NSW fixed cameras × open connected-vehicle speeds (June 2026) × same-road matched controls. Descriptive speed association — not a crash or causal study.
We checked 123 NSW fixed speed cameras against open connected-vehicle data. The surprise: nothing special happens at the camera.
Median operating speeds within 200 m of the camera are the same as on the very same road 0.6–1.4 km up-corridor (median gap +0.0 km/h across 17 strict pairs; camera side slower at just 7/17). No slam-brake-and-surge signature in month-aggregated data.
Compare camera corridors against a naive baseline — every same-limit, same-class road in the surrounding tiles — and cameras look +12 km/h FASTER. That is not an enforcement effect; it is selection: cameras sit on high-standard arterials. Any camera evaluation without same-road controls inherits this artefact. We publish it as a warning, not a finding.
| stage | n | rule |
|---|---|---|
| TfNSW fixed-camera list | 123 | published open data (CC-BY) |
| GPS-locatable | 103 | coordinates parseable |
| paired | 60 | CV features ≤200 m + control band 0.6–1.4 km, limit known |
| non-school-zone | 41 | time-variable limits excluded |
| strict same-road pairs | 17 | OSM-verified identical road name, same posted limit, grade-similar |
Refuse conditions: no claim without a mart binding · no causal or crash language (machine-blocked) · selection exhibit always labelled.
Every answer is planned as SQL on the marts above, then machine-verified; unverifiable → it refuses. Same 5-rule manifesto as ask.crashstats.com.au.