WEIMI / A MAXIMUM IS NOT A BOUNDARY
A short lot.
A separate next field.
Make the parser show where one attribute ends and the next begins.
Introduction
An incoming stock message contains an invented lot value, L7A, followed by another attribute. The receiving system places both values in the batch column because its extraction rule takes a fixed twenty-character slice. The lot looks longer than the label, while the next column is empty. This is an illustrative data-design scenario, not a WEIMI software incident or customer case.
A maximum field length is useful for validation, but it is not necessarily the boundary of a shorter value. Vending purchasers who receive supplier batch attributes should ask how the offered interface keeps neighbouring fields separate. A parser can accept a message without producing the intended batch record.
The official references are GS1 Application Identifier 10 and GS1 Application Identifier 17, reviewed on 11 October 2026. AI 10 is batch or lot number, while AI 17 is expiration date. This article addresses their field-format difference and the buyer’s boundary evidence request. It does not construct a complete barcode or establish full GS1 conformance.
Quick Answer
The AI 10 page gives format N2+X..20. Its displayed regular expression permits one to twenty characters from the specified character set. The data value is variable in length; twenty is the upper limit shown, not a target length for every lot. Preserve the supplier’s actual permitted value rather than adding characters to satisfy a fixed-width assumption.
The same reference lists “FNC1 required?” as Yes. AI 17, by contrast, is an expiration date in YYMMDD form with N2+N6 and FNC1 required listed as No. Those distinctions should inform the implementation provider’s message handling. Ask for the applicable GS1 rules and the actual scanner-to-interface representation, rather than inventing a universal printable delimiter.
For acceptance, use a short lot, a maximum-length permitted sample and a neighbouring attribute. Compare the extracted fields individually at the receiving system. A sample containing only an isolated lot may not expose a boundary defect. None of these sample fragments is presented as a complete valid GS1 symbol.
Comparison Table
| Official reference |
Listed data format |
Procurement implication |
| AI 10: Batch or lot number |
N2+X..20; variable-length data up to twenty permitted characters. |
Show the actual lot and the boundary before a following field. |
| AI 17: Expiration date (YYMMDD) |
N2+N6; fixed six-digit data. |
Keep the six-digit date attribute distinct from the batch. |
| FNC1 flag in the reference |
AI 10 Yes; AI 17 No. |
Provider must explain applicable symbol and receiving-interface handling. |
The table is a narrow comparison of two reference entries. It does not establish all allowed combinations, date-validation rules or the validity of a complete message. The reviewed pages also list mandatory associations. Have the implementation owner review the actual application context.
Who Should Buy This
Use this brief when supplier lot data reaches a vending stock system through GS1 messages, scans or intermediary files. It is relevant to a project where batch and date attributes are extracted from one incoming record. Confirm that the quotation includes such a data interface before assigning the requirement to the cabinet supplier.
The data owner should supply the expected field values. The implementation provider should show how the receiving route represents and separates them. Procurement should retain those expected results with the software scope and version. A reviewer should not repair missing context by guessing a separator or by padding a batch code.
If the workflow uses manual lot entry or a non-GS1 supplier file, define its own data contract. Do not apply the AI 10 field rule to every text field named batch. The purpose is an explicit boundary for the actual interface, not a universal naming convention for all stock software.
How We Evaluate Smart Vending Machines
The hardware shortlist below comes from public listings. We have not scanned a stock pack, tested a parser or imported a lot record into any listed machine. The following proposed workflow makes a supplier demonstration reviewable without using real sales or operational stock changes.
- Identify the incoming route. Record the symbol or message type and the receiving interface supplied.
- Define the sample fields. Retain the intended lot and any following attribute separately.
- Vary the lot length. Include a short permitted value and an agreed twenty-character example.
- Document boundary representation. Ask the provider to identify the applicable rules and representation at the receiver.
- Compare destination fields. Verify that neither the lot nor its neighbour absorbs the other’s content.
- Review invalid inputs. Agree how an oversized or otherwise invalid value is handled instead of silently trimming it.
- Retain the interface version. Link the input and output record to the offered configuration.
The buyer should distinguish field extraction from full message validation. The AI 10 page lists mandatory associations, including one of several identified AIs. This guide does not create a complete test barcode or verify those associations for a deployment. Use the responsible implementation specialist for that review.
