Saturday 28 February 2015

Tricky Tactiles

It's been an interesting and varied week for me (that's why I like being an engineer), but I am posting about one thing on my mind; the tricky subject of tactile paving.

Blister tactile paving - this one at a zebra crossing.
Tactile paving is ubiquitous on our streets, although those of us in the game and some users are the only people who know what it is for. I was at a workshop with Urban Design London earlier this week where we spent the day talking about tactile paving (living the dream!). 

We were fortunate to have Dr Kit Mitchell with us who was involved in the original research into using tactile paving when at the Transport Research Laboratory - specifically blister paving which is the square grid pattern of "bumps". It was proposed as a way of dealing with the important issue of providing flush kerbs at pedestrian crossing points. Of course, with flush crossings comes a significant safety risk for visually impaired people and so something was needed to show that people were at the edge of the footway. The blister paving was born. It is not a panacea as the blisters can be painful or uncomfortable for some people to walk on - the I'DGO information sheet on the subject is well worth a read.

Theory and practice are often different.
These days we have managed to land ourselves with 7 types of tactile surface/ paving which are often used incorrectly and potentially dangerously! "Guidance on the use of tactile paving surfaces" is the current design guidance which has been in place since 1998 and I would recommend that it is read and digested by designers and campaigners alike as it gives far more detail than I will here. 

The guidance is not without its problems as the layouts tend to be "perfect world" with "real world" examples of how one should cut tactile paving flags properly to fit curves. Maintenance is mentioned, but there is no practical information on how tactile paving should be laid - this would be good for designers, installers and maintainers (more on that later).

There is no legal duty to provide tactile paving, but given the need to ensure our streets are accessible to all as implied by the Equality Act 2010, it should definitely be used where visually impaired people are likely to be walking towards dropped and other flush kerbs at a road edge. There are other layouts in the guidance each giving a slightly different message and certainly, some uses are also of real benefit in terms of the safety of visually impaired people.

I have mentioned the square-grid blister paving which is used wherever people might be crossing the street and the kerb is flush with the road - it could be at a dropped kerb (ramp) or where the road (or a cycle track) is brought up to footway level. Blister paving should always be provided in pairs as we are sending the message to people with reduced or no vision that they are about to enter a road. For zebra crossings and signalised crossings, red blister paving is normally used and normally in an 'L' shape as in the first photo to indicate the push button being on the right for a signalised crossing. The "stem" which runs away from the crossing at right angles is to assist people in locating the crossing position. At other crossings (uncontrolled), a colour contrasting with the surround paving (other than red) is recommended, although people often use buff or light grey. 

Corduroy paving used at steps by the Cutty Sark in
Greenwich, London. The row in the foreground is at the
top of the steps, but there is none at the bottom which
leads into the road. In the distance, there is corduroy top
and bottom. In this layout, there should be two rows
(800mm) as people can miss them when stepping over.
Next we have the "corduroy hazard warning surface" which is a series of "bars" with round tops. It is intended to mean "proceed with caution" and is best used at the top and bottom of steps, but is also used at the bottom of a ramp leading up to on on-street "light rapid transit system" (trams etc), to warn that people are about to walk onto a railway platform, at railway level crossings and where a footpath joins a shared-use (unsegregated) cycle track. You can see why blister paving would be dangerous here!

Off-street platform edge at Cannonbury Overground
station. It is the single line of buff paving.
We then have platform edge (off street) blister paving (to be complicated) which is used to warn of the platform edge at railway stations. The blisters are offset in each row to form a triangular grid (rather then the square grid used for crossings). Only laid in a single line, this paving is used a white platform edge line and often a yellow line back from the edge. There is less risk of people stepping over this type of paving (so 800mm not needed) as people in a station will be taking care to find the edge.

From inside the tram, we can see the row of lozenge-
shaped bumps for the LRT tactile surface.
Following the rail theme, we next have a similar approach for railway stations, but this time we have "lozenge" shapes for on street LRT stops (trams etc). It is different from the off street blister paving because these stops are on the street and the railway blister could confuse people into thinking they are about to cross a road. However, as with railway stations, visually impaired people will be there by choice and so one row is needed.

Old Shoreham Road, Hove. Ladder is to the left for the
pedestrian side and tram to the right for cycling. A bit
confused here as people are entering a bus stop area

which is nominally going to be shared.
The 5th type of tactile paving has a series of flat-topped "bars" on them (similar to corduroy) and is used on shared-use, segregated cycle tracks. When the bars are laid across the line of travel (known as "ladder"), this shows the "side" for pedestrians, giving a rumble to people cycling over them. When laid with the direction of travel (known as "tram", this indicates the cycling "side", although the bars can catch bicycle tyres, worse when they are wet. The paving is used at the start and end of these segregated tracks and as repeaters.

