Two rollators, same drawing, same supplier, same shipment. On the showroom tile one stops when you squeeze; the other slides half a metre. Pads, wheels and lever travel match. What differs is what nobody wrote down: surface, load, squeeze speed, cable adjustment. That is the gap ISO 11199-2 brake requirements left when the 2024 amendment removed them.
Why does the same batch brake differently?
The variables that decide braking were never on the drawing. A brake is a friction pair: pad against rim, cable pulling a lever, spring pulling it back. The force that stops the wheel depends on how much pad area makes contact, how hard it is pressed, and what sits between the two surfaces.
Two units off one line can differ on all three without anyone making a mistake. Cable stretches over the first hundred squeezes, and the bite point moves. A pad that sat in humid air for a month has a different surface from one fitted last week. The unit the sales team carried to the showroom got its cable adjusted; the one still in its carton did not.
Now add the test side. Dry tile and wet tile are two tests. A 60 kg load and a 120 kg load are two tests. A squeeze applied at walking pace and one applied standing still are two tests. If the supplier's brake check names none of these, "it passed" describes one undefined experiment.
The 2024 revision left a hole where those numbers used to be. The brake requirement entry was removed from the standard text. The amendment states what it did — removal — and makes no judgement about whether braking is now better or worse. What actually moved is who writes the numbers: the standard used to, and now you do, in the rollator brake specification signed with the purchase order.
What has to be fixed before a brake test means anything?
Six fields. Miss any one and a second person cannot repeat the result — and being repeatable by someone else is the only property that makes a test useful in a dispute. The rollator test standard 2024 revision kept the framework whole and removed the brake entry; these six fields are what fills the space it left.
• Surface. Name the material and its condition. Glazed tile, dry. Glazed tile, wet. Concrete paving, dry. "Indoor floor" is not a surface.
• Load. Name the mass and where it sits. Loading to the declared maximum user mass, positioned where a person would sit or lean, is not the same as a sandbag dropped on the seat.
• Speed or slope. A moving stop needs an entry speed. A holding test needs a gradient. Write at least one; write both if both affect the outcome.
The first three fields describe the test. The last three decide whether its result counts.
• Actuation. Which lever, how far, how fast, held for how long. Squeeze-to-slow and push-down-to-lock are two mechanisms and belong on two lines.
• Failure criterion. What observation counts as a fail, expressed in millimetres or degrees, agreed before the test runs.
• Record. Who tested, which serial number, which surface, which load, which result. Without the serial number the record proves nothing about your shipment.
What slope and load belong in the specification?
Start from the declared maximum user mass, not from the lowest figure the scope allows. The scope begins at 35 kg; if your supplier declares 136 kg, then 136 is the number that matters. A brake that holds an empty cart tells you nothing about the unit your customer will lean on.
Then pick a slope. A 1:12 gradient — about 4.8 degrees — is a common built-environment reference for ramps and a defensible starting point for a holding test; treat it as a suggestion both sides agree, not a figure the standard hands you. On that slope, loaded, parking brake engaged, the unit should stay put for an agreed period — 60 seconds is long enough to expose creep — with the wheels marked and displacement read off a gauge rather than estimated by eye.
For the moving stop, fix an entry speed instead: 0.8 to 1.0 metres per second covers ordinary indoor walking pace. Mark the floor, squeeze at the mark, measure from where the wheels stop. Both figures are suggestions you are free to move. What cannot move is that they are written down, and identical for every unit being compared.
How do you test the brake when the rim is wet?
Wet is not a harder version of dry; it is a different mechanism. A dry pad on a dry rim grips through adhesion and abrasion. With water between them, the pad must first displace the film before it touches metal — and a wet wheel on glazed tile also loses the tyre friction that was quietly helping you stop.
So the wet test changes in three places, and all three get written down. Wet the rim and the pad, not only the floor. Keep load and entry speed identical to the dry run, so water is the only variable. Measure stopping distance on the same marked floor and record it beside the dry figure — the pair is the specification, the wet number alone is not.
Here is the practical reason to insist on it. A brake that passes dry and slips wet is not a rare edge case; it is the predictable result of testing only one condition. If the product is sold for shower rooms, poolside use or outdoor paving, the dry test is not your test.
What counts as a fail, and who decides?
Decide before the run starts. Three observable outcomes cover most purchase specifications: the wheel rotates while the brake is engaged, beyond an agreed angle; the whole unit travels beyond an agreed distance on the slope; or the lever needs noticeably more travel than it did when new to produce the same bite.
Each needs a number you choose and both sides accept. A quarter turn of the wheel. Fifty millimetres of travel during the hold. Twenty per cent more lever travel than the commissioning record. These show the form, and carry no authority of their own — the authority comes from being written into the contract before the first unit ships.
The second half of the question is who signs. A test result with no serial number, no surface and no load is an opinion. A record that survives a claim names four things: the unit, the conditions, the measured value and the person who ran it. If your supplier's brake report cannot supply those four, ask for the format to change before you discuss price.
Why does the parking brake need its own line?
Because the two functions fail differently, and a specification covering one leaves the other unmeasured. A running brake fails by fade: the pad heats, the cable stretches, grip declines with repeated use. A parking brake fails by geometry: a locking tooth skips, a frame flexes under load, or the lock bears on the tyre instead of the hub and the wheel creeps.
A wear test separates them. Lock and release repeatedly — five hundred cycles is a workable starting figure, again a suggestion — and after each block re-check that hold still meets the criterion you set. Then repeat the hold test on the wet rim. A parking brake that locks on day one and creeps after a season is the failure mode no delivery inspection ever catches.
The opposite error is worth naming too. A specification that lists only a wet-rim stopping distance leaves the parking function with no requirement at all — which is the state the amendment left behind, reproduced inside your own document.
FAQ
After the 2024 amendment, is brake performance still a requirement?
Yes, but it is now yours to define. The amendment removed the requirement entry from the standard text. The product still has to be safe, the manufacturer still has to assess the risk, and the buyer has to state what is acceptable — that last part is now written down instead of inherited.
Is ISO 11199-2 a mandatory standard?
No. It is a voluntary standard; what is mandatory is the regulation applying to your market. Certifying to it is one route to showing conformity, not a legal obligation by itself.
Can we just accept the supplier's own brake test report?
Only if the report names those six fields. A report reading "brake tested, passed" cannot be repeated by anyone, including the supplier six months later.
Our volumes are too small to justify a full test protocol. What then?
Test one unit per shipment on the wet rim at declared maximum load, and record the stopping distance. It costs an afternoon and turns an unverifiable claim into a number you can compare across shipments.
Do we need a test rig?
Not for the holding test: a marked ramp, a known load and a tape measure are enough. For a repeatable entry speed, releasing from a marked height is steadier than a person pushing by hand.
Sources and standards status
• ISO 11199-2:2021, third edition, published 2021. Scope covers rollators for a declared maximum user mass of 35 kg and above; walking tables fall under ISO 11199-3.
• ISO 11199-2:2021/Amd 1:2024, "Removal of brake requirements", published 2024, European adoption 2024-12. The amendment removes the brake performance requirement from the standard text and makes no statement about the resulting safety level.
• ISO 14971:2019, risk management for medical devices, the recognised method for justifying residual risk where a standard is silent.
• All test values in this article — gradient, entry speed, hold period, cycle count, displacement limits — are suggestions for a purchase specification. None is a value taken from the standard.
Gavren
Posted on
Top comments (0)