Reproducible Geocomputation for Policy-Relevant Research and Tools
CASA, UCL, London
Robin Lovelace, University of Leeds
2022-07-13
1 / 23

Session plan

1 Introduction to my work and previous tools

2 / 23

Session plan

1 Introduction to my work and previous tools

2 New and emerging tools

2 / 23

Session plan

1 Introduction to my work and previous tools

2 New and emerging tools

3 Live demo and testing: ActDev

4 Coding and questions

Beginner in R: install R and follow instructions at https://itsleeds.github.io/rrsrr/
Intermediate level: work through the code in Chapter 13 of Geocomputation with R: https://geocompr.robinlovelace.net/transport.html
Advanced: Add a new site in the ActDev tool: https://github.com/cyipt/actdev/blob/main/code/tests/add-new-site.md

2 / 23

1 Introduction to me and my research

Further info: https://www.robinlovelace.net/

3 / 23

We have 5 research groups

Working with TASM

Links with WYCA

Academic input

Sarah Sharples

Data can make a difference

4 / 23

Where I am now: PCT used by many in government

Too much data?

A nice problem to have?

Data historically a limiting factor

New approaches needed to tackle 'big noise'

Boyce, D.E., Williams, H.C.W.L., 2015. Forecasting Urban Travel: Past, Present and Future.

5 / 23

Too much data?

A nice problem to have?

Data historically a limiting factor

New approaches needed to tackle 'big noise'

Boyce, D.E., Williams, H.C.W.L., 2015. Forecasting Urban Travel: Past, Present and Future.

Lovelace, Robin, Mark Birkin, Philip Cross, and Martin Clarke. 2016. ‘From Big Noise to Big Data' https://doi.org/10.1111/gean.12081.

5 / 23

For most of the history of transport planning data scarcity has been a major concern
Now we have an abundance of datasets, many of them incompatible
Concrete example: OD to WPZ data in central London (could ask if anyone knows, it's a mess hehe)

Problems worthy of your time6 / 23

Problems worthy of your time

Ideas

Hackathons

Active travel

Road safety policy

Post COVID recovery

Levelling up metrics

Nature recovery networks

Citizen science and data literacy

6 / 23

Climate change: The elephant in the room

7 / 23

What is climate change? And how does it related to open source software?

It's great to be able to talk about climate change upfront. Usually it's behind the scenes.

To solve problems you need to talk about them.

Everyone including the elephant knows that climate change means the world is getting warmer.

But fewer people talk about 'global drying' and likely impacts such as sudden sea level rise that can only be described as catastrophic

As with health, the most solutions tackle the root causes of the problem. So what is the root cause of the problem? We can go back a few steps...

Climate change is caused by emissions.

Emissions are (primarily) caused by combustion of fossil fuels

But what causes that? Energy consumption.

What causes energy use? Demand for energy intensive things, transport, manufacturing

Example: Climate change, what data science can bring

Emotive and potentially polarising issue

Data can provide a shared starting point

Nationally scenarios are vital

Locally, visions, trust, buy-in and participation are key

8 / 23

The Downward trend during coronavirus saw emissions drop by ~10%

The scale of the challenge is to reduce emissions by 10% every year, every year, and the challenge gets hard with each year

That means: we need transformational

Existing tools: The Propensity to Cycle Tool

Early prototype of the tool developed at Open Data Institute Leeds hackathon, 2015
National deployment of the Propensity to Cycle Tool 2 years later (PCT.bike) (Lovelace et al. 2017)

9 / 23

Key aspects of the PCT

Based on open source code
Based on open data
Results published as open data
Publicly available web application encourages participation and evidence-based debate

Open data and accessible results lead to participation. See https://twitter.com/search?q=cyipt.bike%2Frapid

10 / 23

Reproducibility and extending the tool

Find commuting desire lines in West Yorkshire between 1 and 3 km long in which more people drive than cycle:

11 / 23

How it works

12 / 23

Demo of the PCT

Source: https://www.pct.bike/

13 / 23

Existing tools: the Rapid Cycleway Prioritisation Tool14 / 23

2 New and emerging tools

15 / 23

2 New and emerging tools

Evidence-based policies in government: Data Science Fellowship at N. 10

15 / 23

Turing Fellowship
LIDA internship on open transport infrastructure data
Links with DfT, MHCLG, TfNH, international partners

2 New and emerging tools

Evidence-based policies in government: Data Science Fellowship at N. 10
Future areas of development: Reproducible Bayesian modelling of proportions (Dirichlet regression), Machine Learning, Decarbonisation Agenda

15 / 23

Turing Fellowship
LIDA internship on open transport infrastructure data
Links with DfT, MHCLG, TfNH, international partners

The future of transport tools

Modular

Future proof

Scalable

Vector/

Raster/

Source: Morgan and Lovelace (2020) Implementation: stplanr

16 / 23

The future of transport tools

Modular

Future proof

Scalable

Vector/

Raster/

Source: Morgan and Lovelace (2020) Implementation: stplanr

Approach: OD -> Desire Line -> Route -> Route Networks

16 / 23

So what's the overall approach here?

