Serialization Aggregation

Serialization aggregation is the controlled process of linking serialized saleable units to their parent containers and logistics units, creating a verifiable hierarchy (unit → inner pack → case → pallet). Each relationship is recorded as an event so the hierarchy can be built, split, and re-formed with full genealogy. Aggregation enhances traceability beyond unit serialization by allowing inference of child serials from the scanned parent, provided the hierarchy remains intact and under control.

• Unit: typically SGTIN or equivalent device UDI-DI + serial at the smallest saleable level.
• Parent containers: inner packs and cases labeled with SSCC or case-level identifier.
• Logistics unit: pallets labeled with SSCC, often encoded in GS1-128 or 2D labels.
• Events: Commission, Aggregation, Disaggregation, Shipping, Receiving (often as EPCIS events).

In the U.S., 21 CFR 211.130 requires controls over packaging and labeling operations, including identity and lot/controls to prevent mix-ups, and 211.188 requires complete batch production records. While the Drug Supply Chain Security framework drives interoperable, serialized traceability, aggregation is the practical mechanism to carry serialized identity through distribution with strong records. Computerized systems used to execute and record aggregation must comply with 21 CFR Part 11 and 211.68, including validated state, access controls, and secure audit trails.

EU GMP (EudraLex Vol. 4) places expectations on computerized systems and qualification (e.g., Annex 11/15), and many markets mandate or encourage hierarchical identification at case/pallet level alongside unit serialization. Medical device supply chains increasingly use aggregation to operationalize UDI across 3PL/WMS interfaces and field returns. For cannabis and veterinary products, aggregation supports chain-of-custody, diversion controls, and rapid recall capability in line with good distribution practices.

Robust aggregation relies on global identifiers and interoperable event models. GS1 keys and data carriers are widely adopted. The parent–child links are persisted in event repositories and exchanged to partners to enable verification, receiving by scan, and recall execution.

• GTIN/SGTIN: GTIN identifies the trade item; SGTIN = GTIN + serial for the saleable unit.
• SSCC: Serialized Shipping Container Code for cases and pallets; parent keys in aggregation.
• Data carriers: GS1 DataMatrix (units) and GS1-128 (cases/pallets), with Application Identifiers (e.g., AI (01) GTIN, (21) serial, (00) SSCC).
• EPCIS: Event model to capture Commission, Aggregation, Disaggregation, Shipping, and Receiving, with business step, disposition, read points, and event time.

Minimum record content typically includes parent key (SSCC), child keys (SGTIN list), packaging resource/line, timestamps (synchronized), operator ID, equipment ID, reason codes for disaggregation/rework, and references to batch/lot and order. When multiple hierarchies exist (e.g., rainbow pallets), the data must reflect partial aggregation segments without ambiguity.

Aggregation is orchestrated at ISA‑95 Level 3 (MES) and executed via Level 2 equipment (printers, vision systems, scanners, case packers, palletizers). MES binds serial identities to production context (batch, order, recipe), enforces workflows, and records events; Level 4 (ERP/WMS) consumes the results for distribution, receiving, and returns. ISA‑88 procedures/phases often control the line steps that generate the events.

Typical top-down aggregation prints and verifies unit codes, scans child units into a case, prints the case SSCC (GS1-128), verifies it, then aggregates cases to a pallet (SSCC). Bottom-up (post-verification) aggregation reconciles child reads from cameras/scanners to build the parent. Inference—accepting that scanning a parent implies all registered children are present—requires strict controls to prevent tampering, substitution, and miscounts.

• Print-and-verify: reject on bad print, unreadable code, or duplicate serial detection.
• Case build: enforced counts, mixed-SKU prevention unless recipe-approved for rainbow cases.
• Pallet build: tie to load/ship tasks; allow partial pallets with controlled closure and later completion.
• Rework: disaggregate on open case/pallet; re-aggregate with new SSCC and reason-coded events.
• Exception handling: voided serials, damaged labels, overage/shortage reconciliation, and hold/release.

Every re-aggregation must be traceable to the prior hierarchy and reason-coded, with the ability to prove that no orphan serials exist. Time synchronization across devices is critical to correlate scans with packaging steps and prevent double-count/double-apply failures.

