Start with the decision rule
In wheelchair testing, labelling and the accompanying documents are not annexes to the report. They are inputs to the test. When the laboratory builds a test configuration, it works from the conditions you declare. When it decides which parts apply and how the sample is loaded, it works from the intended use you print on the rating label and write into the manual. Every boundary stated in the manual is treated as a pass/fail line that you set for yourself.
That gives you a working rule in two sentences. If the documents and the hardware disagree, you have a correction on your hands. If the documents overstate the intended use, you have additional testing on your hands. The first costs communication time; the second costs schedule.
Why this bites harder on wheelchairs
A wheelchair carries the user while it moves, so the safety boundary depends heavily on whether the user knows where that boundary sits. The ISO 7176 series gives disclosure of information its own part, and ISO 7176-15 is the one that deals with it. On the domestic route the GB/T 18029 series plays the same role. For powered wheelchairs going the European route, EN 12184 also pulls the content of the accompanying documents into the conformity judgement. What the three routes have in common is this: document content is not a recommendation. If it is missing, the verdict is non-conformity.
The common instinct is to run performance first and write the documents afterwards. On wheelchairs that order is expensive. The configuration and loading arrangement for stability and ramp-related work, and the test state for the control system on powered products, all draw their inputs from the documents. Until the documents are settled, the test state cannot be settled. If the documents change later, work already completed may have to be discarded as a batch. Changing something at the planning stage costs nothing; changing it once the sample is on the rig means requeuing.
First split: what has to be on the product, what can live in the documents
This is a call you make for yourself before submission, and getting it wrong drags a string of other problems behind it.
Markings on the product itself carry the information that has to be readable without the manual in hand: identification, the user conditions, warnings about where optional parts mount, and the electrical warnings on powered products. Put that class of information only in the manual and it will be treated as a missing item during on-site review, because a user in the middle of using the chair does not have the manual next to them.
The accompanying documents carry the information that needs room to be explained: assembly and adjustment, transfer and handling, limits for ramps and kerbs, maintenance and inspection, storage and transport conditions, how optional parts combine, and the specific scenarios behind each residual risk. That content will neither fit on the product nor be legible if forced onto it; crowding the label also dilutes the warnings that actually matter.
The way to decide is plain enough. Drop each piece of information into this picture: the user is standing next to the product with no paper in their hands. If the absence of that information at that moment could lead to injury or misuse, it belongs on the product.
Decision table: common situations and how to handle them
| Dimension | What you are seeing | How to handle it | Cost of leaving it alone |
|---|---|---|---|
| Label and manual consistency | The user conditions on the rating label do not match the body of the manual | Align on the stricter of the two, then ship the samples | The lab builds the test state from the stricter side; a delay at best, a full re-run of the performance work at worst |
| Wording of intended use | The manual claims suitability for demanding outdoor surfaces, but the sample is submitted in an indoor configuration | Narrow the wording, or add the corresponding parts in the outdoor configuration | Report conclusions and public claims diverge, and the gap is challenged again and again during registration and customer audits |
| Declared user conditions | No upper limit is declared for the user conditions | Add the declaration, then re-confirm the loading arrangement | The lab cannot fix a loading arrangement and the job sits on hold waiting for paperwork |
| Optional parts and configurations | Cushions, armrests, footrests and headrests are not listed in the documents | Build a configuration list and state which configurations this round of testing covers | Configurations that are actually sold end up outside the coverage and have to be added after delivery |
| Warnings and residual risk | Only generic wording along the lines of use with care | Rewrite scenario by scenario into actionable instructions tied to specific misuse | Use-error risk never closes, and the risk file and the manual contradict each other |
| Symbols and pictograms | In-house pictograms with no explanation provided | Add an explanation table, or switch to expressions with an established meaning | Reviewers cannot confirm what they mean and treat them as a missing item |
| Language versions | The target-market language version does not match the Chinese content | Translate from one controlled source file and keep version numbers in step | Different language versions state different boundaries, which is hard to explain during on-site review |
| Marking durability | Markings blur or peel after cleaning or ordinary use | Change the carrier or the marking process, then re-verify | Assessment of marking legibility fails and fresh samples have to be drawn |
The self-check sequence before submission
Walk this order once and you will stop the large majority of rejections.
