Records and traceability: what to log, what it tells you over time

The four operator-controlled levers from Lesson 1 (preheat, lube, inspection cadence, change-out timing) all drift over time in ways no single die can show. Lube concentration drifts when a new barrel goes in. Operator habit drifts when a senior operator goes on vacation. Shop humidity drifts with the season. Press ram alignment drifts on a months-long curve nobody is watching, because the press is "the press" and nobody runs a dial indicator on it weekly.

A single die going through 50K hits sees one slice of those drifts. A logbook covering ten dies over a year sees the drifts as patterns. Lube barrel 7B correlated with a 0.5% scrap-rate uplift across four dies on three different jobs. The night-shift change-out criterion is consistently 8K hits later than day shift on the same job. Scrap rate climbs in July and August on the press in the bay without HVAC. The press 3 die-life average dropped from 95K to 78K over six months and nobody noticed because no one die fell off a cliff. None of those patterns are visible in any single teardown. All of them are visible in a logbook that someone actually filled out.

The reason logging fails in most shops is not that operators cannot write. It is that the log is treated as paperwork instead of as the only instrument that measures the slow variables. A shop that runs a calibrated thermocouple on the die face but does not write down which barrel of lube was running that shift is instrumenting the fast variable and ignoring the slow one. The fast variable is easier to see and rarely the one that bites.