Following on from yesterday’s ‘background’ post, I thought I’d make a go at putting my money where my mouth is, and suggest some ideas for how we might look at bike congestion as a concept in itself (rather than just ask how car congestion affects bikes and pedestrians, which I plan to think about later).
Part of what I got out of yesterday’s reading was that congestion in scientific terms doesn’t mean angry drivers (or cyclists!) sat in a queue, although that’s the tabloid image. It is about broader network performance, usually thought of in relation to motor vehicles. And the big problem for motor vehicles, in terms of excess delay, wait times, etc. is that you simply have more motor vehicles than can easily pass through the capacity available (which is often temporarily restricted, in London collisions and planned maintenance being two biggies).
However, as with many transport terms that also have a popular meaning, it’s easy to think of congestion only as its symptom; cars stuck behind each other failing to move. And to think, well, in terms of cycling maybe we’d measure busyness on cycle tracks. Which would be useful (the new tracks in London are indeed already getting busy at peak).
But it’s not the only thing. If we want to think about network performance for cyclists, and delays, then what happens in the under 3% of London’s ‘Network of Interest’ with separated cycle tracks is really not the main thing. (It’d be different with a comprehensive cycling system, but we don’t have that yet in London).
It’s about the whole network, and how that enables (or doesn’t) cycle journeys. It’s about whether cyclists can pass through lights without delay – as with Copenhagen’s famous ‘Green Wave’. It’s about the routes they are able to follow.
And here it’s crucially important that this isn’t just about prohibiting cyclists, it’s about where people are willing and able to ride. My colleague Robin Lovelace has coined a nice term ‘cirquity’ to reflect the distance penalty often imposed by taking a ‘quiet’ route to avoid hostile roads. You’d want also to think about hilliness, not so important in most of London perhaps, but another penalty imposed on cyclists for choosing a less hostile route. The measurement would include the delaying impact of poor roadworks management, such as making cyclists dismount and walk around an obstruction. You could even think about whether the network’s so illegible that we get lost (many times I’ve arrived at meetings on the minutes because a diversion leads me onto a confusing one-way system, and I’ve ended up back where I’ve started).
Such a measure wouldn’t contrast daytime to night-time journeys, because network performance for cyclists is about so much more than car congestion. It’s about how well the network copes with cycling, and how much cycling is delayed compared to a theoretical ideal. Night-time journeys as an ideal are also theoretical, as Richard points out in a comment on my last post. Imagine London traffic flowing as smoothly in the daytime as it does at night, when there’s a fraction of the traffic: that would necessitate an impressive modal shift!
For cycling then, if we start with routes actually made (collected via GPS) my theoretical comparison might involve each journey instead being made using the most direct route legally possible (perhaps also adding in some currently impossible connections), assuming no capacity-related delays (bike or car) and assuming minimal waits at junctions – that traffic lights are green, and that turns can be made immediately. The desired speed can be derived from the speed the person actually travelled at – so you’re not assuming that I, an 8mph cyclist, could have been travelling at 12mph. Rather you’d assume that I was able to take a much more direct route, was not held up at lights, and did not have to wait behind other cyclists or cars.
OK, it’s a theoretical ideal (although I do not think making direct routes pleasant for cycling’s that utopian) but certainly no more so than imagining that cars during the day in London could achieve the same journey times as people travelling at night (when for one thing the light phasing’s different – which gave me the idea for this post!)
How might we measure pedestrian system congestion? Maybe I’ll think about this on my ride in…