Start with an inventory of inputs. Pull every statement in the manual that the laboratory could treat as a test input and put it in a table of its own: user conditions, intended environments, limits for gradient and surface, optional part combinations, and for powered products the control method and operating modes. That table becomes the baseline for every conversation that follows.
Then compare against the hardware. Take the table and walk it line by line against the sample. Where they disagree, decide whether the document is wrong or the hardware is wrong. Do not default to editing the document. The document is a promise and the hardware is a capability; shrinking the promise may cost you sales, while changing the hardware forces you to assess whether completed test items are still valid.
Next, converge the scope. Anything you are not confident you can verify should not appear in the intended use. Writing that the product suits all surface types is a way of adding test items to your own bill; writing the boundary out clearly makes the verdict cleaner.
Then align the warnings. Map the use-error scenarios identified in risk management one by one onto the warning entries in the manual. Where a scenario has no match, either add the warning or explain why the scenario does not apply. This step sits on the same chain as human factors and usability, which is a topic in its own right; the only thing worth repeating here is that a warning is a risk control measure, not a disclaimer.
After that, take stock of symbols. Every symbol appearing on the product or in the documents goes into an explanation table submitted alongside them.
Finish by freezing the version. The document revision you submit with the samples must be frozen and archived. If anything changes on the product side while testing is under way, assess first whether it triggers a document change, then decide whether the laboratory needs to be told. A great many retests trace back to an optional part quietly swapped during the test window without the documents following it.
Where manual and powered products differ
For manual products the documents concentrate on assembly, adjustment, transfer and handling, and on the division of tasks between user and attendant. Attendant-related wording is easy to overlook, and it is precisely the area where things go wrong in ramp and kerb scenarios. For the supporting test work on manual products, see manual wheelchair testing.
Powered products carry an extra layer: control method, switching between operating modes, switching between power-off and manual pushing, batteries and charging, and electrical limits on use. That content goes into the manual and also, in quantity, onto the product markings. Powered products also have one recurring problem: the description of how to switch between push mode and drive mode is written vaguely, and a mistake in the field means the chair rolls away. Wording like that needs to read as a sequence of actions rather than an explanation of the principle. For the supporting test work on powered products, see powered wheelchair testing.
When retesting is unavoidable
Three situations leave essentially no room to manoeuvre.
The first is an enlarged intended use. If the documents describe a use scenario that the submitted configuration does not cover, either delete the scenario or add the corresponding parts in that scenario.
The second is user conditions revised upward. The loading arrangement is fixed from the declared value; change the declared value and everything tied to load carrying has to be repeated.
The third is a change to a key configuration. Changes to the seating system, the drive arrangement or the control method are not wording adjustments at document level; they go through change assessment and the scope of test items is set again.
Most other situations can be closed out through correction, but keep one thing in mind: corrections take up schedule too. Document problems tend to surface in a cluster just before the report is issued, and editing at that point pushes delivery back.
On accreditation and the limits of a conclusion
It should be stated plainly that an accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope; it does not constitute a commitment regarding the outcome of market access in the target market. A test report answers the question of whether the submitted sample meets the requirements of the cited standards under the stated conditions. It does not replace registration review, quality system audit, or a customer's own acceptance judgement. The same applies to labelling and manuals: a report can show that document content satisfies the corresponding requirements, but whether an external claim holds up depends on whether the claim itself goes beyond what the report covers. If you want to work out which route you are on and which parts you need to cover, start with standards and test items.
If you want help setting the scope
On documents and labelling, the earlier you get involved the less it costs. Spend half an afternoon at the planning stage aligning intended use, user conditions, the optional part list and the warning entries, and there is usually no rework afterwards. Wait until the samples arrive and any move you make means requeuing. Before submission we can run through the input items in your manual and markings, flag which statements will expand the test scope and which information has to sit on the product itself, and set the test item list and schedule from there.
To talk through a specific product, call +86 132 4819 8029, or send the product documentation through request a quote and we will come back with proposed test items and timing for your target route.