KDE – Key Data Elements (FSMA 204) – Traceability Atoms You Must Capture

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Food Traceability Rule • Records, EPCIS, & Rapid Recall Readiness

Key Data Elements (KDEs) are the specific pieces of information the U.S. FDA’s Food Traceability Rule (FSMA 204) requires firms to capture and keep for defined supply-chain steps. KDEs travel with traceability lot codes across Critical Tracking Events (CTEs) like receiving, transforming, creating, and shipping. When a recall or outbreak happens, KDEs let you identify exactly which lots were touched, by whom, when, and where—so you pull only what’s risky, not entire categories or months of production. In practice, KDEs become structured fields inside your MES/WMS/ERP and are serialized, barcoded, and exchanged electronically via standards like EPCIS, using identifiers such as GS1/GTIN and, where relevant, batch/lot numbers enforced by Barcode Validation.

KDEs are not optional metadata; they are the defensible evidence that ties an ingredient’s path from supplier to finished package. The rule tells you which data to keep at each event and how quickly you must render it. Your job is to design processes and systems so KDE capture is automatic at the moment work happens—scans, signatures, device reads—not reconstructed later from emails and memory. With a proper data model and controlled labels, KDEs also enable surgical quarantine and dynamic lot allocation, keeping good inventory moving while suspect product is contained.

“KDEs are the atoms of traceability. If you don’t capture them when and where things happen, you won’t assemble the molecule during an investigation.”

1) What It Is

A KDE is a discrete, typed field—product ID, lot code, quantity, date/time, location, business step, and counterpart (supplier or customer)—that must be recorded for products on the FDA’s Food Traceability List and for items transformed with them. Each KDE attaches to a CTE, creating a chain of evidence you can traverse bidirectionally: back to sources and forward to recipients. In an integrated stack, KDEs are captured by Goods Receipt scans, Incoming Inspection results, eBMR consumption records, and EDI/EPCIS messages during shipping. The end goal is not a pretty dashboard; it’s a complete, queryable graph of which lots went where and when, with proof.

2) KDEs by Critical Tracking Event (CTE)

Receiving. Capture product description/ID (GTIN), supplier name/address, supplier lot/traceability lot code, quantity and unit, date/time received, and receiving location. Include carrier/transport identifiers when available. Validate identity via barcode scans and link to CoA and Identity Testing results if applicable.

Creating. When you assign a Traceability Lot Code (TLC) for first production or repack, store the product ID/description, the newly created TLC, the date/time and location of creation, and any component links if the “create” step includes ingredients (e.g., producing WIP from bulk). This is where label control under Document Control matters.

Transforming. For mixing, reprocessing, or assembling, record inputs (each component lot and quantity/unit), outputs (new TLC/lot, quantity/unit), the transformation date/time, and location. This step is the heart of Batch Genealogy—consumption must be contemporaneous in the eBMR.

Shipping. Capture recipient name/address, product ID, shipped lot/TLC, quantity and unit, ship date/time, ship-from location, and transport identifiers. Exchange the data electronically—preferably via EPCIS events or EDI—so your customer’s system ingests and validates the chain.

3) Data Model: Identifiers, Codes, and Context

Your KDE schema should normalize product identity (GTIN + description), lot/TLC (single field with type), quantity with units, timestamps in UTC with timezone context, GLN/addresses for locations, and partner identifiers. Barcodes must be durable and scannable; avoid human-only labels. Guard master data with Document Control and integrate with WMS bin logic (Bin Location) so KDEs always reference the current physical truth. For perishable items, couple KDEs with expiry/production dates and enforce FEFO to avoid embedding spoilage risk into the chain.

4) How to Capture KDEs (Without Slowing the Floor)

Make KDE capture invisible to operators. At receiving, scan supplier case/pallet labels and the ASN; your system should parse GTIN, lot, and quantities automatically. In production, let the Job Traveler and device integrations write consumption and output KDEs directly into the record. At shipping, generate EPCIS events from your pick/pack/ship scans; never retype. Enforce Dual Verification only where risk warrants (e.g., first print/apply, allergen changeovers) to avoid fatigue.

5) Evidence, Rendering & Time-to-Trace

FSMA 204 expects you to produce KDE-linked traceability records quickly (commonly within 24 hours) for defined time windows and scopes. That means your system must render: (1) all recipients for a given shipped TLC, (2) all sources for a given finished lot, and (3) a time-bounded slice of events at a site. Store immutable audit trails (GxP Audit Trail) and signatures (Part 11) so your answers are not just fast but defensible.

