Transport network visualisation hackathon

Introduction

This will be an informal hackathon on generating, modelling and modifying geographic representations of transport networks.

The hackathon will be held on Thursday 3rd October at Institute for Transport Studies (34-40, University Road) room 1.11 from 11:00 to 17:00.

Aim

The aim of the hackathon is to bring people working on and with an interest in visualising transport networks together to develop new and evolving ideas, techniques and tools.

Agenda

The event will be split into 2 parts: demo of existing tools plus sharing of ideas and the hackathon itself. The rough timings will be as follows (these can be adjusted on the day):

• 11:00 - 12:30: Demo of existing tools and sharing of ideas

  • 11:00 - 11:15: Introduction and welcome (Robin Lovelace)
  • 11:15 - 11:25: Demo of parenx package for network simplification and visualisation of networks with Python (Will Deakin)
  • 11:25 - 11:35: Demo of reproducible traffic flow visualisation with R (Juan Pablo Fonseca Zamora)
  • 11:35 - 11:45: Demonstration of ‘core network’ generation and visualisation in the Network Planning Tool (Zhao Wang)
  • 11:45 - 11:55: Demonstration of visualisation with MapLibre, with reference to od2net and the Network Planning Tool for Scotland (Robin Lovelace)
  • 11:55 - 12:00: Introduction to TGVE and the hackathon (Layik Hama)
  • 12:00 - 12:15: Q&A of ideas for the hackathon

• 12:15 - 13:00: Lunch and networking

• 13:00 - 16:00: Hackathon

  • At around 15:00 we will ask teams which teams want to present and take a break

• 16:00 - 17:00: Presentations and discussion

  • 16:00 - 16:40: Presentations
  • 16:40 - 17:00: Discussion and wrap-up

Input datasets

We will use existing input datasets including:

• OpenStreetMap
• Ordnance Survey Open Roads
• Data from network planning tools, PCT and Network Planning Tool for Scotland

Although these datasets are available nationally we will provide subsets of them for ease of use:

• Small area (e.g. covering LIDA and direct surroundings)
• Medium area (e.g. covering York)
• Large area (e.g. covering Yorkshire and the Humber)

Links to tools and reproducible examples

Parenx-related

• parenx
• Extract GB mainline rail data from OpenStreetMap
• QGIS data visualisation tool
• GB rail centre-line track-model with example showing conversion from SHP to GPKG
• Project worldpop and Global Human Settlement Layer population data onto a hierachical layers of hexagons

R-related

• Code with reproducible code for visualising traffic flow data: https://github.com/juanfonsecaLS1/netvis-trafficflows

Data

• OpenStreetMap Leeds
• OpenStreetMap Princes Street, Edinburgh
• Rail network model Doncaster
• Rail Data Marketplace requires registration
• World Population data
• Global Human Settlement Layer data

Miscilaneous

• cruse.bike web application and associated open access paper

• Illustration of network simplification (in-progress open source paper): https://nptscot.github.io/networkmerge/

Hack ideas

• Comparing network simplification tools to support visualisation of networks
  • For example parenx vs spgeo
• Visualising traffic volumes
• Visualising traffic speeds and other attributes
• Visualising population, urbanisation or pollution characteristics

Outcomes

The hackathon will be an opportunity to develop new tools and ideas for network generation and analysis. It will also build skills and foster collaboration between researchers and practitioners in the field.

• Skills development: hackathons are a great way to learn new skills and share knowledge.
• New tools and better documentation and understanding
  • Better documentation of existing projects (e.g. NPT)
  • More use cases, identification of bugs and improvements to the tools
• Ideas for future work
  • How the tools can be used in policy
  • Research directions

Pitches

Each team will be given the opportunity to present, we will aim to have a maximum of 10 presentations.

Assessment criteria

The pitches will be judged based on the following criteria:

• Reproducibility
• Attractiveness
• Usefulness

The judges will be:

• Will Deakin (Network Rail, Data Digital and Technology IT Delivery)
• Robin Lovelace (University of Leeds, Institution for Transport Studies, Active Travel England)
• Layik Hama (University of Leeds, School of Computer Science)

Prerequisites

We expect you to:

1. bring your own machines
2. have installed either “Python and Jupyter” or “R and Rmarkdown”. For better results/success it would be great to have sf in R and geopandas in Python

To cover the above, the recommended reading book is Geocomputation with R and Geocomputation with Python which we will encourage attendees to have a look at.