Connecting the Dots

EthanolFour years into Oxfordshire’s latest transport strategy and here we are embarking on another multi-stage consultation for Local Transport Plan 4. The official reasoning makes a case for it but can the goals and objectives of a strategy covering two decades really have changed so much that it warrants starting from scratch? We’ll be taking a look at the detail following tomorrow’s Connecting Oxfordshire meeting at 7pm in Henry Box School.

Connecting Oxfordshire conjures up the idea of places joined together, settlement to settlement, Carterton to Witney, Witney to Oxford, but that’s only one level of local transport. The hyperlocal is at least as important – how we move around within settlements – and it’s the one we consistently fail to make more efficient, pleasant and healthy. Travelling along better local transport links is one part of the solution, not having to sit in 15 minutes of congestion at either end is the other.

The only major road project during four years of LTP3 to be planned, signed-off and (soon to be) built in Witney is the Ducklington Lane junction improvement. It started out with a £2 million budget and a brief to future proof it for a predicted 2030 level of motorised traffic. When the engineering contractor Atkins ran the junction through its capacity modelling software ‘computer said no'; it didn’t have enough lanes for stacking. That’s a technical term for temporarily parking vehicles at a junction so a queue doesn’t back up so far that it adversely effects the next junction upstream.


(Edited to update diagram to Option 4 in Annex 5)

What to do? Quite a large area around the junction is deemed highway land and during a site visit Atkins’ engineer noticed the B&Q corner “doesn’t go anywhere” (the business park being hidden behind a very large hedge) so no crossings or pavement to it would be required. That was a boon to a modeller of motor traffic because it meant:

  • the pavement removal would give more space for traffic lanes,
  • more time could be retained for traffic movements.

Stacking traffic is only part of the equation for a signalised junction, there also needs to be a enough time available to clear the stack during each light sequence so a growing remainder doesn’t back up over time. Adding a proper Toucan crossing to reach B&Q would scupper the junction’s maximum capacity in the traffic model.


Despite bikes being regularly parked at B&Q, and people beating paths through the bushes to it, the traffic counts showed low numbers of people walking and cycling on the western arm of the junction, supporting the removal of the pavement along with the potential to cross there. If you’ve tried it you’ll know why; it’s relatively risky and far from pleasant. It’s so unpleasant and off-putting in fact, a sympathetic observer might conclude that the people counts ought to be multiplied by a factor of about 100 to take that into account. Traffic models can’t empathise neither can they model people walking or cycling.

During the consultation bike users requested better access which resulted in a 2m shared path along Thorney Leys to an informal pedestrian crossing, with a refuge, to Thorney Leys Business Park. Oxfordshire County Council could not provide an official, safe and convenient bike route for the employees and visitors to B&Q and the rest of the businesses there. ‘Encouraging’ people to cycle and relieve congestion doesn’t yet stretch to giving them a place to do it.

I don’t think we’ll get another chance to put that right for a very long time and this corner of Witney will continue to be cut off from the bike network for now, and will remain so when the junction eventually starts to operate at its designed capacity in the late 2020s. All that tarmac poured, rolled and marked out, sat there through rain and shine, developing potholes, waiting for that day. Does that sound innovative to you?

So perhaps Ian Hudspeth, the architect of Connecting Oxfordshire, is right, the Local Transport Plan does need a revamp from the top down. I’m not so sure. Cycling and walking were high priorities for all settlement types in LTP3, including Witney, but that had no influence when it came to providing people with a safe and convenient option to cycle to Thorney Leys Business Park. Not considering walking and cycling until after the road layout for the junction was complete was a failure to plan for walking and cycling; good policy made no difference to an old process.

Come along tomorrow and ask Ian how he’s going to fix that for short journeys at the hyperlocal level, as well as for the local transport links to and from Witney.

Cycling in the Matrix

As a concept Space for Cycling can be applied anywhere, even in the virtual world of design, and thanks to some pointers from Andy R commenting on The Dimensioned Cyclist it seems clear to me that there is little or no space for cycling in the Matrix either…

Vehicle Tracking from Autodesk:

Autoturn from Transoft:

My guess as to why we don’t see a bike in these promotional videos is that historically, with vehicular cycling being the dominant paradigm, bikes have never presented a problem to highway or street designers in the way that a firetruck or an articulated lorry has. Again, the author of the official british cycle infrastructure design guidance puts it like this…

Highway designers consider the dimensions of motor vehicles and their swept paths to determine carriageway widths, junction dimensions and parking layouts. The sizes and swept paths of cycles are usually irrelevant in the design of on­road cycle routes, but there are occasions where they need to be considered.

And six years later, with vehicular cycling on busy roads falling further out of favour, the occasions where the dimensions and swept paths of cycling vehicles need to be considered is increasing, particularly for inclusive bikes. Designing successfully for cycling vehicles on segregated routes is no longer a trivial problem when space becomes a limiting factor.

I don’t see 3D modelling as a panacea. In my limited experience as soon as a model gets to a level of complexity where it should be a useful design tool it simultaneously becomes so ‘heavy’ that it takes forever to render an image or simply crashes. I’ve ended up relying on cardboard and pritt stick constructions to determine cable runs and check for access issues during my time as a designer. That was about ten years ago so buggy software and lack of computing power are probably irrelevant now, but a 2D representation of a 3D model still has its limitations. However, I think it’s fair to say that most designers these days spend a great deal of their working life in these virtual environments and I can only speculate on what the longterm effect of designers never seeing a bike in their virtual transport worlds might be. Surely it can’t be positive.

Another niggle that comes to mind is that if bike vehicle types don’t exist in the available software then presumably the Dutch and the Danish aren’t using them either. Perhaps their standards/guidelines are such that widths and swept paths of particular bike types are still largely irrelevant, and they’re more likely to stick to the guidance. In the british context, where space for cycling appears to be more highly contested and we have a plethora of examples of badly designed bike infrastructure, I’m told that the widths and turning radii in our design guidance is generally OK but compliance with it isn’t. So perhaps that’s where the modelling of bikes in computer aided design could play a useful role – as an educational tool.

AndrewGilliganGetting decision makers hands-on with other bike types and giving them a sense-memory of what it’s like to ride them is possible – Wheels for Wellbeing are doing it successfully – but that’s probably an unrealistic strategy when it comes to reaching everyone involved in the design of a cycle network.

Giving designers the option to quickly and easily try out what will and won’t work when things get tight for trikes, handbikes, tandems and tagalongs might be a more productive way of gaining compliance with the guidance. And who knows, people may even come up with insights that lead to better guidance.

Something like a freely available tutorial for Autocad that designers can access from their desks and run through the design of off-road and segregated bike paths using models of different bike types. Is that a doable blue-pill-solution, or should I take the red pill and dig out the pritt stick?