6) Common Failure Modes (and How to Avoid Them)

Free-text chaos. Uncontrolled descriptions break joins. Countermeasure: standardize on GTINs, partner IDs, and controlled vocabularies; validate on scan.

Repacking without TLC assignment. Breaking bulk then shipping “mystery lots.” Countermeasure: force a create CTE with new TLC and label issuance under control.

Late/retro KDE entry. Backfilling from memory. Countermeasure: capture at the moment of work via scanners and travelers; block Job Release if required fields are missing.

Label/print drift. Wrong template or illegible codes. Countermeasure: controlled templates, print verification, and Barcode Validation on print and apply.

Partial genealogy. Missing one input breaks the trace. Countermeasure: enforced component scanning on dispense/issue and consumption—no manual bypass without Deviation/NC.

Data silos. WMS, MES, and ERP disagree. Countermeasure: event-driven integrations using EPCIS so everyone shares the same truth.

7) Metrics that Prove KDE Discipline

Track scan adherence at receive/consume/ship, label verification pass rate, genealogy completeness (% jobs with all inputs captured), time-to-trace (minutes to render sources/recipients for a lot), ASN match rate, duplicate/illegible barcode rate, and exception closure time for traceability deviations. During drills, measure time to compile recipient lists and to issue targeted holds in WMS (Bin Location moves) without halting unaffected product.

8) Practical Walkthrough (Leafy Greens to Packaged Kits)

A processor receives romaine (on the Food Traceability List). At Goods Receipt, the dock scans pallet labels: GTIN, supplier lot, quantity, carrier, date/time, location—KDEs posted automatically. Incoming micro results and CoA are linked; Component Release flips status to Approved. In prep, operators scan romaine lots into a wash mix; the eBMR records a transform CTE with inputs/outputs and assigns a new WIP TLC. Later, a salad kit work center mixes washed romaine with cheese and dressing (non-FTL). The traveler logs all input lots and creates a finished TLC for each kit batch; label print/apply pulls the TLC and GTIN from masters under Document Control. At shipping, pick/pack scans generate EPCIS ship events—recipient, lots, quantities, and timestamps. A week later, a farm-level alert triggers a targeted trace: QA queries recipients of the affected supplier lots; WMS issues holds by Batch-to-Bin; unaffected kits keep moving. Total cycle: under an hour, with a complete genealogy and audit trail.

9) How This Fits with V5

V5 by SG Systems Global treats KDEs as first-class citizens. In V5 WMS, Directed Picking, FEFO/FIFO, and Bin Location rules ensure the right lots move; scans populate KDEs automatically. In V5 MES, the Job Traveler enforces component scans and contemporaneous consumption; outputs receive TLCs and controlled labels. Event services publish EPCIS captures for receive/transform/ship; plan and ASN data flow via EDI. QA reviews by exception using Data Integrity rules and Audit Trail. During drills or events, V5 renders recipients/sources in minutes, and integrates holds with WMS so containment is precise—not blunt.

10) FAQ

Q1. Are KDEs only for Food Traceability List items?
They are mandated for FTL items and for products transformed with them, but many firms extend KDE discipline to all SKUs to standardize processes and speed investigations.

Q2. What’s the difference between a TLC and a lot number?
A traceability lot code (TLC) is the code required by FSMA 204 and may map to or differ from your internal lot; ensure your system can link them and display the right one on labels and records.

Q3. Do spreadsheets satisfy FSMA 204?
Not safely. You need contemporaneous capture, identity controls, and rapid rendering with audit trails. Spreadsheets struggle with scale, integrity, and response time.

Q4. Do we have to use EPCIS?
It’s not the only path, but it’s the most interoperable way to exchange KDEs and CTEs with partners so traceability works beyond your walls.

Q5. How long should we keep KDE records?
Retention must meet regulatory and customer requirements; align with your document/data retention policy under Data Retention & Archival and ensure readability over time.

Related Reading
• Standards & Regulation: EPCIS | GS1/GTIN | EDI | 21 CFR Part 1
• Execution & Records: eBMR | Batch Genealogy | Barcode Validation | Directed Picking | Bin Location Management
• Quality & Governance: Audit Trail (GxP) | Data Integrity | Hold & Release | CoA | Incoming Inspection