• Context
• The Propensity to Cycle Tool
• Tools to prioritise infrastucture

• Problem is operationalising this data
• Needs to be provided in a format that can be acted on at the local level

• 3 years in the making
• Origins go back further

Concept (PhD) -> Job at UoL (2009 - 2013)
 Discovery of R programming and shiny (2013)
  'Propensity to Cycle' bid by DfT via SDG (2014)
    Link-up with Cambridge University and colleagues (2015)
     Implementation on national OD dataset, 700k routes (2016)
      Completed LSOA phase (4 million lines!) (2017)

• 2018: (Global PCT?)

• A multi-university project

"The PCT is a brilliant example of using Big Data to better plan infrastructure investment. It will allow us to have more confidence that new schemes are built in places and along travel corridors where there is high latent demand."

• Shane Snow: Head of Seamless Travel Team, Sustainable and Acessible Travel Division

"The PCT shows the country’s great potential to get on their bikes, highlights the areas of highest possible growth and will be a useful innovation for local authorities to get the greatest bang for their buck from cycling investments  and realise cycling potential."

• Andrew Jones, Parliamentary Under Secretary of State for Transport

Included in Cycling and Walking Infrastructure Strategy (CWIS)

• Hilliness and distance are (relatively) unchanging over time
• Model based on polynomial logit model of both:

\[ logit(pcycle) = \alpha + \beta_1 d + \beta_2 d^{0.5} + \beta_3 d^2 + \gamma h + \delta_1 d h + \delta_2 d^{0.5} h \]

logit_pcycle = -3.9 + (-0.59 * distance) + (1.8 * sqrt(distance) ) + (0.008 * distance^2)

• Clear link between infrastructure and uptake, but inconsistent

• We can use 'backcasting' to estimate long-term potential under ideal questions
• But transport authorities need forecasts of future uptake
• From specific interventions in order to do this
• There is much existing work on this
• But not able to automatically estimate uptake resulting from new infrastructure

• A $64,000 question

• Very simple model of uptake (Bristol):

## lm(formula = p_uptake ~ dist + exposure, data = l, weights = all11)
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  2.372e-02  4.207e-03   5.639 2.28e-08 ***
## dist        -1.671e-07  8.424e-07  -0.198  0.84283    
## exposure     4.147e-02  1.523e-02   2.724  0.00658 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.7972 on 906 degrees of freedom
## Multiple R-squared:  0.008318,   Adjusted R-squared:  0.006128 
## F-statistic: 3.799 on 2 and 906 DF,  p-value: 0.02274

See: https://www.cyipt.bike (password protected)

stplanr lives here: https://github.com/ropensci/stplanr

Package can be installed from CRAN or GitHub (see the package's README for details), it can be loaded in with library():

install.packages("stplanr") # stable CRAN version
# devtools::install_github("ropensci/stplanr") # dev version

• The PCT in action: http://www.pct.bike/
• A prototype of the CyIPT: http://cyipt.bike/
• A paper on the PCT (Lovelace et al 2017): https://www.jtlu.org/index.php/jtlu/article/view/862
• An article explaining stplanr (for people interested in the tech): https://cran.r-project.org/web/packages/stplanr/vignettes/introducing-stplanr.html