Separating planned maintenance from hours the line simply did not run
A four-hour ‘engineering’ block on a Friday can hide two hours of waiting for a part that was never booked. The chart only helps if the diary and the log agree.
Journal
Notes from books and from the line
Short pieces on how lost hours hide in codes, diaries, and first hours after a change of pack.
What a Pareto of stop reasons usually hides
The tallest bar is often the most popular code, not the most expensive wait. Supervisors reach for ‘other’ when the radio is busy.
The cycle-time board and what the line actually did
A laminated seconds-per-piece figure is a target, a memory, or a leftover from a different pack. Throughput cares which of those it is.
What to have in the shift logs before an outside review
We would rather have messy complete books than a tidy spreadsheet that dropped the night shift. Completeness beats presentation.
Plotting the first hour after a changeover
The run rate at hour four is a poor alibi for hour one. Start-up has its own waits: tickets, film, the previous SKU still in the hopper.