OEE for injection molding is Overall Equipment Effectiveness, Availability × Performance × Quality, measured on a molding cell where the mold, not the press, sets the ideal cycle time. Output per cycle equals the number of good cavities, so cavitation drives the count, mold changes are the big availability loss, and short shots and flash are the quality losses.
Injection molding is closer to textbook OEE than a wet food line, because it is discrete and cyclic: the press closes, injects, holds, cools, opens, and ejects, over and over. But two things trip up molding OEE. The ideal cycle time belongs to the mold, not the machine, so the same press has a different baseline for every tool it runs. And output is multiplied by cavitation, so a blocked cavity quietly cuts your good count on every shot. This guide works through both, plus how to treat short shots and flash. If you want the arithmetic done for you, the free OEE calculator runs the standard formula.
How is OEE calculated for an injection-molding cell?
It is calculated with the standard formula on the cell, counting molded parts, with the mold's optimized cycle time as the ideal. Availability is run time over planned production time; Performance is actual parts over what the run should have produced at the mold's ideal cycle time and cavitation; Quality is first-pass good parts over total parts molded.
- Availability captures mold changes, material and color changes, dryer and feed issues, and machine faults. Mold changes are usually the single largest block, which is why they map to the setup line of the six big losses.
- Performance captures cycle-time drift: the press running a longer cycle than the mold's validated optimum, plus short interruptions. A cycle that creeps from 30 to 33 seconds is a 9% Performance loss that never shows as downtime.
- Quality captures short shots, flash, sink, warpage, contamination, and color streaks. Only first-pass good parts count, the same logic as first pass yield; regrind recovery does not make a scrapped shot good.
Why does the mold set the ideal cycle time, not the machine?
The mold sets the ideal cycle time because cycle length is governed mostly by cooling, and cooling depends on the part's wall thickness, the mold's cooling design, and the resin, all properties of the tool and part, not the press. The same 200-ton machine might run a thin-wall lid at an 8-second cycle and a thick fitting at 45 seconds; a single machine-level "ideal rate" would be meaningless.
Because cooling typically accounts for a large majority of the cycle, the mold's validated optimum cycle time is the right ideal, and it has to be stored per mold and defended. Set it too soft, from a slow historical average, and Performance flatters the cell while real speed loss hides; set it from the mold's scientifically established best cycle and Performance finally measures cycle drift honestly. This is the molding-specific version of the "soft ideal cycle time" trap in the core OEE calculation guide: on a shared press the trap multiplies, because every tool needs its own correct baseline.
How does cavitation change the count?
Cavitation changes the count because a multi-cavity mold produces one part per cavity per cycle, so ideal output equals cavity count times cycle rate, and a blocked or shut-off cavity reduces the good count on every single shot. An eight-cavity mold running at a 20-second cycle has an ideal output of eight parts every 20 seconds; block one cavity for a defect and you are making seven, a 12.5% hit that repeats until the tool is fixed.
This creates a decision like the multi-head filler's. If a cavity is intentionally shut off, you can either treat the run as reduced-cavitation with a re-baselined ideal output, or treat the missing parts as loss, but you must pick one and record the cavitation with the run, or Performance and Quality both drift. Short shots and flash are the other side of cavitation-linked quality: a short shot is an incompletely filled cavity, flash is excess material forced past the parting line, and both are scrap that first-pass Quality has already lost. Because a blocked hot-runner drop or an unbalanced fill can turn one cavity bad while the rest are fine, cavity-level defect tracking is what tells you whether Quality loss is a process problem or a single-cavity tooling problem.
The standard behind the factors. The international KPI standard ISO 22400-2 keeps the familiar Availability × Performance × Quality structure but pins down each input precisely, so two plants following it reach the same OEE from the same facts. For scale, the Federal Reserve's G.17 release put U.S. manufacturing capacity utilization at 75.8% in April 2026 about 2.4 points below its 1972–2025 average, real plants, molding included, run well under theoretical maximum.
Why are mold changes the biggest availability loss?
Mold changes are the biggest availability loss because pulling one tool and hanging, connecting, and qualifying another can take hours, and the press makes nothing the entire time. Add the material and color change that often rides along, plus the startup scrap while the new tool stabilizes, and a single changeover can dwarf a shift's small stops.
That makes molding an ideal candidate for setup reduction. Quick-changeover methods, staging the next mold, pre-heating it, converting internal steps to external, and standardizing water, hydraulic, and electrical connections, routinely cut mold-change time by large fractions without capital. Every minute saved is a minute of Availability recovered, and shorter changeovers make short-run molds economical, which is often the real business goal. The startup scrap after the change is Quality loss and worth its own reason code; a stable, well-documented startup sequence shrinks it. For the wider tooling and process picture, the injection-molding process and plastics-manufacturing guides go deeper.
What drives Performance loss between changeovers?
Between changeovers, Performance loss on a molding cell comes from cycle-time drift and short interruptions rather than long stops. A cycle that quietly lengthens, an operator adds a second of cooling to chase a marginal defect, the tool runs hotter as the day warms, the robot take-out hesitates, produces fewer shots per hour without ever registering as downtime.
The reason drift is dangerous is that it is invisible on a manual downtime log: the press never stopped, it just ran slow, so the loss lands silently in Performance where nobody investigates it. The fix is to compare actual cycle time to the mold's stored optimum, shot by shot, and flag sustained drift. Automatic cycle capture also catches the sub-minute interruptions, a mispick, a brief door-open, a feed hiccup, that manual logs miss entirely. This is the same argument for measuring at the source made about machine downtime generally: reconstruct cycles from memory and the biggest, most fixable Performance losses never surface. Material drying and feed problems belong here too, since a starved or wet-material run slows cycles and adds scrap at once.
How do you build honest injection-molding OEE step by step?
Build it by storing the ideal cycle time and cavitation per mold, logging mold changes, and tracking defects by cavity before you multiply. Here is the procedure:
- Count parts at the cell. Take the shot counter and cavitation from the machine, not a manual tally, and record which mold ran.
- Fix planned production time. Subtract planned breaks and maintenance; decide once whether planned mold changes are excluded or counted, and apply it consistently.
- Store ideal cycle time and cavitation per mold. Use the mold's validated optimum cycle and full cavity count; re-baseline if cavities are shut off.
- Log mold changes and stops with reason codes. Mold change, material change, dryer fault, and machine fault all count against Availability.
- Track defects by cavity. Short shots, flash, and other scrap are Quality loss; cavity-level detail separates process problems from single-cavity tooling.
- Compute and cross-check. Availability = run ÷ planned; Performance = actual ÷ (run × ideal rate); Quality = good ÷ total; OEE = A × P × Q. Confirm good count × ideal cycle time ÷ planned time agrees.
What should a molder do with the number?
Use molding OEE for trend and decomposition on one cell, reading the factors against the mold that ran. A drop in Availability points at mold-change time; a Performance dip means cycles are drifting off the mold's optimum; a Quality slide points at short shots or flash and, with cavity data, at which cavity. Because the ideal cycle time is per mold, comparisons only make sense within the same tool, not across a whole press's mixed schedule.
All of it rests on capturing shots, cycle times, cavitation, and cavity-level defects at the machine rather than reconstructing them from memory. That is the real-time operational layer Harmony provides, machines, systems, and paperwork connected without rip-and-replace (see the platform or read the CLS case study). Decide what target you are chasing with a good OEE score watch mold-change losses the way you would any machine downtime and put your own numbers through the OEE calculator.