Sunday, 9 February 2020

Who Is Liable

I wrote about risk and liability from the designer's point of view back in 2013, but this week I thought it would be worth revisiting the subject. 

The reason for revisiting is because there are designers out there worrying about using continuous footways and cycle tracks over side streets in case they are sued. The fear comes from a concern that should there be a collision after doing something seen as innovative or unusual, then they are in the firing line.

I do understand their concern because let's be honest, people like to operate in their comfort zones and for a highway engineer, carrying on with a design approach which has decades of use behind it is very comforting. Let's look at four junction layouts (some of you will recognise this graphic);

The two on the left are really familiar and from a driver's point of view, one is invited to assume priority over anyone crossing the side street, despite what the Highway Code recommends that they only ever read when learning to drive. This layout being UK custom and practice means that people walking are told to be subservient through the design.

This approach was never critically reviewed before it was rolled out, it has its roots in history where at some point (which varied across the world), we started building footways (sidewalks) to keep foot traffic up out of the mess and filth of the street or to offer a little protection from carts. Over time, footways became commonplace, but in modern times they have evolved to be the bit that people stick to while walking leaving the relative high power of the motor vehicle to claim the space.

The top right image is also familiar and part of a generally modern idea of trying to give some priority back to walking. The layout may suggest that both people walking and driving could take priority, although it would very much depend on the person and the location. From that point of view, a potentially ambiguous layout and one where where someone making an incorrect assumption means the most vulnerable party in a collision comes off worse.

The bottom right is of a continuous affair which suggests that drivers may be guests, although we'd probably look to provide some tactile paving to assist visually impaired people who require assistance.

Across the North Sea, the layout is old hat and when correctly applied one can stand at a junction like the one above (Amsterdam) and see it work without any significant issues. It's a design that you do find in the UK, sometimes very old layouts, but it's never really taken off as an approach. There are some contemporary examples such as Walton-on-the-Naze (below) which was built before we started talking about continuous treatments as a "thing".

So, if we were nervous about doing something new, does this nervousness come from ourselves not having tried it or because nobody else has ever tried it? Clearly with lots of examples in the UK and elsewhere, being nervous has to be an issue for the individual. What's the way forward then?

At the heart of all of this, I don't actually think this is a radical design approach because if you think about the number of driveways we have in the UK which still look like the footway, then there must be millions of crossings of the footway every day with very little risk. The "new" risks will be applying a fairly unusual design to side streets and in the case of continuous cycle tracks, adding cycle traffic to a conflict point.

The key is to be clear in setting objectives, how we use evidence and how we record our decisions. The next part of this post will be to set out a methodology for developing a design for the unfamiliar using continuous treatments as an example. There are of course other ways to do it and you may pick a different order or other stages.

This is key. As a designer, how confident are you that you have enough knowledge of design, legislation and road safety to be able to adapt your to a new situation? What training and experience do you need to design this "thing" you are interested in?

It is a bit chicken and egg I know, but if you're design response is to revert back to bellmouth junction design then I wonder if you are competent to design that given how things are changing. Be honest with yourself and others. Get some help in and learn from that help. Being involved means that you'll gain experience and understanding.

Setting objectives - What are we trying to achieve? 
Pretty simple really, I think we're trying to provide side road treatments that assist people walking and cycling to continue to move smoothly with interactions with motor vehicles taking place in a managed way.

We are not providing absolute priority because that is only possible with traffic signals and zebra crossings, but we are providing implied priority which should be understood by all users.

Literature and policy review
Are there any studies into the thing you are trying to design? What were the conclusions? How old was the study? For continuous treatments, I found this study commissioned by TfL after a minute of Googling. There's some interesting discussion based on different design layouts. There is more out there if you look a bit harder. Unfortunately we have yet to have any UK guidance

Is there legislation covering the use of continuous treatments? Not specifically because they are not a "thing" defined in law as a zebra crossing is. Well I wrote about my view on the law and continuous treatments here - I am not suggesting citing anything I have written as policy or legislation, but there is a thought process to go through and you need that knowledge of design, legislation and road safety.

Robert Weetman goes into a lot of detail on what we should be looking to achieve in this blog post. Robert, like me, is at pains to point out that it is essentially for the designer to understand the risk, but there is nothing stopping us from reading articles and forming considered opinions.

Who else has done the thing you want to do? What worked and didn't work? Is there data available? If it is an idea from another country, how does the legal system translate to the UK? Does the design rely on administrative rules or does the design explain how to navigate?