A raised delineator strip on CS3 at Beckton.
The ladder and tram paving is used with the 6th type of tactile paving - the raised "delineator strip". It is essentially a small kerb or even a raised road marking (to the same dimensions) used to help visually impaired people to keep to the pedestrian "side". Of course with protected cycle tracks (stepped, or kerb-separated) the ladder/ tram and delineator are not needed at all!

Guidance path surface - from the
We finally have the "guidance path surface". Used in the line of travel, this paving type is easily confused with tram and indeed corduroy; but the shape is slightly different. It is intended to help visually impaired people navigate open spaces where building lines or kerb lines are not available by either following the paving or for cane users, using the ridges to follow with their canes. It doesn't often get used, but could be useful to guide people at public transport interchanges for example.

I have briefly set out the "official" advice on using the different types of tactile paving, and clearly, there is lots of room for confusion. As I mentioned at the start that I was at a workshop and it was to discuss in general terms what people thought was good and bad about tactile paving. Having engineers, urban designers and visually impaired people (plus some groups representing them such as Guide Dogs) was a great idea as I think everyone learnt a bit that day. 

There was some discussion on "non-standard" use of tactile paving for level surface (shared space schemes) which mark a nominal edge of footway to assist visually impaired people. For example, Exhibition Road in London uses corduroy paving. Of course, being a shared space scheme, one questions the need to demarcate anything as all users are meant to be sharing, right?

I will be blogging about shared space in the future and so in the case of Exhibition Road, I will simply state that groups such as Guide Dogs were extremely concerned about the concept (and remain concerned more generally), although the Royal Borough of Kingston & Chelsea belatedly engaged with those with concerns and did some testing. Personally, I am not convinced.

Oxford Circus, London. Note the tactile paving cut to
follow the curve of the kerb.
There was also discussion about some proposed changes to the guidance which the Department for Transport will consult formally on soon. This included the following points;

  • Flexibility to allow the back of a row of blister tactile paving to follow the curve of the dropped kerb (for crossings), rather than being straight, although the paving would need to be 800mm deep to make sure visually impaired people don't step over it.
  • Replacement of the requirement for blister paving at a controlled crossings (zebras, signalised) to be red, with a requirement for at least a 50% contrast ratio with the surrounding paving
  • Introduce a requirement for the boundary between carriageway and footway to be demarcated with tactile paving wherever they are at the same level

The first point in theory reduces the amount of paving area needed and in terms of how a street looks, follows the less is more idea without impacting on visually impaired people. I suppose this is fine if the paving is well-detailed and installed, but it will introduce more cut pieces of paving which are notorious for failing - a view expressed by quite a few people.

The second point is driven by aesthetics to some extent, although red tactile paving surrounded by red paving doesn't give contrast. Many people (including users) felt that contrast and the 'L' shape (when it should be used) was more important than the colour; frankly, if users are happy, then so am I.

The third point would cover any flush kerb situation. I suspect it was primarily driven by shared space schemes not providing some kind of demarcation between the "footway" and "carriageway" (which is oxymoronic), but would apply to speed tables and dropped kerbs more generally. I sometimes see speed tables which are poorly designed or installed which are flush, but with only a small area covered by blister paving (not necessarily on the desire line). Shared space issues to one side, this is a good idea.

An Israeli push button - note the arrow on the top
showing the direction of travel.
There was also a discussion on a suggestion to add a tactile arrow to the push button boxes to show the crossing direction with an example from Israel being given. Finally, there was a suggestion to provide a rotating tactile cone on push button boxes when used on both sides of signalised crossings (convention is only to provide on the right hand side). Although users are using the 'L' shape to find the push button, on wide crossings, if people found the push button on the left, why not have a cone. Seems to make sense to me. These part of the discussion prompted concerns that too many staggered signalised crossing are difficult to navigate by visually impaired people and straight through crossings were preferred.

We then spent some time debating how tactile paving is used with cycling schemes. Many people felt that mixing people walking and cycling was not desirable and separated infrastructure was required which was a bit difficult to get away from when the discussion was about using tactile paving to deal with segregated and non-segregated shared-use cycle tracks. There was less consensus here and probably not worth going into detail about to be honest.

I will be looking out for the consultation by the DfT as I want to raise the need to provide some design guidance on detailing and construction to avoid some of the poorly built layouts which require constant maintenance. By this I mean giving practical drawings showing how to properly cut tactile paving units, how to avoid diagonal cuts (which get used when ramp gradients are not properly set up) and how to bed the paving units to stop them breaking up (lay on concrete or mortar on concrete, never sand). Happy paving!


