Invitation To Cross

Last week, I looked at the PV² method of determining whether a pedestrian crossing is warranted and it's nice to say, the post got a fair bit of traction with many people commenting on how their local authority still uses it, even though it hasn't been in official guidance since 1995.

Some people have explained how their local authority has modified and weighted the process which essentially makes a crossing more warranted than under the original PV² thresholds, but it was still a numeric way of justifying and ranking interventions based on the number of people crossing now. It is tough for local authorities given the state of budgets, but the process is flawed for the reasons I set out last week.

The framework for assessment in LTN1/95 The Assessment of Pedestrian Crossings is still very useful in my view, but having to complete one for every request is resource-intensive and it doesn't lead to something which can be compared with another site, and besides it seems that local authorities want to be able to prioritise requests somehow. 

My own view is that prioritisation is a political role with staff providing appropriate evidence and guidance. In practice, that is easier said than done. In a past life, I took over The Requests List and the first thing I did was ask how it was dealt with. The answer was it just kept on growing and although sometimes a scheme came up which could accommodate a request, it was essentially a placebo to keep people quiet.

My approach (in the absence of any real budget for investigations) was to close the list and refer people to their ward councillors who would need to decide whether a request would be put forward to an advisory committee. If a request made it through committee, then it was added to the forward bid programme. OK, there was a little bit of professional advice that went to the committee, but the rationing was at least decided by elected people. 

From a staff resource point of view, not having to maintain a list and simply providing standard replies significantly cut down on workload, freeing staff to concentrate on funded schemes. This is probably of little comfort to those campaigning for improvements, but at least it did cut through the wild goose chase and got to the nub of the matter that decision-making is politically driven and that technical advice is just that. However, it was an unsatisfactory position to be in as it's always staff who have to deal with the complaints in the long term.

I digress, the key thing here is to ask ourselves from a technical point of view if there is a simple way of assessing and ranking requests for crossings, regardless of how the governance in decision making operates. The short answer is that any system we use which attributes any sort of ranking system will be flawed, but maybe the longer answer is there are some low cost things we could do which would still be fairer than PV².

My suggestion is that we use a framework based on driver behavior, "crossability" (traffic flow, crossing width etc) and local land use - in effect, let's actually take pedestrians who are crossing now right out of the discussion entirely and look at the environment that people have to negotiate. The rest of this post is really a work in progress and my thoughts. Maybe others can comment and adapt these ideas, but I'm after something which is relatively simple because of resource constraints and I'll take a little inspiration from LTN 1/95.

Physical Data

For a local authority, there will be some data available and some needs to be collected. In terms of physical information, there are some key metrics which we'll need. I should state at this point, anyone collecting on-street data should be undertaking a risk assessment, assessing risk dynamically on site and if alone, having a "buddy" system in place so that someone else knows where they are and how long they'll be with escalation if they don't check in.

• Driver speed - unless location specific information is already available, then something is needed. Investing in a speed gun and taking some quick readings of say 100 vehicles off-peak would be fine for what is essentially a screening stage. A basic speed gun for a local authority isn't expensive and doesn't need to be of evidential (enforcement) quality. Faster speeds would mean more of a need for a crossing and off-peak, because that's when behaviour will generally be at its worst.

• Traffic flow - a manual peak time traffic count is easy to do and there are plenty of apps available to record vehicle class because we'll ideally want to weight vehicle type using Passenger Car Units. Of course, it is also worth checking to see if there is data published by the Department for Transport. Higher flows mean it's harder to find a crossing gap.

• Width of carriageway (including number of lanes in each direction) and footways - easily done with a measuring wheel. The width of carriageway is important as the wider it is, the harder it is to cross and the more lanes, the harder to cross. Footway width, because it's worth picking that up in case the scheme goes forward.

• Length of unobstructed visibility from the general proposed crossing point - this will be used to check against driver speeds in terms of crossing selection or options because it's pointless if a crossing doesn't work in the location being checked.