For continuous treatments we have lots of UK examples now and in fact, I don't think we even need to look to any other countries. They are established and understandable treatments and you can easily visit them and see how they work. Ask the people who designed them about their experiences too.

What does the data say? Look at casualty data for the existing site and look at the traffic flows (people walking and cycling included). Does motor traffic dominate the junction? Will motor traffic flow need to be reduced? What are your expected walking/ cycling flows and will they outnumber motor traffic flows?

In some ways, not having formal guidance might push us to be more critical of our approach. If there was a table that said with X vehicles and Y pedestrians the approach is safe, then we risk people simply ticking boxes.

Data also extends to using video surveys, site interviews and repeating data gathering for the layout after it's built.

I don't mean chuck out a letter to the neighbours and see if they object. Discuss the proposals with people who might be most affected such as visually impaired people, wheelchair users, local schools and so on. Be prepared to adjust and refine your approach.

I was of the view that we can just push on with a footway across a side street until I was told that some visually impaired people want to know that they are in a location which would have a higher likelihood of encountering a motor vehicle and counting side streets help with navigation.

The solution is blister tactile paving and in fact, not having a radius kerb to cut paving into (because of the continuous design) means a tider job with the tactiles (no little pieces) which means a lower risk of them failing (look above at the example from Walton-on-the-Naze to see what I mean).

Risk assessment
I've also covered risk assessment previously, but it's a useful design tool. The aim of a continuous treatment is to give people walking and cycling a smooth experience across a side street and that is influenced by how each user interprets what they encounter.

For example, running a cycle track across the side street will invite a person cycling to continue because it looks like their space;

The photograph above (Blackfriars Road, London) shows a carriageway level cycle track which invites one to keep cycling through the design, but perhaps it's not as good as my earlier Dutch example.

The principal risk is a person cycling being hit by a driver turning out of the side street and so using a matrix I've talked about before, here is a quick assessment (just for illustration, I have no background data to rely on);

The likelihood is linked to the number of drivers emerging and cycle traffic flows. The consequence is injury, but drivers will be emerging slowly, so an injury is probably in the "medium" area.

So I reckon a medium risk of someone being hit. I've given this assessment because perhaps the side street is quite busy and for a 2-way cycle track, drivers don't always look both ways.

A way to reduce the risk is to reduce the conflict and so we could make the cycle track one-way to remove the problem of drivers not looking in the "unfamiliar" direction. Let's reassess the risk;

The consequence of someone being hit is the same, but getting rid of the unfamiliar direction reduces the likelihood and so using the simple red/ green/ amber approach we are in the green and so we can be reasonably happy.

But wait, we can't make the cycle track one-way, so do we go with design? There is no reason why we cannot because we are judging all sorts of things, but we might decide to adjust the layout of the side road to slow drivers down even more - we could put up a sign explaining a two-way cycle track, but a sign is essentially an administrative control which does help us in the long term.

Along a whole scheme, we might have fewer problem locations and so in the larger scheme of things where we are trying to transform an area we might note the risk, but judge it to be acceptable because because we don't consider the outcome of a collision as one which would result in a serious injury.

Road Safety Audit
Get an independent safety review undertaken, but it needs to be with people who understand the objectives and are competent in that type of design. Having people who do nothing but audit motorway layouts look at this type of layout probably isn't going to help.

The worry at the back of the designer's mind is being sued. You either work for a local authority or a consultant working for a local authority. You will never be working with total authority or autonomy because building stuff on the street has to go through some sort of process and at some point there is an agreement to proceed. In other words, it will never be 100% down to you.

But let's imagine the worst (and this is pretty hypothetical). A person cycling is hit by an emerging driver and hurt. The cyclist takes out a civil claim against you personally. Your defence will be that you are competent in what you are doing and that you've followed a structured design process which has taken into account the things I written about (and probably a great deal more).

It's a good idea to keep records of what you've done, but even if you haven't you will be able to explain the process logically (although writing stuff down is a good idea). You will be able to show how you have mitigated the risks through good design and understanding how your design works.

I really don't think that continuous treatments are innovative, but their design does change how risks might be viewed because they are not custom and practice. The good news is that the more of them that we build, the more that the people using them will become familiar with them. The tiny risk of being sued will be even smaller.


  1. Thanks - this came up with reference to disabled people at a council Active Travel group. Useful to have knowledgeable guidance.