  1. I spoke to someone at Kensington and Chelsea council who confirmed that Exhibition Road is not "Shared Space" but "Single Surface". As such there are still separate defined areas for vehicles and pedestrians delineated by the tactile paving, Pedestrians must give way to vehicles when crossing between the pedestrian areas either side.

    1. The scheme was sold as shared space - look at the contemporary news reports, this just sells people out. Low and behold, Council's website talks about "pedestrian zones" and "lanes for traffic" -

  2. Actually, I'd be very interested to know about the discussion regarding shared use cycle paths. Our council rarely if ever gets the tactiles right, but they never use the median strip.

    1. The discussion was along the lines of sharing is a problem for visually impaired people and so separation would make "them" feel safer and if properly designed, there would be (virtually) no need for cycle tactile paving. But, given there is shared (segregated and unsegregated) properly laid out tactiles will help visually impaired users.

      But, agreement was not universal!

    2. Generally speaking we see both separation for cyclists and pedestrians and tactile paving used in the Netherlands. For instance, the tactile paving can be seen quite clearly in photos of the centre of Assen.

    3. David, thanks - you have a really good photo here which shows a tactile route along the street, with a guide taking visually impaired people to the crossing point (which also has tactile to show the cycle track edge).

    4. There should never be tactile paving to say you are entering an unsegregated shared use area. This is something that has been misinterpreted for donkeys years. Finally though people are catching on and doing it right.

    5. Yes, we've over complicated and misinterpreted it. My own view is that kind of tactile should be used for steps/ slops rather than saying "you are entering a shared area"

  3. Quote "Tactile paving is ubiquitous on our streets, although those of us in the game and some users are the only people who know what it is for."
    Actually, just as designers make mistakes and some don't seem to know the differences, I'd be interested to know if anyone asked the question how well the differences - both between different types of tactile paviour and between the differing layouts - are known, and detected, by the 'average' blind or partially-sighted person (realising that such a person doesn't really exist).

    Quote "The guidance is not without its problems as the layouts tend to be "perfect world" with [no] "real world" examples of how one should cut tactile paving flags properly to fit curves. "
    Nor does it help that, when I last needed advice on layouts for example, at least two of the contact numbers/organisations in the back of the guidance appeared to no longer exist.

    Andy R.

    1. Even the users at the meeting had some disagreement, although the advocate groups felt much was down to how things are publicised. Yes, it is old and a printed document - it needs to be web based and kept up to date.

  4. "When laid with the direction of travel (known as "tram", this indicates the cycling "side", although the bars can catch bicycle tyres, worse when they are wet." - Yup. It happened to me so many times on the Portway (A4, by Cumberland Basin) in Bristol that I started going over the pedestrian side of the paving so I didn't run that risk. Maybe my tyres are just the wrong size...?

  5. I'm led to believe that work will soon start on a new EN standard on tactile paving. Unlike the stalled draft of a few years back (DD CEN/TS 15209:2008) that was essential a library of different patterns used around the EU, this one will apparently attempt to codify use too. It will be interesting to see how this will be reconciled with the requirements of the UK Equality Act 2010 and what occurs if there is a difference of opinion between the UK (via its indigenous committee) and CEN. Anyone else heard about this?

    Pleased to see you picked up on the issue about poor design and delineation of raised tables. Ignored by far too many authorities.

    Finally, it will be interesting to see how any guidance in the new tactile doc on delineating 'flush' surfaces will be reconciled with the addendum to MfS published by the DfT on behalf of the NFBUK in Dec 2013 (Access for Blind People in Towns - SS1401). That rather implies that kerbs with upstands lower than the 'standard' 100-150mm (their definition) have a significant negative impact on blind and partially sighted people and, as per the Equality Duty, should not therefore be used without an EqIA etc... That'll be 90% of kerbs in London requiring some new angled type of tactile to their length then :). Interesting also that this clashes with the 60mm finding from the 2009 UCL/Guide Dogs research on 'Effective kerb heights for blind and partially sighted people'.

    1. The whole thing has got messy, and while I wouldn't want guidance to be interpreted at standards, it doesn't replace good design; and of course, good design is from the user's point of view.

      For me, the greatest area of risk must be where people are walking along and can go into a road without knowing - junction mainly and so in practice, level surfaces may be less of a risk than is made out (including by me) so long as there is a properly recognisable tactile surface, 800mm deep. Such a surface is potentially more use than a kerb which in theory could be stepped over. But, over course, guide dogs use a kerb and kerbs are good to discourage overrun by vehicles.

