Step 3 of 4
57%
Operator and crew procedures: the small habits that double die life
Ten shop-floor habits, grouped into four routines, that separate a 40K-hit shop from a 120K-hit shop on the same die. None of them require new equipment. All of them require that the crew actually do them, every shift, and log the result.
Step 3 of 4Handoff and records
The shift ends. The next crew arrives. This is the highest-leverage and most-frequently-skipped routine in the lesson. Three habits.
Shift-handoff communication, structured. Not "everything's fine, see you tomorrow." A standing handoff checklist: current hit count on each die in service, any anomalies logged this shift (referencing the observation log entries), any process changes made and why, current lube tank level and last concentration check, any pending maintenance items, any parts the QC sample flagged for review. Five minutes, written, signed by both shifts. The cost of skipping this is the night-shift dial-up scenario from Step 1: information that existed on one shift does not survive to the next shift, and the new crew compensates blindly for a problem they cannot see.
Log every anomaly, not just failures. A failure log captures the moment the die was pulled. An anomaly log captures the chain of small drifts that led to the pull. The anomaly is the louder release at hit 320, the part that came off 30°C hotter than typical, the spray pattern that took a second extra to stabilize after the lube refill, the brief moment the alignment check showed 0.003 inch of parting-line variation that returned to 0.001 inch on the next cycle. A year of anomaly logs across a die program tells the shop things a teardown report cannot: lube batch effects, seasonal effects, slow drift in press alignment, operator effects. Lesson 13 covers the full log specification. This habit is about writing the entry while the anomaly is happening, not at end of shift when the memory has faded.
Photograph every die at every change-out. Standardized lighting, standardized angle, standardized scale reference in the frame, taken before any cleaning so the surface state is preserved. The single most useful inspection record a shop can build is a trend album of the same die at the same angle at every change-out across the die's life. Heat-check evolution that takes 50K hits to develop is invisible from one inspection to the next; it is obvious from photo one to photo six. The discipline is in taking the photo every time, not in waiting for a teardown.
The pushback here is that handoff and photo discipline feel like paperwork. They are not. They are the shop's memory. The shop without them is rebuilding context every shift, every die change, every inspection. The shop with them compounds.
Quick check
A die failed at 68K hits with a heat-check network that looked, in the teardown photographs, like it had developed over the last 15K hits. The shop has change-out photos for this die at 10K and at 60K hits and nothing in between. What does the missing 30K of trend data cost the program, and what is the corrective action for the next die?