    1. I get people are nervous, but a bit of thought and application will see them through!

  2. Thank you for this useful overview. I have a few comments:
    1/ Use of tactiles. You say above of continuous footways that "we'd probably look to provide some tactile paving", however in your blog 16/2/13 you said: "this layout has not got tactile paving as pedestrians have full priority". I would tend towards your earlier statement as long as design priority was established for pedestrians as it is on countless driveways & entrances to non-adopted developments and accesses to shops and businesses. The only issues seem to be giving pedestrians (esp. visually impaired people) confidence in 'design priority' & overcoming the issue of visually impaired people that use counting of side roads as their only way of navigating.
    2/ I was interested in the notion that "absolute priority is only possible with traffic signals and zebra crossings". I am not so concerned with signalised crossings and the 'green person's invitation to cross', but on non-signalised crossings of side roads I tend to think in terms of a hierarchy of pedestrian priority with Rule 170 of the Highway Code providing the lowest level of pedestrian priority and the 'design priority' of continuous footways providing the highest level of pedestrian priority. I would place the various forms of marked priority, including zebras, somewhere in between. Additionally there are designs and markings that intentionally or otherwise create ambiguous priority for some or all users.
    3/ Finally, when thinking of risk I think that it is important to consider 'risk for who'? In your examples you consider the risk to cyclists, but the same designs have very different risk profiles for drivers and pedestrians. The likelihood of the event occurring is the same for all, but a medium risk for a cyclist might be a negligible risk for someone protected by 1.5 tonnes of steel. Similarly a uni-directional cycle track might reduce risk for cyclists, but does it increase risks for pedestrians who will still cross in both directions, while the presence of a uni-directional track may make it even less likely that a driver turning left will look left?

    1. 1) is tricky, but if we are being asked for help by VI users and we are eager to roll the design out widely to rebalance things, then I'm happy to oblige - the Walton example is very neat and I think doesn't present too much of an issue with given drivers a cue. 7 years and I've learned along the way!

      2) It's a purist legal position in terms of legislation which gives priority. Of course, the continuous footway (if done well) give implicit priority because the administrative Rule 170 priority (strong guidance) is poorly understood. Certainly designs and markings can alter perceptions.

      3) Absolutely, we should assess for a variety of modes and potentially types of people. Drivers do tend to expect to see people walking as cycling is a rare thing - I imagine sites would change over time. I've read something about drivers pulling out on lorries because they are not used to seeing them - there is psychology at play here too.

  3. WRT 3. Risk assessments for highways are now in the DMRB (GG104), which defines 3 levels of safety risk by how common the feature is. Type A is the most straightforward and is for schemes for which there are many examples. Continuous footway schemes could be argued to be Type B, wherein examples exist but they are not as common as Type A schemes and Type C are truly novel schemes which are new to the UK network. Whilst the design engineer can carry out a Type A risk assessment, Type B and C require levels of input from road safety professionals due to the number of issues which need to be assessed. Like RSA this process is not meant to prevent innovation but support it with evidence.

    Andy R.

    1. It's a good point, although I wonder whether this helps people stay in their comfort zone?

  4. I would hope not. One thing risk assessment and a study of the data can do is produce counter-intuitive results, i.e. results which appear to go against 'common sense' or 'gut feel'.

    Last year we looked at, what at first glance appeared to be an insanely dangerous access on to the A1 from a housing estate, immediately prior to an off-slip. Previous advice had been to close it.

    However, upon digging in to both the injury collision data and the damage-only call outs to the HETOs, there was absolutely no evidence that this was in any way dangerous. In fact by far the greatest danger came from the stop-start nature of the flow on the mainline during the morning peak hour leading to shunts. Our conclusion, to be blunt, was that the access appeared to be dangerous so drivers (on both the access road and the mainline) took extra care when making any manoeuvre. Given the complete absence of collisions, this appeared to be the case both during the peak hour congestion or off-peak free flow. This gave everyone some confidence that appearancs were deceptive and the access could be kept open.

    (BTW, risk assessment allowed the alternative route - to be used should this access have been closed - to be assessed and rated alongside it. It's injury collision record was much worse and we would have been adding a significant volume of traffic to it).

    I don't know how Dutch collision statistics compare with our STATS19 data, that they could be used over here, but I'd be happy to write a qualitative risk assessment looking at any potential safety risks associated with continuous footways. I have a feeling that, when comparing them against the more normal side road layout with vehicle priority, they'd actually come out safer.

    Andy R.