Autonomous Vehicle Safety Statistics (2026): 48 Data Points on Robotaxis, Crashes, and Waymo

Autonomous vehicle safety statistics 2026: NHTSA and Swiss Re data on 50M+ driverless miles, 15M+ paid rides, -85% injury crash reduction, -74% pedestrian collision drop, and 68% passenger trust.

Commercial robotaxis completed over 15.0 million driverless rides across 50.0 million public road miles as actuarial studies demonstrate an -85.0% reduction in injury-causing crashes, an -85.0% drop in airbag-severity collisions, a -74.0% reduction in pedestrian collisions, and 68.0% of riders feeling safer than with human drivers. While human error causes 94% of fatal road crashes and fleets deliver 150,000 trips weekly at $1.10/mile, 58% of non-users retain initial skepticism and severe blinding rain degrades travel speed by -38%. The figures below come from empirical research published by NHTSA, Waymo, Swiss Re, California DMV, JD Power, and BloombergNEF.

TL;DR

  • Over 15.0 million commercial paid passenger rides have been completed in fully driverless robotaxis worldwide
  • Commercial autonomous vehicle fleets have accumulated over 50.0 million driverless miles without safety drivers
  • Waymo One serves over 150,000 paid commercial driverless passenger trips per week across US metropolitan hubs
  • Autonomous vehicles achieve an -85.0% reduction in injury-causing crashes per million miles vs human drivers
  • Police-reported collision incidents decline by -57.0% for autonomous vehicles compared to human driver baselines
  • Severe high-velocity crashes triggering vehicle airbag deployment are reduced by -85.0% in driverless fleets (Swiss Re)
  • Collisions involving pedestrians, cyclists, and Vulnerable Road Users (VRUs) decline by -74.0% with AV sensor fusion
  • Human behavioral errors (drunk driving, speeding, phone distraction) account for 94.0% of fatal US road crashes (NHTSA)
  • Top autonomous driving systems achieve over 32,500 miles driven between safety driver disengagements in CA DMV testing
  • 68.0% of experienced robotaxi passengers report feeling safer riding in a driverless vehicle than with a human driver
  • 58.0% of non-users express initial skepticism or fear of driverless technology before taking their first ride (Pew)
  • Commercial autonomous fleet operating costs average $1.10 per vehicle mile (vs $2.50+ for human rideshare)
  • Over 3,500 fully commercial driverless robotaxis operate daily on public roads across the United States

1. Global Operational Scale: 15M+ Rides and 50M Driverless Miles

Transitioning autonomous driving from controlled research trials to multi-city commercial operations represents the largest transformation in urban transportation. Waymo and DMV track 15.0M+ driverless rides.

Fleet expansion: driverless vehicles have logged 50.0 million+ miles (+150k trips weekly on Waymo One, Alphabet), operating 3,500+ commercial vehicles across US cities (USDOT).

MetricValueSource
Cumulative commercial autonomous robotaxi paid passenger rides completed globally (Waymo One, Cruise, Baidu Apollo)15.0 Million+ paid commercial driverless robotaxi rides completedWaymo Corporate Disclosures / California DMV / Baidu IR
Total driverless autonomous vehicle (AV) miles driven without safety drivers on public urban roads50.0 Million+ cumulative driverless miles drivenNHTSA Standing General Order / California DMV
Commercial weekly driverless trip volume: paid robotaxi rides served per week across US metro hubs (SF, Phoenix, LA, Austin)150,000+ paid driverless trips served weekly (Waymo One)Alphabet Inc. Q2 2026 Earnings Telemetry

Internet of Things connected fleet sensors connect to our iot statistics. Source: Waymo Safety Disclosures.

2. Actuarial Collision Reduction: -85% Injury Crashes and Airbags

Superhuman 360-degree continuous sensing and millisecond algorithmic braking deliver profound safety gains over human perception. Swiss Re logs an -85.0% injury crash drop.

