The 62% Problem Hiding in Plain Sight

When one of our customers, a contract packaging operation in Suzhou, first asked us to help improve the OEE on their four Zray ZX-650 carton forming lines, they told us their number was “around 75 percent.” We installed data loggers on each machine for two weeks and came back with the real number: 62.3%. The gap between perceived and actual OEE is almost always this wide. People remember the good hours and forget the creeping losses.

Breaking down that 62.3% was revealing. Availability was 87% — not terrible. Performance was 78% — the machines rarely ran at nameplate speed because operators throttled down when they sensed “something not right.” Quality was 91% — reject cartons were being manually reworked downstream. Multiply 0.87 × 0.78 × 0.91 and you get 0.617. The math is unforgiving.

Three-shift operation amplifies every OEE loss factor. Shift handovers introduce communication gaps. Night shift crews, working with less supervision, tend to run conservative speeds. Third-shift maintenance response is slower because the senior technicians work days. And the machine never gets a proper cool-down cycle, which accelerates wear on heat-sensitive components like vacuum pumps and hot-melt systems.

The Six Big Losses, Mapped to Three Shifts

We categorize carton forming machine losses using a modified version of Nakajima’s Six Big Losses framework, specifically adapted for continuous packaging lines.

Availability losses: (1) Setup and changeover between carton sizes — averaging 23 minutes per change on the Suzhou lines, versus a target of 12 minutes. (2) Unplanned stops from jam-ups, glue string contamination, and vacuum failures — 47 recorded events across four machines in two weeks.

Performance losses: (3) Reduced speed operation — night shift ran at 82% of nameplate speed. (4) Micro-stops under 2 minutes that operators cleared without logging — we counted 184 of these in the data, none in the logbook.

Quality losses: (5) Start-up rejects after changeovers — the first 20-40 cartons after each change were consistently out of spec. (6) Process drift during long runs — squareness gradually moved out of tolerance between the 6-hour and 10-hour mark of a continuous run.

The Interventions That Moved the Needle

Shift handover protocol (addressed availability and quality): We implemented a mandatory 10-minute overlapping handover with a standardized checklist. The outgoing operator must demonstrate three consecutive in-spec cartons while the incoming operator watches. The glue temperature, forming pressure, and vacuum level are read aloud and logged on a shared tablet. This single change reduced changeover time from 23 to 16 minutes and cut start-up rejects by roughly 60%.

Dynamic speed floor (addressed performance): Instead of letting operators choose their own speed, we set a hard minimum speed floor of 88% of nameplate for normal production, with a documented override process for “abnormal conditions.” Override frequency is reviewed weekly. Within three weeks, night shift speed climbed from 82% to 91% with no increase in rejects — because the operators actually learned that the machine could run faster reliably once they were forced to try.

Predictive vacuum monitoring (addressed availability): We installed vacuum transducers with trend logging on each machine’s suction cup circuit. When the steady-state vacuum level during a pick cycle drops below 85% of the “clean cup” baseline, an alert prompts cup inspection at the next scheduled pause. This caught 11 degrading cups before they caused jams in the first month, cutting unplanned stops by roughly 40%.

Measuring What Works

Six months after implementing these changes, the Suzhou plant’s combined OEE across four machines measured 78.4% — a 16-point improvement. More importantly, the variation between shifts narrowed considerably. Day shift ran at 79.1%, night shift at 77.8% — a gap of only 1.3 points, down from a previous spread of nearly 9 points. Consistent OEE across shifts is the real sign that a three-shift operation has its processes under control.