds.
WHAT TO ESTABLISH
A route for them to raise improvements, and a response to every one.
WHAT SMALL CONTINUOUS CHANGES ACHIEVE
More than occasional large projects.
WHAT TO AVOID
Improvement that depends on people being more careful Changes made without measurement
WHAT TO DO
Measure before, change, measure after.
WHAT TO STANDARDISE
Whatever worked.
WHAT TO REVIEW
Whether improvements held, months later.
WHAT TO PROTECT
Time for improvement work.
WHY
It is displaced by production otherwise.