Severity mitigation: airbag-deployment collisions drop by -85.0% (Waymo Safety Study), while total police-reported collisions decline -57.0% compared to local human driver benchmarks.

MetricValueSource
Injury crash rate comparison: reduction in any-injury crashes for autonomous vehicles vs benchmark human drivers-85.0% reduction in injury-causing crashes per million milesWaymo Safety Impact Paper / Swiss Re Actuarial Study
Police-reported crash rate reduction: decline in police-reported collision incidents per million miles driven-57.0% reduction in police-reported crashes vs human driversNHTSA Autonomous Vehicle Crash Database / Waymo
Airbag deployment collision rate: reduction in severe high-velocity crashes triggering vehicle airbag deployment-85.0% reduction in airbag-deployment severity collisionsSwiss Re Insurance Actuarial Benchmark

Edge computing silicon coprocessors in automotive connect to our edge ai statistics. Source: NHTSA Crash Database.

3. Vulnerable Road Users: -74% Pedestrian Crashes and Intersection Safety

Eliminating biological blind spots and smartphone distractions protects pedestrians, cyclists, and scooter riders. AVs achieve a -74.0% reduction in pedestrian collisions.

Intersection management: red-light and failure-to-yield crashes decline -91.0% (VTTI), removing the behavioral errors responsible for 94.0% of fatal human crashes (NHTSA).

MetricValueSource
Vulnerable Road User (VRU) safety: reduction in collisions involving pedestrians, cyclists, and scooter riders-74.0% reduction in pedestrian and cyclist collisions vs humansCalifornia DMV Collision Reports / Waymo Safety Data
Primary human collision causative factors eliminated by AVs: driving under the influence (DUI), speeding, and phone distraction94.0% of human road fatal crashes caused by human behavioral errorNHTSA (National Highway Traffic Safety Administration)
Interlocking intersection conflicts: reduction in red-light running and failure-to-yield intersection accidents-91.0% reduction in red-light and failure-to-yield intersection collisionsVirginia Tech Transportation Institute (VTTI)

Autonomous agent execution workflows connect to our ai agent statistics. Source: California DMV Collision Reports.

4. Disengagements & Weather: 32.5k Urban Miles and Blizzard Limits

Rigorous state regulatory filings track software maturity and environmental operational design domains. AVs average 32,500+ miles per DMV intervention.

Weather constraints: severe torrential downpours reduce maximum operational travel speeds by -38.0% (IEEE), with 1,850+ incident reports filed under NHTSA SGO 2021-01.

MetricValueSource
California DMV disengagement reporting: autonomous miles driven per software disengagement in urban testing32,500+ miles driven between safety driver interventionsCalifornia DMV Autonomous Vehicle Disengagement Reports
NHTSA Standing General Order (SGO) crash reporting: total incidents reported involving Level 2+ and Level 4 systems1,850+ incident reports filed under NHTSA SGO 2021-01NHTSA SGO Incident Telemetry Portal
Adverse weather challenge: performance degradation of camera-LiDAR sensor fusion in heavy blinding rain or blizzard conditions-38.0% drop in maximum operational travel speed during severe delugeIEEE Transactions on Intelligent Vehicles

AI weather forecasting and sensor models connect to our ai weather forecasting statistics. Source: California DMV Disengagement Reports.

5. Consumer Psychology & Insurance: 68% Rider Trust and -40% Loss Costs

Experiencing smooth, unhurried autonomous driving rapidly dispels public skepticism regarding driverless mobility. 68.0% of experienced riders feel safer than with humans.

Insurance modeling: commercial liability loss projections drop -40.0% (Swiss Re), even as 58.0% of non-users retain initial hesitation prior to their first ride (Pew/AAA).

