An approved disposable AIO sample can prove that a supplier built several units the buyer was willing to accept. It does not prove that the same configuration will move through the intended production route without mixed revisions, undocumented substitutions or variation that the sample set never exposed.
That is the job of a pilot run.
For an empty AIO program, the decision depends on whether the buyer and supplier can reconstruct what was produced, under which conditions, what the evidence actually covered and which open issues must close before mass production.
Write that decision before the run begins:
Decide whether the approved AIO configuration and its intended production controls are ready for repeatable mass production, conditionally ready with named restrictions, or not ready.
A precise decision statement keeps the pre-production run from becoming a showroom sample exercise or an ambiguous small production order.
Prerequisites / controlled inputs
Enter the pilot with an approved sample record for the exact AIO configuration you intend to produce. The checks should match its design. For a dual-chamber model, the dual-chamber AIO sample-testing guide shows how to record separate fills, supported switching modes and chamber-isolation observations. Confirm separately that the supplier is approved for the defined program through program-specific vape hardware supplier qualification. These upstream records become pilot inputs; the pilot retains its own mass-production-readiness decision.
Start with an approved configuration, not a product nickname
A pilot cannot test repeatability when the baseline is still moving.
The input should be a controlled configuration connected to the approved sample, specification and commercial order. Depending on the AIO, that identity may include:
- supplier, manufacturing site, model and project code;
- drawing, specification and bill-of-material revision;
- capacity, postless or center-post architecture and intake configuration;
- reservoir and other oil-contact materials;
- heater identity and approved resistance range;
- battery, printed circuit board (PCB), activation logic, charging interface and controlled power behavior;
- closure method, finish, color, screen or indicator behavior;
- artwork, label and packaging revisions;
- the sample ID and evidence used for approval.
This identity record prevents a familiar-looking AIO from entering the pilot with a different cell, intake, seal, firmware, surface finish or package than the buyer evaluated.
Define what “representative” means before the run
The record must show which pilot conditions represent the intended production route.
The supplier should identify which conditions match the intended production route and which do not. Useful context can include the assembly line, tooling or cavity, component sources, setup controls, production sequence, inspection methods, firmware-loading route, artwork application and packaging operation. The relevant fields depend on the actual design and the failure modes that matter to the program.
A hand-built unit may answer an engineering question; its record should identify that it does not represent line repeatability. A pilot using temporary packaging can inform hardware production while the production-packaging gate remains open. Record the departure instead of stretching the conclusion.
There is no responsible universal rule that every AIO pilot needs the same number of units or runs. The rationale should consider the novelty of the configuration, process complexity, tooling and line coverage, destructive examinations, previous evidence, observed variation and the consequences of a missed failure mode. A copied quantity without that reasoning creates the appearance of discipline rather than evidence.
Pharmaceutical FDA process-validation requirements do not legally govern empty vape hardware manufacturing. Their underlying engineering logic—understanding the process, collecting objective evidence and controlling variation at scale—offers a transferable manufacturing principle. FDA’s Production and Process Controls Q&A also makes clear within its own regulated context that a justified, science-based rationale matters more than a universal minimum batch count. This article does not import pharmaceutical batch-count expectations into AIO hardware.
Build one pilot evidence board
Scattered photos, chat messages and spreadsheet tabs are difficult to review as one decision. Use a board or controlled record that keeps four evidence layers connected.

Card 1 — Frozen identity
- Establish: the exact configuration and approved references entering the pilot.
- Force the question: Did the pilot use the approved battery, intake, finish and artwork revisions?
Card 2 — Production context
- Establish: where and how the units were made, including every known nonrepresentative condition.
- Force the question: Did the run use the intended line, tooling, component route and packaging process?
Card 3 — Evidence and observations
- Establish: what was checked, on which identified units, by which method and under whose ownership.
- Force the question: Does each result describe empty hardware, the buyer’s filled system or a defined interface between them?
Card 4 — Exceptions and decision
- Establish: affected scope, containment, owner, closure evidence and the current decision state.
- Force the question: Can mass production start, and what remains restricted until closure?
Each observation should point back to identifiable units or a defined production segment. “Five devices failed” is weak if the record cannot show the revision, line, time window, method or affected population. Conversely, a polished chart is not useful when its sample identity cannot be reconciled with the pilot build.
Do not force every result into pass or fail. A measurement near a limit, an intermittent activation observation or a packaging mark concentrated in one cavity may justify investigation even when a simple attribute count has not triggered rejection. Preserve the actual reading or observation when it helps the team see drift.
Separate empty-hardware evidence from the buyer’s filled system
An empty AIO supplier and an oil brand do not control the same variables.
The supplier can document the hardware configuration, component identities, assembly route, in-process controls and empty-device inspections. The buyer or filling partner controls formulation identity, conditioning, filling equipment, fill settings, closure timing, post-fill handling, finished packaging and the distribution study appropriate to the intended program.
Some questions sit at the interface. Leakage, wicking, activation and aerosol behavior can depend on the hardware, formulation, filling process, closure, storage and use conditions together. A pilot record should name which party generated the evidence and which configuration it covers.
Avoid either extreme:
- Do not accept “hardware passed” as proof that every formulation and filling process will work.
- Do not attribute every filled-system observation to the hardware before checking filling, capping, conditioning and handling records.
