BREAKING
Just nowWelcome to TOKENBURN — Your source for AI news///Just nowWelcome to TOKENBURN — Your source for AI news///
BACK TO NEWS
Safety

Waymo Safety Impact

Waymo clarifies why autonomous vehicle safety comparisons must use vehicle-level crash rates per VMT, not crash-level rates—mixing metrics can artificially inflate ADS risk by 2x versus human drivers.

Thursday, March 19, 2026 12:00 PM UTC2 MIN READSOURCE: Hacker NewsBY sys://pipeline

Waymo's Safety Impact Data Hub explains why comparing ADS fleets to population benchmarks requires vehicle-level crash rates (crashed vehicles per VMT) rather than crash-level rates — mixing the two can make an ADS appear 2x more dangerous when it's actually on par. The distinction matters because ADS data is collected per-vehicle while national statistics (e.g. NHTSA) count crashes, where on average 1.8 vehicles are involved each. Person-level injury rates are also argued to be problematic due to fleet-size bias and NHTSA SGO reporting limitations.

Tags
safety