Measure the process, not the person
In most facilities the time a unit spends being worked on is a minority of the time it spends in the building. Fifty notes on finding where work waits, how far it travels and where errors originate — and on the line between measuring a process and watching people.
Foundations
6 notesThree different things share the name: measuring a process, tracking equipment, and watching people. They need different justifications, and most of the value is in the first two.
What to measure
9 notesOutput per person is easy to compute and explains little. The measures that explain performance are about waiting, travel and variation — and all of them are aggregate.
Capturing the data
7 notesMost facilities already generate more than they use. What each source sees, what it misses, and the data quality defects that produce confident wrong answers.
Monitoring people
8 notesAttributing measurement to a named person is workplace surveillance with obligations attached, and it costs the cooperation that determines whether any of the data is accurate.
Safety
5 notesThe clearest justification in the field, and the place where monitoring is most often substituted for the physical control that should have been there.
Making sense of it
5 notesMost of the errors in this field are analytical rather than technical. Reacting to ordinary variation, comparing without normalising, and producing findings nobody acts on.
Running a programme
6 notesNinety days without a purchase, then buying for the gaps you named. Plus the failures that recur, each visible early and each with a specific correction.
Reference
4 notesSafety, data protection and employment law all apply and pull in different directions. Plus the boundary: questions the data cannot answer, and what to use for each instead.
What this is
Fifty notes on measuring what happens in a warehouse or a production facility, written for the people who have to run one rather than for the people selling sensors.
No vendor material, no product recommendations, no sponsored content.
Four things that hold across facilities
Handling is a minority of elapsed time. Waiting, travel and rework usually dominate, and none of them is visible without timestamps.
Output per person explains almost nothing. A slow figure in a badly slotted aisle is a measurement of the slotting.
Almost every operational question is aggregate. Where the bottleneck is, why the afternoon is slower, whether the new layout helped — none of them requires attributing anything to a named individual.
Rate pressure changes how people work, and the changes increase injury risk. If meeting a target requires abandoning the safe method, the target is wrong.
Where to start
Scoping a programme: the foundations, then the first ninety days in the programme section. No purchase required and it usually redirects the project.
You have data and no findings: the measures section, particularly the elapsed time decomposition.
Individual measurement is being proposed: the section on monitoring people, before anything about technology.
A deployment has stalled: the note on how these programmes fail.
What this is not
This is not legal advice. Descriptions of safety duties, data protection obligations and consultation requirements are general and vary substantially by jurisdiction.
It does not describe how to monitor people covertly, how to set rates that outrun the safe method, or how to work around the obligations that attach to workplace monitoring.