Battery documentation has a similar boundary. A traceable UN 38.3 test-summary review can connect a battery type to the AIO configuration for its stated document purpose. It does not prove oil compatibility, finished-device shelf life or readiness for every shipping route.
A fictional pilot file shows how the decision works
The following fictional case demonstrates the decision mechanics. It does not represent an ILEVA customer, factory run, test result or recommended acceptance threshold.
Case file: AIO-PILOT-07
Controlled input — A frozen 2 ml screen-equipped empty AIO configuration. Its record identifies the hardware revision, battery type, resistance range, intake configuration, firmware behavior, black finish and retail artwork revision B.
Evidence confirmed — Component and empty-hardware records match the frozen revision. The intended assembly route and firmware-loading process were used. Traceable in-process readings and final empty-device observations are available.
Deviations discovered — Some pilot cartons carry artwork revision A. The buyer also reports one filling-line observation whose unit identities and filling settings are incomplete, so causation cannot yet be assigned.
Containment — The supplier segregates the affected cartons and identifies the packaging time window. Hardware-production evidence and the buyer’s filled-system investigation remain separate workstreams.
Disposition logic — Conditional go. The evidence supports only the scope stated below.
- Hardware production readiness is supported for the frozen configuration and recorded route.
- Packaging release remains blocked until revision identity is corrected and reconciled.
- The filled-system question remains open until the buyer reconstructs the units, oil, filling settings, closure timing and observation method.
- Mass production cannot silently absorb either issue. Each has an owner, affected scope and closure requirement.
A pilot earns its value by improving the quality of the next decision.
Use three disposition states
Go
Use “go” only when the approved configuration, representative route, required evidence and decision responsibilities are sufficiently complete for the defined mass-production scope. The record should state exactly which model, revision, site, process and package are covered.
Go does not mean every future lot is automatically acceptable. Routine production still needs lot identity, supplier release controls and the buyer’s receiving system.
Conditional go
Use “conditional go” when the evidence supports a limited next step but named items remain open. Conditions must be operational, not vague reassurance.
Record:
- what may proceed;
- what remains blocked;
- the affected units, configuration or time window;
- containment already in place;
- the owner and due date;
- evidence required for closure;
- who has authority to remove the restriction.
If the condition involves a temporary departure from the approved baseline, route it through a controlled supplier deviation request. A pilot is not permission to normalize an exception.
No-go
Use “no-go” when configuration identity is uncertain, representative production was not demonstrated, a material failure remains uncontrolled, or the evidence cannot support the intended scope. State what must change before another pilot or review can begin.
No-go is more useful when it separates the failure mechanism from the remedy. “Pilot failed” does not tell the supplier whether to correct the configuration, process, method, packaging route or evidence chain.
Do not use AQL to answer the pilot question
An AQL sampling plan for empty vape hardware supports acceptance or rejection of a defined lot under a documented attribute-sampling system. ISO 2859-1:2026 addresses statistical lot acceptance under an established attribute-sampling plan and defect classification. A pilot instead examines whether the configuration and production route are repeatable and can expose process failure modes that routine defect categories have not yet captured.
A pilot may use inspection methods that later appear in routine lot acceptance, and its results may help refine the defect catalogue or receiving plan. But a small lot passing an AQL table does not establish process capability, close a configuration gap or explain whether the production conditions represented the intended route.
Keep the handoff explicit:
PILOT EVIDENCE → CONTROLLED PRODUCTION BASELINE → LOT-SPECIFIC RELEASE → BUYER RECEIVING DECISION
Each gate answers a different question.
Convert pilot evidence into production controls
The pilot earns its value when its accepted conditions survive into later orders.
Carry the approved configuration and process references into the purchase order, quality agreement or controlled supplier file. Define which changes require notice, which records accompany future lots and how mixed or suspect revisions are contained. Preserve appropriately identified physical references through a retained-sample program so later investigations do not depend on memory.
The production handoff should identify, as applicable:
- the configuration and revision released for production;
- approved supplier, site and route;
- critical component or process identities that must not change silently;
- inspection methods and controlled acceptance criteria;
- packaging and label revisions;
- open conditions and their closure status;
- retained references and record locations;
- change-notification and deviation routes;
- the controlled change-control route for permanent configuration or process differences identified after the pilot;
- inputs to incoming inspection and lot history.
When the first production shipment is complete, the pre-shipment release review decides whether that specific lot has the evidence required for factory release. It should not rely on the pilot outcome as a permanent substitute for shipment evidence.
What the evidence must show
A pilot run is valuable because it creates a controlled bridge between an approved sample and routine production. The bridge fails when the configuration drifts, the conditions are not representative, filled-system responsibilities blur or open issues disappear into a general “pass.”
The strongest pilot file lets a new buyer, quality engineer or supplier contact reconstruct the same decision later: what was built, how it was built, what the evidence covered, what remained open and why production was—or was not—allowed to proceed.
Buyers preparing an OEM, White Label or custom hardware production program can review ILEVA’s empty disposable AIO hardware options after defining the configuration and pilot evidence they need. Programs seeking stocked finished disposable inventory instead follow a different commercial route through bulk disposable vapes.
The buyer remains responsible for its formulation, filling process, finished-product evaluation, market-specific requirements and release decisions. No Blog framework replaces the current technical, quality, legal or regulatory controls applicable to the actual program.




