We used videotaping of a line operator performing an operation, and used this to effectively test gage repeatability. Especially, it tested the operational definition of start time and stop time for the cycle.
Another approach is to create a mock system that responds like the real one. But then the “repeatability” variation is still made up of measurement variation and short-term process variation. Fact is, cycle time of a process is inherently a “destructive” measure. You can still have two or more people measure the event and get reproducibility.
.