3 Live demo and testing: ActDev

ActDev: a data driven tool for evidence-based planning

17 / 23

3 Live demo and testing: ActDev

ActDev: a data driven tool for evidence-based planning

Interactive demo: https://actdev.cyipt.bike/

Talbot, Joseph, et al. 2021. ‘Active Travel Oriented Development: Assessing the Suitability of Sites for New Homes’. https://osf.io/7fuq5/

17 / 23

The ActDev project: https://actdev.cyipt.bike/

A/B Street - R integration, see https://github.com/a-b-street/abstr
Give it a try!
- Find a new housing development that has good active travel provision and potential
  - How could active travel provision be made better?
- Find a housing development that has poor active travel provision
  - How could active travel provision be improved, based on the data?
  - Should new houses be built in this location?

18 / 23

19 / 23

Discussion: scalability vs resolution

Source: UKRI CREDS project repo

20 / 23

Also complexity vs simplicity and readability

Making transport data come to life

Source: Lovelace, Tennekes, Carlino (2022)

21 / 23

4 Coding and questions

Beginner in R: install R and follow instructions at https://itsleeds.github.io/rrsrr/
Intermediate level: work through the code in Chapter 13 of Geocomputation with R: https://geocompr.robinlovelace.net/transport.html
Intermediate: demo of the {simodels} R package: https://github.com/Robinlovelace/simodels
Advanced: Installation and use of the odjitter package: https://github.com/Robinlovelace/odjitter
Advanced: Add a new site in the ActDev tool: https://github.com/cyipt/actdev/blob/main/code/tests/add-new-site.md

22 / 23

Thanks!23 / 23

Thanks!References23 / 23

Thanks!

References

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

23 / 23

Thanks!

References

Morgan, M., Lovelace, R., 2020. Travel flow aggregation: nationally scalable methods for interactive and online visualisation of transport behaviour at the road network level. Environment & Planning B: Planning & Design. https://doi.org/10.1177/2399808320942779

23 / 23

Thanks!

References

Lovelace, R., Tennekes, M., Carlino, D., 2021. ClockBoard: a zoning system for urban analysis. https://doi.org/10.31219/osf.io/vncgw

23 / 23

Thanks!

References

Lovelace, R., Tennekes, M., Carlino, D., 2021. ClockBoard: a zoning system for urban analysis. https://doi.org/10.31219/osf.io/vncgw

Lovelace, Robin, Rosa Félix, and Dustin Carlino. “Jittering: A Computationally Efficient Method for Generating Realistic Route Networks from Origin-Destination Data.” OSF Preprints, January 13, 2022. https://doi.org/10.31219/osf.io/qux6g.

↑, ←, Pg Up, k	Go to previous slide
↓, →, Pg Dn, Space, j	Go to next slide
Home	Go to first slide
End	Go to last slide
Number + Return	Go to specific slide
b / m / f	Toggle blackout / mirrored / fullscreen mode
c	Clone slideshow
p	Toggle presenter mode
t	Restart the presentation timer
?, h	Toggle this help

Reproducible Geocomputation for Policy-Relevant Research and Tools

CASA, UCL, London

Robin Lovelace, University of Leeds

2022-07-13

Session plan

1 Introduction to my work and previous tools

Session plan

1 Introduction to my work and previous tools

2 New and emerging tools

Session plan

1 Introduction to my work and previous tools

2 New and emerging tools

3 Live demo and testing: ActDev

4 Coding and questions

1 Introduction to me and my research

Data can make a difference

Too much data?

Too much data?

Problems worthy of your time

Problems worthy of your time

Ideas

Climate change: The elephant in the room

Example: Climate change, what data science can bring

Existing tools: The Propensity to Cycle Tool

Key aspects of the PCT

Reproducibility and extending the tool

How it works

Demo of the PCT

Existing tools: the Rapid Cycleway Prioritisation Tool

2 New and emerging tools

2 New and emerging tools

2 New and emerging tools

The future of transport tools

The future of transport tools

3 Live demo and testing: ActDev

ActDev: a data driven tool for evidence-based planning

3 Live demo and testing: ActDev

ActDev: a data driven tool for evidence-based planning

Discussion: scalability vs resolution

Making transport data come to life

4 Coding and questions

Thanks!

Thanks!

References

Thanks!

References

Thanks!

References

Thanks!

References

Thanks!

References

Session plan

1 Introduction to my work and previous tools

Help