Introduction

• Who am I? Professor of Transport Data Science at the University of Leeds, with 2 years experience working in the UK Civil Service (Head of Data and Digital, Active Travel England, part of the Department for Transport).
• My focus: Building open, reproducible, and policy-relevant transport planning tools.
• Today’s topic: Building a community of practice for data-driven transport decision-making in Ukraine.
• Why it matters: Supporting evidence-based transport planning, recovery and EU integration.

The challenge: the data-to-policy gap

We have more data than ever before, but are we using it effectively?

Common Barriers:

• Data Access: Data is often siloed, proprietary, or not in a usable format.
• Skills & Tools: Lack of training and accessible tools to analyse the data.
• Institutional Culture: Resistance to new methods and a disconnect between analysts and decision-makers.
• Procurement: Over-reliance on ‘black box’ commercial solutions.

Problem: stop-start funding for transport

Source: github.com/itsleeds/uktransportauthorities

Problem: proprietary ‘black box’ tools

Source: Lovelace (2021), source code: github.com/robinlovelace

Problem: ‘just build on top of it’ (dependency)

A solution: open & reproducible workflows

Openness builds trust and empowers collaboration.

• Open Source: The code is free to use, modify, and share.
• Open Data: Public data is accessible to everyone.
• Open Methods: Methodologies are transparent and documented.
• Open Access: Research and educational materials are freely available.
• Community of Practice: People learning from each other.

Solution 1: open data & standards

Good analysis starts with good data.

• Data Standards are Key:
  • GTFS for public transport schedules.
  • GBFS for shared micro-mobility (bikes, scooters).
  • OpenStreetMap for detailed street network data.
• National Data Portals: A vital resource for official statistics.
• Example: Sourcing road network data from OpenStreetMap for transport analysis across an entire country.

Solution 2: open source tools and materials

Powerful, free, and adaptable tools for transport analysis.

• Python and R Ecosystems: Mature libraries for data science, statistics, and visualisation.
  • Free teaching materials including Geocomputation with R’s Transport chapter (Lovelace et al. 2025), Geocomputation with Python (Dorman et al. 2025) and Applied Geostatistics with Python (Pyrcz 2024).
    
  • There is so much content out there, updated every month, it’s worth searching online for the latest resources, and getting in touch with authors.
• Python libraries: osmnx for OpenStreetMap, geopandas for spatial data, grid2demand for demand modelling, madina for urban network analysis (Sevtsuk and Alhassan 2025).
• These tools can be adapted to local needs and data sources, and extended

Solution 3: communities of practice

Capacity building is not just about tools, it’s about people.

• A Community of Practice is a group who “share a concern for something they do and learn how to do it better as they interact regularly.”
• Open source projects naturally foster these communities.
• Examples: rOpenSci, QGIS user groups, transport modelling forums.
• To build capacity, invest in building communities, they don’t just happen, they must be built.

Personal connections are more important than platforms.

Solution 4: openly available results

Results that are open access can be seen ‘de-silo’ transport planning.

Solution 5: open and reproducible environments

Flexible and open reproducible development environment using devcontainers and GitHub codespaces (Google Colab alternative).

Source: robinlovelace.net/itfworkshop

Case study: open OD data in Spain

• The spanishoddata R package provides access to open origin-destination data (Kotov et al. 2026)
• It is a community-driven project, adopted by rOpenSpain, ensuring long-term maintenance.
• Used by researchers and public bodies, including the Spanish Ministry of Transport.
• A successful model of collaboration between government, academia, and the open-source community.

Example: Propensity to Cycle Tool

pct.bike now used by local authorities across England (Lovelace et al. 2017)

Variants building on it deployed in Ireland, Scotland, Portugal and beyond.

Making it happen in Ukraine

Caveat: No one-size-fits-all solution, may be possible to leapfrog some challenges!

How can these principles be applied?

1. Data Audits: What data exists? What format is it in? What can be opened?
2. Pilot Project: Choose a specific problem/region to test solutions that can scale nationally.
3. Invest in Training and collaboration: Build future-proof skills, including efficient use of AI.

What we can learn from Ukraine

• Rapid adoption of new technologies.
• Dealing with network disruptions.
• Building resilience in the face of external shocks.
• Specifically: dealing with infrastructure damage, variable fuel supply, and rapidly changing demand patterns.

These are all things many countries are likely to need

Source: wartranslated

Links and future directions

• Sustainable capacity is built on open data, open tools, and an open community.
• bridgeukraine.org an initiative to build capacity
• Slides and code: github.com/robinlovelace/itfworkshop
• Related materials: tdscience.github.io/tartu26/
• Contact: r.lovelace@leeds.ac.uk

Example: refugee displacement flows from Ukraine (video)

An interactive flow map. Reproducible code: github.com/robinlovelace/itfworkshop