We’ve put a huge amount of thought into what cycle lanes should be prioritised. This has been based on a number of criteria and a healthy pinch of our own subjective preferences. In this post, we’re taking an analytical approach, to see if that gives us any different insights.
In this article, we’re looking at journeys from one point to another – from a source to a destination. This is a very limited approach, as many journeys involve multiple locations, and here we’re only making one-way trips, not round-trips. This approach is easy to do, and with more time we might improve this later.
Step 1 – Select Sources and Destinations
The source for journeys are residential locations. I selected these by dividing Navan into segments and then arbitrarily picking places that would give good coverage. This approach is again very subjective, and I took this approach because it was simple. I wasn’t able to find a good high-granularity data source for population, and didn’t have time to create the data from existing zoning maps.
The locations for journeys are a mix of:
- Leisure and recreation areas, Shopping Centres and Supermarkets
Step 2 – Generate Routes
To get from a Source to a Destination, you travel along a Route. There are services that will give you the route, given a source and destination. I used this very neat solution using Google Sheets and Google Maps because it was easy to use.
There are 31 locations listed above, which means 930 possible routes. I wanted to focus on travelling from a residential location to a destination, which cuts this down to 184 routes.
Step 3 – Get Some Statistics!
Now that I had data for 184 routes, I looked at the distances and travel times. Here’s what I found:
Most journeys are under 5km, and the majority seem to be between 2-4km. How long does that take? Glad you asked!
So the most common journeys are 8-12 minutes. It wouldn’t be unreasonable to suggest an average cycle time of 10 minutes across all routes.
Just for fun, I looked at how much time people would save if they cycled to school instead of walking.
Step 4 – Maps!
Now that I had a list of routes, I wanted to plot them all together on a map. Unfortunately I couldn’t find a way to do this with Google Maps, so I switched to using Open Street Maps, and their cousin Open Route Service. From that I was able to merge together all of the routes and plot them on a map. Here’s what I got:
Hmm. Not bad, but I really wanted to see which segments were more popular than others, so that we potentially prioritise these segments. For this, I used Mapbox.com. I’ve only scratched the surface of Mapbox, but it’s pleasingly easy to use, and I was able to get to something quite useful in a couple of hours.
Here’s where I ended up.
A little better. You can distinguish routes that are used more than others.
Limitations & Improvements
This analysis is very dependent on the locations of the residential homes, and there are only a handful in my data-set. So I hope to add a few more, which should be more representative of the population distribution. Also, one or two of the locations could use a bit of tweaking.
Apart from fixing the mistakes mentioned above, the cycle routes are based on existing map data, and often don’t include places where you can cycle through one housing estate to the next (“low traffic neighbourhoods”). Because Navan is quite compact, it’s possible that many cycling routes can pass through existing housing estates, which are often safer than main roads.
Finally, the routes aren’t very smart. I’m assuming each home is going to each destination, and that each journey is equal. It’s more likely that someone living in Johnstown will go to the Johnstown Shopping Centre more often than Blackwater shopping centre. And each destination attracts the same number of journeys, which probably isn’t true.
Maybe you’ve got some ideas or suggestions? We’d love to hear from you!
I was inspired to write this article after watching this presentation by Kevin Baker of Dublin Cycling Campaign, which is the single best explanation of how to create a cycling network that I’ve seen.
The Google Sheets / Google Maps API was super useful to get started. This article by Amit Agarwal got me up and running in minutes.
The Open Street Map ecosystem is amazing, and there are a huge number of contributors, too many to thank. Kudos!