An Electronic Product Code gives software a way to name an object. For a serialized trade item, the key distinction is between the product type and one individual instance of that product. Encoding puts that identity into a form a tag and reader can exchange.
Start with the identity assigned by your organization or trading partner. Establish its GS1 Company Prefix length, product key and serial-number rules before choosing an encoder. A plausible-looking number or a tag that responds successfully is insufficient evidence of a correct identity.
Separate the product from the instance
A GTIN identifies a trade item at a defined packaging level. Combining it with a serial number distinguishes an individual instance. The SGTIN scheme expresses that serialized trade-item identity. Other objects can need other schemes: selecting an identifier starts with what the object is and how partners manage it.
GS1’s EPCIS implementation guideline explains the components of a pure-identity SGTIN URI: company prefix, indicator/item reference and serial. The GTIN check digit is omitted from that URI and reconstructed when returning to the GTIN. Preserve prefix length explicitly; it determines how the digits are divided.
On a small screen, scroll sideways to read both columns.
| Representation | Meaning and useful check |
|---|---|
| GTIN plus serial | The individual trade-item identity. Compare both fields when checking a serialized item. |
| Pure-identity EPC URI | A system-independent representation of identity, such as urn:epc:id:sgtin:… |
| EPC tag URI | Adds the selected binary scheme and filter value used for tag encoding. |
| EPC hexadecimal value | The encoded EPC bits shown as hexadecimal; distinguish these from surrounding protocol-control or CRC fields. |
| TID and other memory | Tag-manufacturing information and other data are separate from the assigned EPC. Check the chip specification for supported memory and serialization. |
Walk through a checked encoding
The following values are the published demonstration in GS1 RFIDcoder’s API documentation, reproduced here as a compact interoperability example. They are test data for learning, not identifiers allocated for your products. This walkthrough uses filter 1 and a seven-digit company prefix; those choices are part of this example.
The 24 hexadecimal digits hold 96 bits. For this seven-digit prefix, partition 5 allocates 24 bits to the company prefix and 20 bits to the indicator/item reference. The full layout is 8 header bits, 3 filter bits, 3 partition bits, those 44 identity-prefix bits and 38 serial bits. Their sum is 96. These are EPC payload bits; reader exports may include additional fields separately.
A correct round trip returns the same GTIN and serial after encoding and decoding. Compare parsed identity fields, not only strings copied from a reader screen. The downloadable demonstration JSON preserves the inputs and expected outputs with its source and restrictions. It can serve as a known test case for an adapter.
Preserve serials and memory boundaries
The GS1 EPC Tag Data Standard 2.3 defines encoding constraints. SGTIN-96 represents a numeric serial smaller than 238; leading zeros other than the single digit zero cannot be preserved by that numeric representation. An alphanumeric serial or significant leading zeros requires a suitable supported alternative encoding. Never silently turn “00400” into “400” to make a field fit.
The filter is an encoding aid; it does not create a new business identity. Likewise, a TID identifies tag-related information according to the chip’s capabilities, while the EPC names the assigned object. Replacing a damaged tag can change tag hardware without changing the intended item identity. Record that replacement relationship in the application.
Lot, expiration and other attributes may live in business records or supported additional tag data. Determine the actual encoding agreement. A decoded SGTIN alone establishes neither the full product history nor authenticity. Assess provenance using the surrounding records and controls.
Diagnose a mismatch from the outside inward
- Capture the raw reader output, its field names and the decoder error before making corrections. Keep access to real production identifiers controlled.
- Confirm the expected scheme, EPC length, prefix length and serial format. Check that the decoder is receiving EPC payload rather than a concatenated dump of several memory fields.
- Run the known demonstration through the same adapter. A failure here points to parsing or configuration before product-specific data is considered.
- Decode a real sample and compare GTIN and serial with the authorized source record and visible label. Separate wrong product, wrong serial, duplicate identity and unreadable tag.
- Reconcile physical labels before issuing replacements, then repeat the round-trip check and the actual application transaction.
A duplicate identity on two objects is a data-control fault even when both radio reads are perfect. Conversely, repeated reports of the same correctly labeled object may be normal reader behavior. Preserve the distinction when reviewing inventory discrepancies. Use the RFID integration guide for deduplication and accepted business events.
Take the guide into your workflow
Download the working checklist (plain text). The editable text file includes its purpose, instructions, evidence fields and a link back to this guide. Record observed results and unresolved questions before making the decision.
Related reading
- Understand the radio protocol and standards layers
- Verify labels and manage reprints
- Document customer tagging acceptance
- Explore all RFID guides
Researched and updated September 11, 2026. Recheck the linked primary sources when equipment, standards or operating requirements change.