Something crystalised in my mind a little while back as I contemplated crossing a side street at a signalised junction - the idea of double standards in traffic signal design.
The idea had been in my head for some time, and it developed from my work with Robert Weetman on the design of side road junctions. Consideration of signalised junctions wasn't part of our scope per se (as we are looking at priority junctions), but we did recognise that they are still a tool that would be needed within the overall design of a road network.
The actual junction that finally nudged me to write this is the junction of the A12 Eastern Avenue with Aldborough Road North & South. The A12 is a Transport for London strategic road and a dual carriageway with a 40mph speed limit. It's a two lanes in each direction affair, but flares out to four approach lanes (including right turn filters) and three exit lanes. This junction is very much about stuffing motor traffic though.
The side streets have two lane approaches and single lane exits. The northern arm has a left turn slip road which curiously has a zebra crossing over it, but which uncontrolled for pedestrians over the northern arm (see above). You can also just see the staggered pedestrian crossing over the main carriageway (the eastern side of the junction).
The crossing of the main carriageway is pretty standard with people crossing the southern half with a green man when main road traffic is held and crossing the northern half when the side roads run with traffic turning from south to east being held at an internal stop line.
The western side of the junction has a period subway to cross the main road which is as attractive as you might imagine (above). And like the northern arm, the southern is uncontrolled (below).
The zebra crossing over the left turn slip road is a little unusual, but this type junction with a green man crossing on the main road crossing, but nothing on the side roads is very common and essentially there for motor traffic flow.
Because this is is a flow-focussed design, having single stage pedestrian crossings over the side roads would require an "all-red" to motors which isn't apparently acceptable because it takes too much capacity from driving. You do find junctions which at least provide green men on all arms with multi-stage crossings even if still flow-focussed. Barley Lane to the east is a good example, but this will always be a space and motor-capacity debate.
You might think the double standard here is just that drivers get most of the capacity at the junction. That is of course true, but the issue goes far deeper than that when we think about this in terms of pedestrian-friendliness. There are a set of factors which affect the pedestrian experience which I will restate here:
- Background vehicle speed
- Through traffic volume
- Turning speed
- Turning vehicle size
- Turning levels and complexity of traffic movement
- Visibility character and crossing distance
- Presence/absence of physical barriers to pedestrian movement
I would recommend you go and read Robert Weetman's blog post on "factors and features" if you want a more in-depth explanation of these, but I will refer to some of them when applied to crossing the side road arms of a signalised junction.
Signals are there to control flows of traffic, including pedestrians and cyclists. In the context of junctions like Aldborough Road, signals enable side road traffic streams to gain access to, and the ability to cross the main road. Traffic signals also provide time within their set-up to give crossing opportunities by holding motor traffic streams to allow people to cross, even if that isn't specifically with a pedestrian signal, although you almost need to know the sequence to stand a change.
It is important to understand that traffic signals are not safety devices, because they rely on people obeying the rules, rather than being inherently safe. I am not suggesting that traffic signals are unsafe, but they are a rules-based feature that can have catastrophic impacts from non-compliance, often with those carrying the least energy coming of worst.
If we wanted to create a junction which is safer than traffic signals, then we would be grade-separating everything which is complex and costly in established urban areas. There are things we can do within traffic signal arrangements to mitigate the risks, but this will always be in balance with capacity and efficiency. Back to my junction example where pedestrians are not given a green man to cross, notwithstanding the point that signals cannot be perfect.
If you start to think about the factors above and how they might influence conditions for people trying to cross the side roads, you soon realise that the issues are not there in the background more generally as you would have with an ordinary side road junction, they become concentrated because of the traffic signals.
While a crossing pedestrian won't have to judge traffic gaps coming from the full range of turning movements at once, the underlying volumes of passing traffic on the main road are ever present. My example has a refuge in both side roads which means crossing traffic leaving the side road will at least be simpler as it is only one stream from one direction which is often directly held at a red signal. Crossing the second stage is much more difficult because the signals will operate to provide an almost constant stream of traffic, but with their directions ever changing as the junction runs though its cycle.
It is arguable that crossing from the refuge might be simplest because one can see oncoming traffic which makes judging gaps easier than crossing to the refuge. Younger and older people have more trouble assessing speed and therefore gaps, and of course visually impaired people and people who cannot move quickly will also have a hard time crossing. There will also be signalised junctions without refuges and these could be even more difficult to cross.
