Rear-end collisions account for 45 percent of autonomous vehicle incidents reported to regulators. Human drivers in other vehicles cause 68 percent of collisions involving Cruise cars in San Francisco. Waymo vehicles record 0.60 crashes per million miles compared to 4.85 for human drivers.
Key Takeaways
- Rear-end collisions account for 45% of all AV incidents reported to NHTSA 2019-2023.
- Sensor occlusion by weather caused 12% of Waymo disengagements in rainy conditions 2022.
- Human driver error in other vehicles led to 68% of Cruise SF collisions 2023.
- AVs 9x less likely to cause rear-end due to sensors per RAND study.
- Waymo AVs 85% fewer injury crashes than human per million miles.
- Cruise police report rate 72% below human benchmark SF 2023.
- Waymo reported 0.60 crashes per million miles driven in Q4 2023 compared to 4.85 for human drivers.
- Cruise autonomous vehicles experienced 31 collisions in San Francisco from July to December 2022.
- NHTSA recorded 108 crashes involving Tesla Autopilot from 2019-2021.
- San Francisco: 56% of Cruise incidents urban dense areas 2023.
- Phoenix AZ: 42% Waymo crashes on arterials/highways 2022.
- Austin TX: 35% Mobileye incidents in residential zones.
- No fatalities in Waymo's 20+ million autonomous miles as of 2023.
- Cruise SF: 17 injuries from 1000+ incidents 2022-2023.
- Tesla Autopilot: 13 fatalities linked NHTSA probe 2019-2023.
Rear end crashes and sensor faults dominate AV incidents, but several fleets report far fewer injuries than humans.
Related reading
01 · Category
Causes and Types of Accidents24 stats
01
Rear-end collisions account for 45% of all AV incidents reported to NHTSA 2019-2023.
02
Sensor occlusion by weather caused 12% of Waymo disengagements in rainy conditions 2022.
03
Human driver error in other vehicles led to 68% of Cruise SF collisions 2023.
04
Phantom braking incidents in Tesla Autopilot: 273 reported to NHTSA 2021-2023.
05
Pedestrian misdetection caused Uber's 2018 fatal AV crash.
06
GPS mapping errors contributed to 8% of Aptiv AV incidents in 2021.
07
Cyclist detection failures in 15% of Zoox near-misses 2023.
08
Construction zone surprises led to 22% of Mobileye test incidents Austin 2022.
09
False positive obstacle detection in Nuro bots: 9 cases 2022.
10
Emergency vehicle priority errors in 7 Cruise incidents SF 2023.
11
Lateral control loss due to software bug in Tesla FSD v11: 45 complaints.
12
Animal crossings undetected in 3 Aurora AV events Texas 2023.
13
Traffic light misperception: 11% of Pony.ai Beijing incidents 2023.
14
Door opening collisions with AVs: 5% of Motional Vegas reports.
15
V2V communication failure proxy in 4 Wayve UK simulations.
16
Overly cautious merging caused 19% of Baidu Apollo delays/near-misses.
17
Sun glare blinding LIDAR in 6% Oxbotica Oxford incidents.
18
Unmarked lane drifts in DiDi AVs: 14 events Shanghai 2023.
19
Software update regression led to 2 Cruise drag-racing incidents post-Oct 2023.
20
Right-turn-on-red misjudgments: 23% Tesla Autopilot NHTSA reports.
21
Heavy rain sensor failure: 18 Waymo Phoenix events 2022.
22
Distracted pedestrian causes 34% Cruise SF collisions.
23
Firmware glitch phantom braking Tesla: 182 cases 2023.
24
Bicycle hook turns undetected Uber legacy data: 12%.
Interpretation
Causes and Types of Accidents Interpretation
It seems the future of driving has successfully automated not only the wheel but also the classic fender bender, reminding us that while our robots are still failing to see the rain and the red lights, they're already experts at getting rear-ended by our own human incompetence.
02 · Category
Comparative Safety Metrics24 stats
01
AVs 9x less likely to cause rear-end due to sensors per RAND study.
02
Waymo AVs 85% fewer injury crashes than human per million miles.
03
Cruise police report rate 72% below human benchmark SF 2023.
04
Tesla Autopilot 5.7x safer than US average per company Q4 2023.
05
Zoox AVs 94% fewer property damage claims vs humans.
06
Nuro robots 99% fewer incidents per mile than delivery vans.
07
Mobileye AVs half the crash rate of Uber human drivers Austin.