• At this stage, I am not looking at casualty data. While it might be a consideration for a scheme going forward, it doesn't necessarily give a good indication of risk exposure as it is simple an absolute at any given location.

Geographic Data

This is about looking at the amenities and any obvious attractors and mapping them. Things like schools, parks, care homes, shops, post boxes, medical facilities and so on are things that people will be walking to locally.

The task needs to be manageable and so I would probably think about what could be within a 5 minute walk around the site that a crossing could help with. 5 minutes is around 420 metres walked (at 1.4 metres per second, a pretty standard assessment of walking speed).

Maybe we could score the number of potential attractors within our selected radius with more being an indication of more need. Yes, this is rationing, but it is based on what people could get to if we built a crossing, rather than how many are crossing now.

Demographic Data

This is easily obtainable information which could be used to gain local insight into age (children and older people), long term illness and disability - probably the groups who most need help crossing the road. For example, children find it harder to judge speed and some people need more time to cross. Scoring here might be challenging, so maybe it could be assessed in comparison to the borough or district average, or maybe a national average. It's generally easy to find a local ward profile.

Processing Data

OK, we have collected a bit of data for the site at the lowest cost we can because we are just trying to screen requests to provide some advice to politicians. What do we report? We can slice this up any way we like and so what follows is just a suggestion;

The way I have set this up is where speed and flow is low, the carriageway is narrow, the number of lanes is low (one each way), attractors are limited and where the area generally has lower numbers of children, older people and disabled people/ those with longer term illness, the score is zero and from that, I think the logic for a score of 1 or 2 makes sense. The maximum score is 20 and so it does make conversion to a percentage really easy. You might need to review the demographic percentages for your local area.

Discussion

As I have said a number of times, this is rationing, but I have tried to couch an assessment in terms of risk exposure, the local demographic and places people might visit. The scoring and presentation echoes that of the Walking Route Audit Tool to give a bit of assessment consistency. Of course, one might then have a minimum score needed to take something forward, or perhaps a year's worth of scores can be used to rank the top five sites or whatever is possible and affordable.

I am not specifying the type of crossing needed because that is where more engineering thought is required, although for example, a 40mph dual carriageway is going to need signals or anything at 50mph+ really needs grade separation. In those cases, it's almost pointless doing an assessment because the answer is obvious.

I remain uncomfortable with rationing as an engineer because that is a political decision, but may be a simple assessment based on how easy it is for the local population to cross is a better way of looking at the issue, rather than the number of people crossing now which PV² skews against. That is for you or your local authority to decide.

Worked Example

Well, rather than have this as a theoretical idea, it's worth going through a couple of examples.

First, we have a location which is fairly busy in traffic terms, with a fair bit of speeding from drivers and a 30mph speed limit. The carriageway is quite wide and the local authority have stuffed two lanes in each direction into the space without a central reserve.

There are fewer children in the ward than the borough average, but more older and disabled people in the ward than the borough average. There are between 5 and 10 amenities in the area and so we get a score of 13.

In the second scenario, we have a much quieter residential estate with a 20mph speed limit, although we have a significant compliance problem which is a worry given that traffic flows are generally low.

There are lots more children in the area because of a primary and secondary school within the estate and in terms of attractors, there is a parade of shops nearby. There are also fewer older people and disabled people in the area versus the borough average. We get a score of 8.

So, this week, we are presenting crossing requests to the highways committee. There really is only budget to take one site forward for more detailed work. The first location has been requested by a local resident who has said they find it hard to cross the road to get to their nearest parade of shops. The second has come from the primary school and is supported by a 150 signature petition from parents. They demand a crossing to slow the drivers down. Which request would you take forward?

Personally, I'm going with the first. The street by the school does have a specific driver behaviour issue which needs addressing, but I don't think a crossing is the right answer in any case. The first site is difficult to cross for sure and there may well be latent local demand. The school would be very appealing with the petition and the emotion, but that's a political consideration. As a professional, my job is easy because I only have to advise, I don't make the decisions.