Throughput and Cycle Time
Two measures that answer different questions and are routinely confused. Which to use for capacity, which for responsiveness.
Reference
9 notes
Where Work Waits
Queueing usually exceeds handling by a wide margin and is invisible without timestamps. Finding it, and the fixes that follow.
Procedure
Travel and Movement
The largest controllable loss in most picking operations. Measuring it without positioning hardware, and the slotting changes that follow.
Analysis
Equipment Utilisation
Fleets are usually larger than the average suggests and short at particular hours. Measuring availability against demand rather than in aggregate.
Procedure
Exceptions and Rework
Errors are found where they surface and caused somewhere else. Tracing origin rather than detection is what makes the data actionable.
Analysis
Reading Variation
Most performance differences are ordinary variation. Reacting to noise as though it were signal produces churn and worse outcomes.
Analysis
Finding the Constraint
One stage limits the whole flow. Identifying it from queue data, and why improvements anywhere else produce nothing.
Procedure
Labour Standards, Done Properly
Expected times for tasks are legitimate planning tools and dangerous targets. What separates a standard used for staffing from a quota used on people.
Analysis
Comparing Shifts and Sites
Raw comparisons are almost always unfair. What to normalise for before concluding that one shift is better than another.
Procedure