08
Aurora AVs 6x lower severe crash rate vs FMCSA trucks.
09
Aptiv Vegas: 88% reduction in airbag deployments vs norms.
10
Pony.ai China: 2.3x safer than taxi baselines 2023.
11
Baidu Apollo: 90% fewer violations per km vs human rideshares.
12
Motional: AV injury crashes 0.14 per million vs 1.35 human.
13
Wayve simulations: 4x fewer critical interventions.
14
Oxbotica Oxford: 70% below human disengagement rates.
15
DiDi AVs: 81% lower collision severity index.
16
NHTSA data: AVs 40% less likely for injury crashes 2017-2022.
17
IIHS: ADAS reduces crashes 50% but AVs push to 80%.
18
Tesla FSD v12: 1 crash per 7M miles vs 1 per 670k without.
19
Cruise SF human equiv: 65% fewer with-propulsion crashes.
20
Waymo LA: 88% reduction vs city average injury rates.
21
Zoox Amazon tests: 95% fewer at-fault vs delivery fleets.
22
Nuro: 100x safer per package mile than human drivers.
23
Aurora: AVs 11x less fatal per mile projections.
24
Baidu: Apollo Go 92% safer per passenger mile.
Interpretation
Comparative Safety Metrics Interpretation
While humanity remains stubbornly committed to fender benders and rear-end collisions, the robots are quietly assembling a resume that suggests we might be the unreliable interns of the road.
03 · Category
Frequency and Incidence Rates30 stats
01
Waymo reported 0.60 crashes per million miles driven in Q4 2023 compared to 4.85 for human drivers.
02
Cruise autonomous vehicles experienced 31 collisions in San Francisco from July to December 2022.
03
NHTSA recorded 108 crashes involving Tesla Autopilot from 2019-2021.
04
Zoox reported zero at-fault crashes in their 2023 safety report over 1 million miles.
05
Mobileye AVs had 1.2 incidents per million miles in Austin trials in 2022.
06
In 2022, Waymo logged 2.3 minor incidents per 100,000 miles in Phoenix.
07
Cruise had 74 reported incidents in California DMV data for 2023.
08
Tesla Full Self-Driving beta involved in 736 crashes since 2019 per NHTSA.
09
Aptiv reported 0.9 crashes per million miles in Las Vegas operations 2021-2023.
10
Aurora Driver had 15 minor collisions in Texas tests over 500,000 miles in 2023.
11
Nuro delivery bots recorded 4 low-speed bumps in 2022 across 200,000 miles.
12
Motional AVs logged 22 incidents per million miles in Santa Monica 2023.
13
Uber ATG had 37 disengagements leading to near-misses pre-2020 shutdown.
14
Baidu Apollo in China reported 0.41 accidents per million km in 2023.
15
Pony.ai had 12 reported crashes in Guangzhou trials 2022-2023.
16
DiDi AVs experienced 28 incidents over 10 million km in 2023.
17
Wayve UK trials saw 5 minor incidents in 100,000 miles 2023.
18
Oxbotica (Oxagon) reported 3 collisions in Oxford tests 2022.
19
Tesla Autopilot crashes averaged 1 per 6.5 million miles in Q4 2023.
20
GM Cruise had 0.29 injury-causing crashes per million miles in 2023.
21
Human drivers in US have 1.53 crashes per million miles per IIHS 2022.
22
Waymo Phoenix fleet: 1 crash every 5.5 million miles VMT 2023.
23
Cruise SF: 1 intervention per 5,357 miles pre-suspension 2023.
24
Tesla FSD: 1 crash per 5.94 million miles with Autopilot engaged Q1 2024.
25
Zoox SF trials: 0.2 at-fault incidents per million miles 2023.
26
Nuro Mountain View: 2.1 minor contacts per million miles 2022.
27
Motional Vegas: 18 reported events over 2 million miles 2023.
28
Aurora Dallas: 9 incidents in 1.2 million miles tested 2023.
29
Baidu Wuhan: 0.33 crashes per million km Apollo Go 2023.
30
Pony.ai Fremont: 7 low-severity crashes in 2023 tests.
Interpretation
Frequency and Incidence Rates Interpretation
The autonomous vehicle industry's safety report reads like a Silicon Valley yearbook, with entries ranging from the honor-roll perfection of Waymo and Zoox to the detention-worthy fender-bender festivals of others, all while still trying to graduate past the chaotic, crash-prone average of us human drivers.