    2. Sounds like an argument for kerbs plus tactile :).

      I see the whole 'death of the 25mm kerb upstand' as one of the more unfortunate unforeseen consequences of the great shared surface delineator debate. It causes no end of design grief at raised tables for anyone who acknowledges it - 25mm being the previous commonly accepted solution (though see below). Increasingly you were seeing boroughs attempt to make local use of corduroy to delineate corners and extensions beyond the blister waiting areas, but these were invariably messy jobs (we really need some smaller unit tactile modules to ease laying around radio if this sort of edge delineation is to be the future). Also, most were done only 400mm wide. Having said that, using 25mm to the corners of tables was only ever acceptable in combination with guardrailing. If you read the extisting DfT tactile guidance it emphasises the fact that this or some similar vertical barrier should always be used in those locations to prevent the visually impaired mistakenly wandering into the road. Somehow that seemed to be completely overlooked in the great rush to de-guardrail (not that I'm a fan of the stuff). Don't think I saw it considered in any guard rail removal assessment methodology. Should we be putting it all back in given Ali vs LB Newham? You begin to lose the will...

      The sooner new DfT tactile guidance comes out to make sense of the mess the better. I only hope they learn the lesson of Ex Rd and the shared space LTNs and properly involve the Disability groups as per the Equality Act so that it has proper weight and boroughs can use it with confidence.

  6. "When laid with the direction of travel (known as "tram", this indicates the cycling "side", although the bars can catch bicycle tyres, worse when they are wet." - Oh, totally. This stuff is terrifying - I usually go across these on the 'ladder' side instead, after losing count of the times I've nearly come off (that feeling of the 'tram' grabbing your wheel and making you skid is horrible!). IMO, guidance should be for it to be fitted the other way around (as pedestrians don't have wheels).

    1. It would make sense to swap them, but then all existing will become confusing!

      I think the aim should be to design them out, but when they are used, only where people cycling are completely on straight lines.

  7. I don't actually understand the logic of the cycle/footpath pattern. I had heard that the basic philosophy is that lines are in the direction of travel (of the visually impaired person). E.g. they run parallel to kerbs to guide people safely along the kerb, but when you approach a kerb they lines will be perpendicular to symbolise a "barrier". With this logic, the lines should be in the direction of travel on the pedestrian side, not the cycle side, and perpendicular on the cycling side.

    1. It would make more sense and the lines running across the cycle line of travel would be safer. Of course, if we design cycle tracks properly, we don't need tactiles at all!

  8. What term would you use for a tactile surface, which is used as a deterrent, to stop people walking into a certain zone, people can still walk on it but it us uneven and uncomfortable to walk on.

    1. Literally known as deterrent paving - there are two main types - anti-pedestrian (various styles)

      and anti-vehicle (less seen)

  9. Noting your mention of Kit Mitchell, I wonder if there might at last be an answer to over a decade of my asking for the report, and research that TRRL carried out for CLT3 of DETR over 20 years ago. This established the material, and its shape (height 5mm +/-1mm : rounded corners : 70mm-30mm-70mm-30mm... pattern : aligned ridges) such that the inherently UNSAFE arrangement could be made reasonably safe for cyclists. This also included NOT installing it where cyclists would be turning a corner, and defining the length to be installed (frequently not complied with). I really do need to get the report and research detail, as it links in with standard also established for level crossings (max vertical step between abutting elements 5mm), tram track installation (Tramways Act 1870 - 6mm current ORR(UK Tram spec +/- 3mm), paving slabs (6mm+ = trip hazard), dropped kerbs (+6mm limit above carriageway pavement).... etc

    With a glorious irony as well the transverse pattern is a pain (literally for some) to use in a wheelchair, or with a pram, do these users often cross over to the cycling side. How on earth did such idiocy slip though the basic design process?

    The hazards are made worse by the use of materials other than a pressure cast concrete. Lethal when wet are specifications like the brass (& smooth) dimple studs set in to York/Caithness stone paving slabs in a number of 'heritage' areas where this "naturally slippery when wet" stone has been specified as a paving material. All I can do I guess, is to slowly build that portfolio of the bad & dangerous, and put on record, the liabilities that councils/clients have from their failure to specify correctly, or manage their contractors delivery to design, when the inevitable happens.

  10. Regarding the second image "Theory and practice are often different", how do you solve this problem ? How can you avoid the cut pieces and also maintain the direction of travel when the pedestrian crossing are located at a curve ?

    1. Curves are awkward because if you are right on the corner, then the direction of travel could be read as two directions, but if you inset into the side street too much then you're off the desire line and maybe hiding people round the corner. For the cutting in, that's easier because a skilled roadworker will have set the thing out and thought about the cuts so only small pieces are thrown away - you can see that if you look closely at the second image.