Key Buying Factors
Maximum versus actual length. A system that allocates twenty characters may still need to recognise a shorter value’s end. Request a demonstration with a neighbouring field. The physical database column size does not prove correct parsing of an incoming message.
Exact batch preservation. Artificial padding, truncation or replacement can change a supplier’s intended lot reference. Agree permitted transformations with the data owner. Preserve the raw input for comparison so an operator can distinguish extraction from later display formatting.
Character validation. X..20 does not mean arbitrary unrestricted text. The official entry supplies a regular expression with a defined character set and one-to-twenty length. Ask the provider to use the relevant specification rather than accepting any string that fits the screen.
Boundary representation. The reference’s FNC1 flag is important, but a purchase contract should identify what the receiver actually obtains. The symbol function, decoded stream and exported file are distinct stages. This article does not assign one visible character to every route. Request the provider’s explicit mapping and evidence.
Date separation. AI 17 is a separate fixed-length date attribute. A parser should not append its digits to a short lot simply because they appear later in the message. Compare the lot and date fields independently. Correct extraction does not by itself decide food disposition or local date-labelling requirements.
Evidence scope. Stock management, vision recognition and a touchscreen are not proof of GS1 field-boundary handling. Identify the actual software and integration responsible for the import. Keep hardware suitability and parser acceptance as separate purchasing decisions.
Best Smart Vending Machines
These are three real manufacturer listings for a hardware shortlist. “Best” means fit for the quoted merchandise and format. No batch parser, scanner capability or recall performance is independently tested here.
CANDIDATE 1
Single-Door AI Vision Smart Fridge for Packaged Drinks
Read public listing →
The single-door AI vision fridge listing describes camera recognition for selected packaged products, five shelf levels and five baskets, plus a top-screen or light-box option. It stores compatible packaged goods and does not prepare fresh juice. Confirm cooling and recognition scope in the offer. If batch import accompanies product onboarding, request the software route and boundary evidence separately. Camera recognition is not proof that a GS1 lot field is parsed correctly.
CANDIDATE 2
WM22 Snacks and Drinks Vending Machine
Read public listing →
WM22 is described with a 21.5-inch touchscreen, inventory management and cooling. Spiral, conveyor, direct-push and hanging mechanisms appear as options and should be identified in the quotation. Match the physical product to the selected mechanism. If the offered inventory interface imports batch attributes, demonstrate a short lot followed by another field. The public listing does not establish that capability.
CANDIDATE 3
Two Cabinets, More Choice: Snack & Drink Vending Station
Read public listing →
The dual-cabinet listing shows a main product display and an additional visible spiral-stock area. Do not infer shared software, independent cooling, a second touchscreen or precise capacity. Identify any included stock-data interface and the selling-section context it uses. A cabinet arrangement cannot prove where an incoming variable-length field ends.
Feature Comparison
| Review item |
Useful evidence |
What it does not prove |
| Short lot sample |
Actual permitted value retained without padding. |
Complete GS1 message validity. |
| Maximum-length sample |
Twenty-character permitted value handled as agreed. |
Correct boundary handling for shorter values. |
| Following attribute |
Neighbouring field remains separately extracted. |
Every possible field combination was tested. |
| Raw-to-receiver mapping |
Provider documents applicable boundary representation. |
Universal scanner or export behaviour. |
| Invalid-value handling |
Oversized or invalid field gets the agreed response. |
Operator recall procedures are complete. |
| Hardware offer |
Products fit the quoted mechanism and storage scope. |
A GS1 parser or scanner is included. |
A successful import should be supported by field-level comparison. The number of accepted rows is not enough to identify whether a batch swallowed neighbouring data. Keep the expected values visible beside the observed results.
Cost & ROI Analysis
For a hypothetical review, assume four sample cases require ten minutes each, plus twenty minutes to record the raw-to-receiver mapping. The active work totals 60 minutes. At an assumed reviewer rate of US$38 per hour, the labour amount is US$38. These values are planning assumptions, not vendor charges or a customer case.
| Assumed activity |
Time |
Illustrative cost |
| Four sample reviews |
4 × 10 minutes = 40 minutes |
$25.33 |
| Boundary-mapping record |
20 minutes |
$12.67 |
| Total |
60 minutes |
$38.00 |
The two cost lines are rounded to cents. Software development, scanning equipment, data cleanup and full message review are excluded. Obtain actual quotations for the interface and the required acceptance evidence before budgeting integration work.