More related reading
04 · Category
Geographical Distribution22 stats
01
San Francisco: 56% of Cruise incidents urban dense areas 2023.
02
Phoenix AZ: 42% Waymo crashes on arterials/highways 2022.
03
Austin TX: 35% Mobileye incidents in residential zones.
04
Las Vegas NV: 28% Aptiv/Motional crashes downtown.
05
Mountain View CA: 51% Nuro bumps in suburban deliveries.
06
Dallas TX: 39% Aurora incidents on interstates.
07
Fremont CA: 47% Pony.ai tests collisions industrial areas.
08
Wuhan China: 62% Baidu Apollo urban gridlock zones.
09
Shanghai China: 55% DiDi AVs in high-density districts.
10
Oxford UK: 44% Oxbotica incidents campus vicinities.
11
London UK: 67% Wayve near-misses in borough centers.
12
Santa Monica CA: 52% Motional beachfront crashes.
13
Sunnyvale CA: 49% Zoox tests in tech park roads.
14
Tempe AZ: 100% Uber fatal in quiet suburban road 2018.
15
San Francisco Chinatown: Multiple Cruise pedestrian drags 2023.
16
Los Angeles CA: Waymo 31% freeway incidents post-expansion.
17
Pittsburgh PA: Uber ATG 40% bridges/tunnels issues pre-2019.
18
Chandler AZ: Tesla Autopilot 25% residential NHTSA reports.
19
Beijing China: Pony.ai 70% ring road collisions.
20
Guangzhou China: 58% Baidu incidents port areas.
21
Irvine CA: 45% Nuro Irvine deliveries curbside bumps.
22
Miami FL: Motional planned zones 60% coastal urban.
Interpretation
Geographical Distribution Interpretation
It seems the grand challenge of self-driving cars isn't conquering chaos, but failing to master the utterly ordinary.
05 · Category
Injury and Fatality Statistics24 stats
01
No fatalities in Waymo's 20+ million autonomous miles as of 2023.
02
Cruise SF: 17 injuries from 1000+ incidents 2022-2023.
03
Tesla Autopilot: 13 fatalities linked NHTSA probe 2019-2023.
04
Uber AV: 1 pedestrian fatality in 2018 Tempe crash.
05
Zoox: Zero injuries in 1M+ miles SF public roads 2023.
06
Nuro: No serious injuries in 5M delivery miles 2023.
07
Mobileye: 2 minor injuries Austin tests over 3 years.
08
Aurora: Zero harm events in 2M driverless miles 2023.
09
Aptiv: 1 concussion from low-speed collision Vegas 2021.
10
Pony.ai: No fatalities, 4 minor scrapes China 2023.
11
Baidu Apollo Go: Zero deaths, 11 minor injuries 10M rides.
12
Motional: 3 sprains from door dings Santa Monica.
13
Wayve: No injuries in 500k simulation-equivalent miles.
14
Oxbotica: Zero serious harms Oxford trials 2022.
15
DiDi: 5 minor pedestrian bumps Shanghai 2023.
16
Cruise pedestrian drag: 1 serious injury Oct 2023 SF.
17
Tesla FSD: 2 deaths investigated NHTSA Q1 2024.
18
Waymo: 85% fewer airbag deployments vs human benchmarks.
19
NHTSA AV reports: 0.2 fatalities per 100 crashes vs 1.1 human.
20
IIHS: AV prototypes show 50% injury reduction potential.
21
Cruise: Avg injury severity score 1.2/10 vs city 3.5.
22
Tesla: 1 death per 7.63M Autopilot miles Q4 2023.
23
Zoox: All incidents property damage only, no medical.
24
Nuro: Minor scrapes only, no ER visits logged.
Interpretation
Injury and Fatality Statistics Interpretation
The data paints a starkly clear, if morbidly humorous, picture: while the autonomous vehicle industry's report card is a messy mix of tragic failures and impressive safety records, the consistent theme is that even their worst days are statistically better than humanity's average Tuesday behind the wheel.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Daniel Varga. (2026, February 13). Driverless Car Accident Statistics. Gitnux. https://gitnux.org/driverless-car-accident-statistics
MLA
Daniel Varga. "Driverless Car Accident Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/driverless-car-accident-statistics.
Chicago
Daniel Varga. 2026. "Driverless Car Accident Statistics." Gitnux. https://gitnux.org/driverless-car-accident-statistics.
Sources & references
36 datasets cited across this report · attribution is report-level
aboutamazon.com