Part of Phase III of the Propensity to Cycle Tool project, awarded to the team by the Department for Transport to add new features, was to ensure the routing data remained up-to-date. In March we launched the School Travel layer, which provides insight into routes with the most cycling potential to inform investment and policies to reduce air pollution and increase activity levels among children nationwide (Goodman et al. 2019). This post describes the launch of another new dataset, more subtle but perhaps equally important as the School Travel layer: updated route data for the commuting layer in the PCT.

Why update the routes?

The quickest way to get from A to B would remain largely static if the network didn’t change. However, the cycle network in many cities is evolving. This is generally a good thing: new cycle infrastructure can reduce journey distances and the PCT should be based on the latest data. A new bridge, for example, could lead part of a network to shift such that investment in ‘feeder routes’ is necessary to create a cohesive cycle network. (Less positively, severance could reduce cycling potential, for instance if a cut-through for cycling is blocked).

A good example of the importance of updating the routes to reflect new infrastructure can be illustrated with reference to Sunderland. The Figures below show a ‘before and after’ image of the LSOA route network at this location: in the top image there are only 2 crossings along ~3km of the River Wear shown in the maps. In the bottom figure there are 3 crossings, showing a substantial increase in cycling potential along the south bank of the river around European Way (see the area in Google Maps and on http://www.pct.bike/m/?r=north-east).

lsoa-network-bridge-sunderland-old
Cycling potential estimates in East Sunderland over the River Wear before the new LSOA route network data was added, under the Go Dutch scenario.

lsoa-network-bridge-sunderland-new

The new routes over the river to the West of the bottom map was made possible thanks to the Northern Spire Bridge, opened in August 2017, after the old route data was generated. This shows that OSM is evolving, thanks to thousands of contributors across the world, and is more than 80% complete according to a recent paper (Barrington-Leigh and Millard-Ball 2017). The OpenStreetMap data underlying the routes generated by CycleStreets.net, who provide routing services for the PCT, is constantly evolving. New contributors add to and improve the data all the time, making the routes generated more realistic over time.

The ‘before and after’ screenshots also illustrate another improvement: the new route network layer (LSOA, image) does not ‘double count’ cyclists where 2 roads meet or run parallel to each other. Therefore, while the bridge to the East in the top image shows a Go Dutch potential of 2000+ cyclists, the bottom image shows 1,000-2,000.

For these reasons over the last few months we’ve been preparing an update to the commute layer. The results may be slightly different in some places, for example, if the fast route along a particular desire line becomes shorter, the estimated levels of cycling uptake will increase slightly, reflecting the fact that people are more likely to cycle short distances (assuming there is good infrastructure).

Updated route network data

Initially, the PCT was based exclusively on data at MSOA level, each inhabited by around 7000 people (Lovelace et al. 2017). The ‘vector route network’ layer is based on this input data and was initially the only route network layer available. In early 2017, however, we expanded the PCT data to include data representing trips between LSOAs, each of which homes around 2000 people, as outlined in the blog article Lower Level Super Output Area route network now up!.

Due to computational constraints, we only provided the route network representation of these new ‘LSOA routes’ as a raster image. The clickable vector layer was still that based on the MSOA route data. Thanks to developments in the R package stplanr (Lovelace and Ellison, 208), the computational constraints preventing us from building route networks on the larger LSOA routes no longer apply.1 We have therefore decided to replace the MSOA route network layer with an LSOA route network layer generated. The results show the benefits: the figure below, for example, shows the before and after route networks in Hereford and Worcester.

Old (left) and new (right) vector route network layers illustrated in a rural region (Hereford and Worcester)

An updated near market scenario

In terms of the user interface, the PCT remains unchanged, with one major exception. We have added a new national scenario called Near Market (equality), which you will notice on the dropdown menus. This will be described in detail in a subsequent post. For now, the main message is that up-to-date route data, and associated route networks, have been added to the PCT, making the data even more useful for transport planning in an ever-evolving transport network. As before, you can download the data, as described in the final section of this blog post.

Data downloads

Making the code underlying the PCT, hosted at github.com/npct, open and transparent was always an important component of the project. Making the data open access ensures that local authorities and others can make the best use of the work. We have made all the new route data available in the web application from the Region data and National data tabs, as illustrated in the screenshot for the West Yorkshire region below (see this hosted at http://www.pct.bike/m/?r=west-yorkshire ).

 

Screenshot from 2019-04-29 21-12-15
The Region data and National data tabs in the PCT shown in context.

The new data, and associated uptake values, replace the old data for all geographic levels: zones, desire lines, routes and route networks. The MSOA route network data is no longer available as it has be superseded by the LSOA vector route network data. The new data are available as .Rds files which can be read into industry standard open source data analysis software such as R or or QGIS.

References

Barrington-Leigh, C., Millard-Ball, A., 2017. The world’s user-generated road map is more than 80% complete. PLOS ONE 12, e0180698. https://doi.org/10.1371/journal.pone.0180698
Lovelace, R., Goodman, A., Aldred, R., Berkoff, N., Abbas, A., Woodcock, J., 2017. The Propensity to Cycle Tool: An open source online system for sustainable transport planning. Journal of Transport and Land Use 10. https://doi.org/10.5198/jtlu.2016.862
Lovelace, R., Ellison, R., 2018. stplanr: A Package for Transport Planning. The R Journal 10, 7–23. https://doi.org/10.32614/RJ-2018-053
Goodman, A., Rojas, I.F., Woodcock, J., Aldred, R., Berkoff, N., Morgan, M., Abbas, A., Lovelace, R., 2019. Scenarios of cycling to school in England, and associated health and carbon impacts: Application of the ‘Propensity to Cycle Tool.’ Journal of Transport & Health 12, 263–278. https://doi.org/10.1016/j.jth.2019.01.008

 

 


  1. This is thanks to the new function overline2(), which is 2 orders of magnitude (100+ fold) faster than its predecessor overline()