Françoise Provencher - How to measure a city | PyData Amsterdam 2024

Learn how Montreal measures urban livability using data science, from transit access to green spaces. Discover 5 key criteria for meaningful city metrics and how to implement them effectively.

Key takeaways
  • Montreal used 6 key indicators to measure city livability: shops/services, public transit, cycling networks, parks/green spaces, schools, and cultural infrastructure

  • Five criteria for good city metrics:

    • Simplicity (easy to understand for stakeholders)
    • Faithfulness (captures intended measurement)
    • Sensitivity (reflects meaningful changes)
    • Precision (allows for nuanced measurement)
    • Cost effectiveness (considers limited resources)
  • Used 15-minute city concept as foundation - measuring what amenities/services are accessible within 15-minute walk radius

  • Population density targets help achieve efficiency in infrastructure/services while enabling transit viability and community amenities

  • Different transportation modes were weighted differently:

    • Metro rated 10x better than bus service
    • Train rated 2x better than Metro
    • Weights based on speed and passenger capacity
  • Data challenges included:

    • Accuracy of points of interest data
    • Need for manual verification of businesses
    • Computing limitations requiring Apache Sedona/Spark
    • Balancing precision vs. practicality
  • Important to acknowledge that metrics contain inherent opinions/biases and be transparent about methodology

  • Social justice considerations must be included - ensuring cities are designed for all residents

  • Solutions need to be tailored to each city’s unique characteristics and needs

  • Success metrics should be both detailed enough for experts and simple enough for general public understanding