Speed cushions

Trial Conclusions

The on-road trials have shown that speed cushions are suitable for use as speed control features. They do not generally cause excessive discomfort to passengers of large buses, or excessive discomfort/delay to fire service vehicles. The discomfort or delay for passengers in mini and midi buses and double rear wheel ambulances is likely to be higher than for the larger buses, particularly at the wider cushions.

Careful consideration needs to be given to the most appropriate cushion design for a particular road, and for the vehicles likely to use that road. Design parameters established from the trials were as follows:

Side ramps gradients

Not steeper than 1:4.

On and Off gradients

It is recommended that on and off gradients should not be steeper than 1:8. The research suggested that where larger buses operate, the use of both of in situ cushions with 1:6 gradients and a proprietary brand of cushion using a curved ramp with an average gradient of 1:5, did not cause any great discomfort or difficulty. However, in some circumstances with the 1:6 gradient cushions, grounding by other vehicles could occur. It is likely that these steeper gradients could be permitted in future, where they are appropriate to the type of vehicles likely to be using the route.


80mm should be considered as an absolute maximum. There have been reports of grounding occurring where cushions have in error been installed at heights greater than 80mm. It is recommended that 75mm should be specified as the maximum height for any cushions to be constructed in situ. A lower height of 65mm may be appropriate for narrow cushions.


Overall lengths varied between 1.7m and 3.4m, but were generally around 2m to 2.5m. Monitoring of the trials did not indicate any great difference in terms of discomfort. However, it has been suggested from off-road trials elsewhere that cushion lengths of 3.5m, with a width of 1.6m, height not greater than 65mm, and on and of gradients not steeper than 1:8 may cause less discomfort to passengers of mini and midi buses. On-road trials have not yet been carried out to substantiate this claim.


A wide cushion (about 1.9m) will have a slightly higher speed reducing quality than a narrower (1.6m) one. However, whilst the wider cushion would be generally appropriate for fire service vehicles, it may not be acceptable by bus operators because of the increased discomfort. Very narrow cushions (1.3m or less) are best used in conjunction with a chicane or narrowing feature, when the effects tend to be more psychological than physical.

Speed cushion dimensions

Speed cushion dimensions

Longitudinal spacing

This should generally accord with the requirements of the Highways (Road Humps) Regulations. Cushions cannot control speeds as much as standard road humps, and complete reliance on them in a 20mph zone may not achieve average speeds of less than 20mph. For the narrower cushions (1.6m), spacing in the region of 60m to 80m would normally be required to ensure 85th percentile speeds of 25mph of 30mph. Closer spacing than this should produce 85th percentile speeds of less than 25mph along the road, offering greater encouragement to drivers to maintain a steady speed which then allows the cushions to be negotiated without discomfort or heavy braking, and with consequent environmental benefits.

Transverse gaps

Minimum gaps of 750mm between the base of a cushion and the kerb, as well as between adjacent cushions, are appropriate to accommodate cyclists and motorcyclists, though 1m is an ideal width. It was found that most cyclists and motorcyclists naturally tended to follow the nearside gap. It is important therefore to ensure that this is reasonably level, and the cushion is not located adjacent to a gully. Where parking occurred cyclists could not take advantage of the nearside gap. In some instances parked vehicles straddled the cushion, reducing the gap available on both sides. As a result cyclists either rode over the cushions or rode to the other side of the carriageway to utilise the gap there. It is not unsafe for cyclists or motorcyclists to ride across cushions of designs agreeing with those specified above, though it is not particularly comfortable. Special care is required in design where three cushions transversely in line are used and regular parking occurs.


Should be in accordance with the requirements of the Road Hump Regulations.

Three abreast speed cushions

Three Abreast Speed Cushions

Set of four cushions

Set of Four Cushions

Cyclist using nearside gap

Cyclist using nearside gap

Indiscriminate parking

This can be a reason for buses being unable to centrally straddle cushions. In these cases it may be preferable to consider using cushions positioned adjacent to build outs or pinch points to prevent parking within the vicinity of the cushions. Another alternative would be to have a central refuge with a cushion either side. However, if the refuge is likely to be used by pedestrians to cross, the cushions will need to be positioned before or after the refuge, usually offset one to the other, to allow pedestrians to cross the road at one level. Groups of three cushions spaced across the road allow at least one cushion to be available to be straddled, limiting the impact of parking directly over the cushions.


Within the design parameter of having forward gradients of about 1:8, mean speeds at the cushions were about 19mph for the narrower cushions (1.6m), and about 14mph for wider cushions (1.8m to 1.9m).

Before the schemes were installed the mean speed of buses was about 2mph to 8mph slower than the mean speed of cars. After the cushions were installed the mean speeds of buses were generally similar to or slightly faster than cars.

For emergency vehicles the mean "urgent" speeds for the fire appliance were higher than that at 75mm road humps, whilst for the ambulance the mean speeds at the wider cushions were similar to that at road humps.

Double pair arrangements of cushions resulted in mean speeds about 1mph less than single pair arrangements. Three cushions in line transversely had mean speeds about 2mph higher than single pair groups.

Single cushions at single lane pinch points (with or without priority signing) resulted in unopposed speeds about 1mph higher than single pair cushion arrangements.

The mean unopposed speeds at the cushion groups which had a "set of three" or a "set of 5" cushions within a narrowing were about 3mph lower than the single lane pinch points where only a single or double single cushion arrangement had been used.

Set of three cushions

Set of Three Cushions

Set of five cushions

Set of Five Cushions

Cushion combined with build out

Cushion Combined with Build Out

Cushion combined with build out (2)

Cushion Combined with Build Out (2)

Driver behaviour

Overall about half the car drivers straddled the cushions. The remainder generally let their wheels ride up onto the cushions on one side only. At other sites where speed cushions have been installed in pairs, it is reported that some drivers choose to drive in the central gap. This is not entirely desirable. particularly where frequent opposing traffic occurs. Obviously having a central refuge would prevent this but would add to the cost. Ensuring the gap was not wider than 1.1m and/or placing hatched markings between the cushions might discourage this action.


Where cushions are located near to bus stops, care should be taken that parked vehicles in the vicinity do not prevent buses straddling the cushion. At junctions there should be sufficient space for a large vehicle to complete its turn and straighten up before crossing a cushion.


The presence of a speed cushion can result in a substantial drop in traffic noise levels. The maximum vehicle noise for light vehicles can also be reduced, as a result of light vehicles slowing down at the cushion.

It is possible that some nuisance could be caused, due to variations of noise that occur between the cushions and at the cushions. This suggests that the spacing between cushions should be chosen so that constant traffic speeds are encouraged along the route, resulting in less variation in noise. Further research has suggested that 50m spacings of cushions results in very little variation in noise levels, with average speeds along the route of around 20mph. The conditions were too variable to be able to form comparisons between the noise generated by different types of cushions.

For related documents, pages and internet links, see the column on the right.

Back to contents

Back to top