Waymo Pulse SF

About This Project

What began as an experiment to test Google's Gemini 2.5 Pro's video analysis skills—identifying car details like color and make from challenging footage—quickly evolved. Encouraged by initial successes, the project took on a real-world challenge: monitoring the frequency of Waymo autonomous vehicles in San Francisco's Lower Pacific Heights. A Raspberry Pi did capture constant 5-minute video loops, which Gemini meticulously analyzes to produce the detailed statistics presented on this site, refined through several iterations of prompting and evaluation leading to interesting insights and stats you can see below.

Read the full story

Overall Car Analysis

Waymo Sightings

Spotted 4069 Waymo vehicles.

This makes up 2.92% of all identified cars.

General Statistics

Total cars analyzed: 139219

Motorbikes: 3260 (2.07%)

Pedestrians: 10741 (6.82%) (514 with dogs - 4.79%)

Cyclists: 304 (0.19%)

Total objects detected: 157593

Car Color Distribution (Top 10)

White: 36309 (26.08%)
Black: 24049 (17.27%)
Grey: 20372 (14.63%)
Silver: 17812 (12.79%)
Dark Grey: 17658 (12.68%)
Red: 6484 (4.66%)
Blue: 6421 (4.61%)
Dark Blue: 3763 (2.7%)
Yellow: 1080 (0.78%)
Light Blue: 779 (0.56%)
Others: 4492 (3.23%)

Car Make Distribution (Top 10)

Toyota: 27402 (19.68%)
Honda: 13580 (9.75%)
Tesla: 11967 (8.6%)
Ford: 10758 (7.73%)
Unknown: 8049 (5.78%)
Subaru: 6872 (4.94%)
Bmw: 6275 (4.51%)
Lexus: 6132 (4.4%)
Mazda: 6009 (4.32%)
Mercedes-benz: 3718 (2.67%)
Others: 38457 (27.62%)

Car Age Distribution

0-2 Years: 1368 (0.98%)
2-4 Years: 28934 (20.78%)
4-6 Years: 54007 (38.79%)
6-8 Years: 31694 (22.77%)
8-10 Years: 11699 (8.4%)
10+ Years: 11517 (8.27%)

Car Value Distribution

$0-$10,000: 6710 (4.82%)
$10,000-$20,000: 40205 (28.88%)
$20,000-$30,000: 46391 (33.32%)
$30,000-$40,000: 26222 (18.84%)
$40,000-$50,000: 12519 (8.99%)
$50,000-$60,000: 2365 (1.7%)
$60,000-$70,000: 838 (0.6%)
$70,000-$80,000: 509 (0.37%)
$80,000-$90,000: 245 (0.18%)
$90,000-$100,000: 180 (0.13%)
$100,000+: 3035 (2.18%)