Systems executing and recording aggregation are GxP-relevant. 21 CFR 211.68 requires controls/validation of automated systems, and 21 CFR Part 11 applies to electronic records and signatures. Follow GAMP 5 risk-based validation: define intended use, perform supplier assessment, and verify functional risks (e.g., duplicate serial detection, orphan prevention, audit trails, exception workflows). Maintain ALCOA+ attributes: attribution to operator/equipment, legibility, contemporaneity, originality, and accuracy.

• Access control and segregation of duties for re-aggregation and disaggregation authorizations.
• Audit trail of who/what/when/why for each aggregation edit, including device IDs.
• Electronic signatures for critical steps (e.g., disaggregate, override) per Part 11.
• Backup/restore tests to ensure genealogy integrity after recovery.
• Change control for label templates, Application Identifier mappings, and EPCIS schemas.

Test negative scenarios: identical SSCC re-use attempts, child in two parents, partial-case closure, late scans, label reprint collisions, and system clock drift. Evidence should include screen captures/log extracts of audit trails and device data, and objective acceptance criteria for read rates and false-accept thresholds.

Post-packaging, MES publishes EPCIS Aggregation and Shipping events, with SSCC and children SGTINs, business steps and dispositions, and links to orders/shipments. WMS uses SSCC to build shipments, generate ASNs, and support scan-minimized receiving: scanning a pallet SSCC can receive all children, provided the hierarchy is intact and the receiver trusts the event source. Disaggregation on receipt (e.g., for QA sampling) must be recorded and shared to avoid stale inference downstream.

When partners are not EPCIS-enabled, a fallback is to share hierarchical packing lists secured to shipments. However, this increases manual reconciliation and risks orphaned serials after partial receipt or damage. Strong partner onboarding and data quality SLAs are essential to maintain end-to-end aggregation integrity.

Serialization aggregation adds new performance dimensions to packaging operations. Beyond OEE, track code quality and read performance at each hierarchy, exception loads, and the stability of inference across handling and transport. Trend these against materials (substrates/inks), equipment setups, environmental conditions, and operator shifts to identify systemic issues.

• First-pass read rate by level (unit/case/pallet) and by printer/vision cell.
• Duplicate/invalid serial detection rate and root cause attribution.
• Aggregation exception rate (over/short/rework) and cycle time to resolution.
• Disaggregation frequency and percent successfully re-aggregated without orphaning.
• Partner receiving mismatches per 10,000 SSCC scans and time-to-correct.

Establish control limits for read rates and exception thresholds; tie out-of-control signals to corrective maintenance, label/ink change, or operator retraining. Use periodic reconciliation (MES vs. WMS vs. EPCIS repository) to detect drift and trigger investigation.

• Inference leakage: pallets opened and reworked in uncontrolled areas; mitigate with tamper-evident seals and mandatory scan workflows on break/make.
• Orphaned serials: disaggregated units not re-associated; mitigate with forced reconciliation queues and auto-holds until closure.
• Clock drift: device timestamps out of sync; mitigate with NTP-hardening and monitoring per ICS security guidance.
• Label collisions: SSCC/SGTIN reprints without reallocation; mitigate with server-side uniqueness and printer lockouts.
• Mixed-SKU creep: recipe changes insufficiently propagated to aggregation rules; mitigate with centralized master data governance and e-signature-controlled releases.

V5 executes aggregation at the MES layer with native device integration (printers, cameras, scanners), enforcing unit→case→pallet workflows tied to batch/order context. It maintains immutable genealogy with Part 11-compliant audit trails and supports controlled disaggregation/re-aggregation with electronic signatures. V5 publishes EPCIS events to WMS/ERP and consumes receiving feedback to keep hierarchies current, eliminating stale inference. Quality events (deviations/CAPAs) and training records are linked to the same execution record; line asset maintenance and calibration readiness are checked in-line.

• Server-side serial/SSCC issuance and collision prevention; label template control with review/approve.
• Real-time read-rate monitoring and exception queues; forced reconciliation before release.
• EPCIS event capture and exchange; SSCC-driven pick/pack/ship and ASN alignment.
• End-to-end audit trail including operator, equipment, and reason codes for every hierarchy change.