I have been around the houses a bit here, but I will at this point suggest that signalised junctions can never be pedestrian-friendly as defined by situations where drivers are likely to yield to pedestrians crossing the side road. This is because there is one additional factor - the green traffic signal.
Green traffic signals for drivers (and cyclists) are an invitation to proceed but only if your exit is clear (unless waiting to turn right) and in any case, caution should be exercised. The problem is, however, that some people treat a green signal as affording total priority and they behave in that way. I am sure we can all provide anecdotes of drivers moving from a green signal using their horns at people trying to cross in front of them!
I'm being a little unfair to lay the behaviour we see completely at the feet of drivers (and cyclists) though, because there is also the ideas of social pressure and cultural pressure. Social pressure is where people react to a green signal in a way where they don't want to hold other people up - horns being sounded by their impatient followers. Cultural pressure comes from a long history of green meaning go (as different from meaning proceed with caution). This might come from learned behaviour from people's peers, what they see in the media or perhaps even how they were taught.
This stuff is hard to deal with in design terms, and one of my favourite examples is where there is the need to give way shortly after a signal such as with the left turn slip road in the photograph above. I guarantee that there will be people missing the give way because they are just reacting to the green signal.
As an aside, a green man signal (more properly called a "pedestrian signal") is an invitation to cross, but in a rules-based system, the safety of someone crossing relies on everyone else behaving. This is why I will always push back when people describe traffic signals being safety devices. The other thing to note is that there is no legal force to disobeying a red man in the UK because (in theory) people can cross where and when they like.
Amusingly Rule 21 of the Highway Code states:
"There may be special signals for pedestrians. You should only start to cross the road when the green figure shows. If you have started to cross the road and the green figure goes out, you should still have time to reach the other side, but do not delay. If no pedestrian signals have been provided, watch carefully and do not cross until the traffic lights are red and the traffic has stopped. Keep looking and check for traffic that may be turning the corner. Remember that traffic lights may let traffic move in some lanes while traffic in other lanes has stopped."
It's amusing because the junction in my example is very much the second two-thirds of the rule which is actually telling people that in fact they are fully responsible for crossing the road and wishing them luck in doing so; this starts to pull at the double standard thread.
When one undertakes design work for a junction, the general approach is to separate conflicting traffic movements in time and sometimes space. While "traffic" is very much all road users, including pedestrians and cyclists, when we get to the actual detail, we find pedestrians are treated differently to drivers and cyclists because they are not vehicle-based (a cycle is a vehicle in this context). This may just be a result of the ability of people to cross where and when they like (in theory) as accrued as right from antiquity.
Each traffic movement will have a "phase" and when a series of phases run together that is called a "stage", but a stage cannot have conflicting phases. That is why main road traffic moving ahead doesn't run on green at the same time as a side road joining or crossing it - we'd obviously have carnage on our hands.
Where we have phases for a specific movement, we'll see green arrows used and if a driver (or cyclist) sees one of these, then they will have an exclusive right of way. This means we can do clever things like having left turns into and right turns out from a side road running together.
There is a bit of an exception where two opposing traffic phases run together and that's where drivers (and cyclists) can enter the junction to wait until they can turn right. In some cases, a driver (or cyclist) might have entered the junction on a standard green signal and then after a while, they get a right turn green arrow. This will be because the oncoming phase will have been stopped to allow the right turn to clear (called an "early cut-off").
Where there are cycle-specific signals, the same applies and that is why in the UK, we cannot use the Dutch-style simultaneous green because of the issue of "ahead" flows being in conflict with those coming from the side.
Pedestrians are treated differently. They can cross all arms of a junction at once, and they can also do that plus crossing diagonally because they are trusted to negotiate with each other. I don't know for sure, but I suspect this arises from the basic ability of people to cross when or where they like and in theory, a bunch of people might just all decide to freestyle it at a junction anyway.
The non-conflicting phase approach also applies to pedestrian phases insofar as we can't give a green to drivers (and/ or cyclists) to proceed at the same time as pedestrians cross the same space because that is risky, but all bets are off when we don't give pedestrians their own crossing phase. This is the nub of the double standard I am talking about.