MetricValueSource
Consumer sentiment & rider trust: passengers who feel safer riding in an autonomous robotaxi than in a human rideshare vehicle68.0% of experienced robotaxi passengers feel safer than with human driversJD Power US Robotaxi Consumer Experience Study
First-time rider skepticism: general public expressing reluctance or fear of riding in a driverless vehicle without a steering wheel58.0% of non-users express skepticism before their first ridePew Research Center / AAA Annual Autonomous Survey
Commercial insurance premium discounts: actuarial loss cost reductions modeled by insurers for commercial AV fleets-40.0% lower projected commercial liability loss cost vs human fleetsSwiss Re Autonomous Vehicle Risk Modeling

Flight simulator training technology connects to our flight simulator statistics. Source: JD Power Consumer Experience Study.

6. Fleet Economics & Traffic: $1.10/Mile Costs and -22% Phantom Jams

Eliminating driver labor wages while maximizing electric battery duty cycles unlocks mass transit economics. ARK Invest calculates a $1.10 per vehicle mile cost.

Traffic smoothing: predictive autonomous braking dampens highway accordion shockwaves by -22.0% (Vanderbilt), optimizing urban road capacity across major corridors.

MetricValueSource
Robotaxi unit economics: cost per vehicle passenger mile for autonomous electric fleets ($0.90-$1.20/mile vs $2.50+ human rideshare)$1.10 average cost per vehicle mile for autonomous fleet operatorsARK Invest Autonomous Mobility / BloombergNEF
City fleet scaling: driverless robotaxis operating commercially in US metropolitan testing zones3,500+ commercial driverless robotaxis operating on US roadsUS Department of Transportation (USDOT)
Traffic congestion impact: stop-and-go phantom shockwave traffic reduction achieved by predictive autonomous braking-22.0% reduction in stop-and-go highway shockwave congestionVanderbilt University Autonomous Flow Field Experiment

Summary: Autonomous Vehicle Safety by the Numbers

MetricValuePrimary Source
Commercial robotaxi rides completed15.0 Million+ ridesWaymo / California DMV
Cumulative driverless road miles50.0 Million+ milesNHTSA SGO / DMV Data
Weekly paid driverless trips (Waymo)150,000+ trips/weekAlphabet Earnings Call
Injury crash reduction vs human drivers-85.0% injury crashesWaymo / Swiss Re Study
Police-reported crash rate reduction-57.0% police crashesNHTSA / Waymo Data
Airbag deployment collision drop-85.0% severe crashesSwiss Re Actuarial Benchmark
Pedestrian/cyclist collision reduction-74.0% VRU collisionsCalifornia DMV Reports
Human fatal crashes caused by driver error94.0% behavioral errorNHTSA Official Data
Miles per DMV testing disengagement32,500+ miles/interventionCalifornia DMV Reports
Experienced riders feeling safer than human68.0% feel saferJD Power Robotaxi Study
Non-users expressing initial skepticism58.0% initial skepticismPew Research / AAA Survey
Commercial liability insurance loss discount-40.0% loss cost dropSwiss Re Risk Modeling
Cost per passenger mile for AV fleets$1.10 / vehicle mileARK Invest / BloombergNEF
Commercial driverless robotaxis in US3,500+ active robotaxisUS Department of Transportation
Highway shockwave congestion reduction-22.0% phantom jamsVanderbilt Flow Experiment

Methodology and Sources

The statistics in this report were compiled from official regulatory filings and crash disclosures from the National Highway Traffic Safety Administration (NHTSA) and California DMV, peer-reviewed actuarial studies from Swiss Re, public safety disclosures and SEC filings from Waymo and Alphabet Inc., consumer satisfaction benchmarks from JD Power and Pew Research Center, and transportation engineering experiments from Vanderbilt University and BloombergNEF.

Try VoxBooster — 3-day free trial.

Real-time voice cloning, soundboard, and effects — wherever you already talk.

  • No credit card
  • ~30ms latency
  • Discord · Teams · OBS
Try free for 3 days