Better Intersections

An expansion of the Better Intersections Cinematic Universe™️

These are a collection of charts picking apart the Better Intersections dataset. They provide multiple avenues to find further patterns in complex and incomplete data, but also as a tool for communicating and demonstrating improvement over time (or perhaps lack thereof).

I've intentionally added charts to demonstrate the limitations of the current accuracy and coverage of the dataset. Please note that the data is crowdsourced by a relatively small number of volunteers - for which I am incredibly grateful to those who have contributed! - and as such should not be used as a primary source of truth for any decision making (yet) - but a useful tool for further investigation nonetheless!

All code used to generate these charts is open source on Github. Contributions are very welcome! Please raise an issue if you find any bugs, or feel free to contact me via email (jake@jakecoppinger.com) or Mastodon.

Please note: this is a living document and is still in a draft stage - I wrote it in 2 days.

Data sources, implementation and caching concerns

All geographic data is from OpenStreetMap. All measurement data is from Better Intersections. Charts are generated using Observable Plot.

Generating these charts requires a large number of queries to Overpass Turbo to fetch OpenStreetMap data (eg. fetching all traffic lights for every council in Sydney). To mitigate load on these community-run servers, all requests are made and cached to the Better Intersections database at compile time.

Any intersections added since the last build will trigger client-side requests to the OSM API, with the exception of categorising intersections by council. This means any measurements since the latest build within Sydney will not appear in charts that filter results by Sydney specifically.

Notes on best practice

More comprehensive details are at Shining a Light on the Traffic Signals of Sydney (July 2023).

The excellent City of Sydney "A City for Walking: Strategy and Action Plan - Continuing the Vision" draft states action 4 (pg. 36) is:

We will work with Transport for NSW to ensure that signal phasing prioritises people walking. The City will advocate for:
  • Automated pedestrian phases
  • A maximum wait time at intersections of 45 seconds for people walking with a target of 30 seconds

I've added lines that display these best practice benchmarks where the maximum wait time is displayed (calculated here as the sum of the flashing red and solid red durations for the same traffic light sample).

Loading...