A traffic/ pedestrian conflict only exists when we give pedestrians their own crossing phase. Where we don't, the conflict apparently disappears, even where we suggest a place to cross with dropped kerbs and tactile paving as with my example.
People undertaking the design of junctions are expected follow certain rules and there is guidance to help them. They are also expected to undertake a level of risk assessment on their designs to help support their approach. When it comes to making a decision not to provide pedestrian phases, I am sure the risk assessments are undertaken, except the balance ends up lying with favouring motor traffic and it then becomes the responsibility of the crossing pedestrian to undertake their own "dynamic" risk assessment whenever they choose to cross.
This is at the heart of the double standard. We have created rules and expend our design efforts to make these junctions as safe as we can for drivers. We do for cyclists too, but that's limited to those happy to mix with traffic and in terms of the A12, I assure you I am exercising the Boateng defence vigorously.
However, when it becomes a bit tricky because we want to maximise motor traffic capacity, all care for the most vulnerable evaporates and any thoughts of ensuring we design out conflicting phases is thrown out the of the window as we throw people under the bus.
I have said that junctions controlled by traffic signals can never be pedestrian-friendly and to some extent, that doesn't really matter. If we tackle roads and streets at the network-level, deploying area-wide traffic management policies, then we can actually make loads of side road junctions pedestrian-friendly, even if we have to go through stages to get there.
This does mean that some junctions, like my A12 example, will require traffic signals and that's fine in terms of the wider prize. However, if we are interested in equality and accessibility, we should be providing pedestrian phases wherever pedestrians are expected to cross the roads within a signalised junction; unless there is a way to provide decent grade-separation.
I won't get into the details around staggered and two-stage crossings, but suffice to say layouts should meet a standard of accessibility, and in the interim we might not provide crossings on every arm, so long as every pedestrian route through is accommodated somehow.
This is not even a radical proposition in my view. This should be the absolute basic standard for new and refurbished signalised junctions. If this means a loss of motor traffic capacity and more congestion, then so be it. Perhaps that will act as a wakeup call for everyone to realise that we have scarified pedestrians at the altar of the motorcar.
I would go as far as to say any designer who is not planning to provide this basic level of service should be held to account for their decisions. To anyone under pressure from their boss on subjects like this, then my message is you really need to start to refuse to work in that way if you can, otherwise you need to carefully document things and get your boss to confirm their instructions in writing. Oh and councillors - this is on you as well.
If not, I guess we could try the tried and tested UK approach of putting up a sign where the push button might otherwise be.
I am not a traffic signals expert, but as ever, this blog helps me learn as well as run through ideas and develop my views. If you are interested in traffic signals, my other posts on the subject and listed below.
Why "traffic signal pie"? Well, for any given junction size and layout, there will be a maximum pie of capacity to be served. It's just we serve too much to drivers and in some cases, pedestrians hardly see a crumb.
First Slice, January 2014 - a look at standalone crossings
Second Slice, March 2014 - the SCOOT system
Third Slice, July 2014 - protected junctions.
A Midnight Feast (For Some), November 2014 - a deconstruction of the Greater London Authority Conservative's idea of switching of traffic signals at night. Which obviously ignores those not driving.
So Near, Yet So Far, March 2015 - a comparison between nearside and farside pedestrian.
The Great Switch Off? No, It Runs Far Deeper, January 2016 - a look at a report from the libertarian and very motor-centric Institute of Economic Affairs.
Staggering! July 2017 - a look at why pedestrian crossings at junctions are often staggered.
Time for T, March 2018 - an exploration of a real life idea to connect a side road junction with a quiet local cycle route to a main road with cycle tracks.
Just Rephase the Lights, March 2018 - a pushback on the parlance.
Partially Protected Turns, January 2019 - a novel approach to a protected junction design where side road space is limited.
More Partially Protected Turns, February 2019 - the second part.
The Long Wait, June 2019 - me moaning about a local toucan crossing.
Automatic for the People, November 2019 - all about push-buttons.
Begging the Question, September 2020 - more on push-buttons and especially how they are a pain for people cycling.
Critical Crossings, February 2022 - my views on how we should react to how we treat different modes in an emergency response to signal failure.
Innie vs Outie, May 2022 - a look at the Dutch vs CYCLOPS protected junction designs.
No comments:
Post a Comment