As a separate sensitivity example, assume reconstructing a corrupted lot record takes fifteen minutes at the same rate. The assumed labour is US$9.50 per record. Four such corrections would equal the US$38 review amount. This arithmetic does not establish that four defects occur or that a particular test prevents them. Use real defect and correction records for a savings claim.
Correct batch extraction does not establish vending revenue, avoided recall loss or equipment payback. Those outcomes require their own operational evidence. This article isolates the field-boundary procurement question.
Best Choice by Scenario
A short lot with another attribute. Make this the primary boundary demonstration. Retain the intended lot and neighbouring value separately and compare both at the destination. Do not substitute a maximum-length lot that makes a fixed slice appear correct.
A maximum-length lot. Include it as a length-limit case, using permitted characters. It complements the short-field case; it does not replace it. The data owner should approve the sample and expected value.
A scanner-to-file handoff. Ask the provider to show each transformation and the receiver’s boundary representation. A printable human-readable display is not automatically the same as the decoded message. Define the actual interface contract.
No GS1 batch integration in the offer. Choose the vision fridge, WM22 or dual-cabinet format on confirmed merchandise and handling fit. If a separate third-party system owns batch import, assign this evidence request to that provider instead of assuming the cabinet implements it.
Applications
At supplier-data onboarding, build a safe sample manifest containing an actual permitted lot value and a separately identified next attribute. Use an agreed non-production route. Avoid constructing a live product barcode or substituting a fabricated GTIN as though it were assigned.
At export review, compare the raw and destination batch text. A file may retain all characters while losing the information needed to distinguish adjacent fields. Record the boundary representation and any approved transformation, not just a screenshot of a populated batch column.
At a software change, repeat the short-lot and neighbouring-field sample against the changed interface. Keep the software version with the observed result. This proposed regression check is not evidence that any current WEIMI system has passed.
At operational handover, explain which owner resolves a mismatch. Correcting the parser and rebuilding affected records are separate tasks. A clean batch field supports a data foundation; it does not replace receiving, stock movement or recall procedures.
FAQ
Does AI 10 require exactly twenty characters?
No. The reviewed reference gives N2+X..20 and a regular expression allowing one to twenty characters from its specified character set. Twenty is the maximum in this field, not a requirement to pad every lot value.
Can a parser locate the next field by taking twenty characters after AI 10?
That assumption does not account for shorter lot values. Ask the implementation provider to demonstrate the applicable field-boundary handling in the actual message and receiving interface.
Does this article prescribe a literal separator for every export?
No. The GS1 AI 10 page lists FNC1 required as Yes. A symbol function and the receiving interface’s representation need an explicit implementation contract. The article does not claim one printable character applies to every scanner or file route.
How is AI 17 different in the reviewed reference?
AI 17 describes an expiration date in YYMMDD form, with N2+N6 and FNC1 required listed as No. It is a fixed six-digit data field, unlike AI 10’s variable-length batch field.
Does a clean lot field prove recall readiness?
No. Correct field extraction is one data result. Product association, receiving records, stock movements and the operator’s response process require their own evidence. No recall exercise is reported here.
Have the listed machines passed this parser test?
No. The shortlist is based on public manufacturer pages. No scan, batch import or field-boundary test has been run on these machines for this article.
Final Recommendation
Review AI 10 as a variable-length field with an actual value and an explicit boundary. Keep its maximum length separate from its received length. Use a short lot followed by another attribute to expose a fixed-slice assumption.
Ask the implementation provider to document the applicable GS1 handling and the receiving representation, then compare extracted values at the destination. Select the vending hardware independently on confirmed product fit. Neither the reference entries nor the shortlist proves a complete message, a scanner capability or a recall-ready deployment.
CTA
Send WEIMI your merchandise, preferred cabinet format and any required supplier-data integration. Include safe sample fields, the incoming message route and the expected destination values. Request a quotation that identifies the hardware and any offered batch-import scope.
Get My Custom Quote
This guide uses official AI references and public equipment listings. It reports no independent machine or parser test, complete GS1 conformance, customer ROI, legal conclusion or Google-indexing result.
