Packaging KPIs

Pharma Packaging and Serialization KPIs and Benchmark Ranges

The KPIs that matter on a serialized packaging line, world-class versus typical benchmark ranges, how to measure them, and the levers that move each one.

Packaging line OEE is the headline KPI. Typical serialized pharma lines run 45 to 60 percent OEE, competent operations reach 65 to 75 percent, and world-class packaging sustains 80 to 85 percent. The gap is usually availability, not speed. Measure OEE per shift from the line meter and split the loss tree, because serialization adds micro-stops (print faults, camera no-reads, aggregation mismatches) that erode availability in one to three second bites. The lever is attacking the top three stoppage codes; on most lines print-verify no-reads and reject-station jams account for over half of unplanned downtime.

Serialization first-pass grade drives salable yield. Aim for 100 percent of codes at ISO 15415 grade B or better, with grade-failure scrap under 0.5 percent typical and under 0.2 percent world-class. Measure it from the verifier log as codes failing grade divided by codes printed. The levers are print head resolution, substrate contrast, and camera lighting geometry. A matte carton stock versus glossy can move average grade a full band. Track no-read rate separately from grade-fail, since a no-read is a detection problem and a grade-fail is a print-quality problem, and they get fixed differently.

Vision false-reject rate is where good margin disappears. Typical lines run 2 to 3 percent total reject with a large false-reject share; world-class holds false rejects under 0.5 percent and total rejects under 1 percent. Measure by auditing a reject bin: reconfirm a sample of rejected cartons and the pass-on-recheck fraction is your false-reject rate. The lever is camera tuning, lighting stability, and tightening trigger timing, not loosening spec. Cutting false rejects from 2 percent to 0.5 percent on a line running 250 cartons per minute recovers roughly 225 good cartons per hour.

Blister yield and defect rate benchmark tightly because the product is exposed. Good lines hold blister reject under 1 to 2 percent, and world-class forming and sealing runs under 0.5 percent. Measure as voided cavities divided by cavities formed, coded by cause: miss-fill, broken tablet, seal integrity, and foil defect. The dominant loss is usually feed-related miss-fill at the tablet hopper. Levers are feeder tuning, forming temperature control, and seal pressure and dwell. A stable forming station typically pushes yield past 99.5 percent, and the last half percent comes from hopper and detabber setup, not the sealing station.

Changeover time gates how small a batch you can run profitably. Typical serialized line changeovers land at 90 to 120 minutes; SMED-mature lines hit under 45 minutes and best-in-class under 30. Measure last-good to first-good, including line clearance and first-article verification, which serialization lengthens through reconciliation and code re-verification. Levers are externalizing setup, pre-kitting format parts, and standardizing verification. Use the Changeover Validation Time metric to track the clearance and verification portion separately, since that is the part serialization inflates and the part SMED alone will not fix without procedure redesign.

Batch release packaging hold measures how fast finished goods clear QA. Typical hold is 5 to 10 days from last carton to release; strong operations run 2 to 4 days and world-class 1 to 2. Measure as median days between packaging completion and disposition. Long holds are almost always documentation and reconciliation review, not testing. The levers are review-by-exception, electronic batch records, and clean serialization reconciliation so commissioned versus decommissioned counts match on the first pass. Every day cut off the hold frees real working capital; track it as standing quarantined pallets to make the carrying cost visible.

Aggregation accuracy is a pass-fail KPI regulators care about. Target parent-child association accuracy at 99.99 percent or better, with commission-decommission reconciliation matching exactly at batch close. Measure as mismatched or orphaned serials divided by total serials. A single unresolved aggregation error can hold an entire pallet. The lever is verified scanning at each hierarchy level and automated reconciliation rather than manual reconciliation at end of run. Lines that automate aggregation typically push error rates below 0.01 percent while cutting the labor that manual scan-and-pack stations demand.

Roll these into one review cadence: OEE and stoppage codes daily, grade and reject rates per shift, blister yield per batch, changeover per event, and release hold and aggregation accuracy weekly. The improvement sequence that pays fastest is availability first (kill the top no-read and jam codes), then false-reject reduction, then changeover via SMED, then release-hold via review-by-exception. Each lever moves a specific number, so hold every target against these ranges and chase the widest gap, not the easiest one. On most lines that means starting with the reject station and the print-verify unit, where the biggest availability and yield points are hiding.

Published 2026-07-02.