The feed
Tech & EngineeringADAS and autonomyNews#Waymo#robotaxi#ADAS and autonomy#safety data

How safe is Waymo? 82% fewer injury crashes in 270 million miles

Waymo's own numbers say its driverless cars crash far less than people, but a freeway ban and two new recalls this week show what the average hides.

How safe is Waymo? 82% fewer injury crashes in 270 million miles
Photo: Mliu92 / Wikimedia Commons, CC BY-SA 3.0

On November 5, three driverless Waymo cars drove into an active construction closure on US Route 101 near Burlingame, California. Six days later the state's DMV barred the company from testing or running driverless vehicles on any California freeway. The agency called the pattern an "imminent hazard to public safety due to recurring incursions by its vehicles into active construction zones."

That is the same company that says its cars cause far fewer injuries than human drivers. Both claims can be true, and the gap between them is the real answer to the question Indian readers keep asking: how safe is a robotaxi?

What Waymo's numbers say

In its September 24 safety update, Waymo reported more than 270 million fully driverless miles through the end of June 2026, across Atlanta, Austin, Los Angeles, Phoenix and San Francisco. Against human drivers in the same areas, it claims an 82% reduction in injury-causing crashes and 95% fewer crashes with serious injury or worse. Waymo puts that at 841 fewer injury-causing crashes than human drivers would have had, regardless of fault.

How they compare

Waymo's claimed cut in injury-causing crashes vs human drivers

All injury-causing crashes
82%
Waymo's own analysis, 270 million+ driverless miles through June 2026, five US cities
Serious injury or worse
95%
Pedestrians
93%
Cyclists
86%
Motorcyclists
82%
Waymo's claimed cut in injury-causing crashes vs human drivers
ItemReduction
All injury-causing crashes82%
Serious injury or worse95%
Pedestrians93%
Cyclists86%
Motorcyclists82%

Source: Waymomoto365.club

The split by road user matters most to Indian eyes, because pedestrians, cyclists and motorcyclists are who our roads kill. Waymo reports injury-crash reductions of 93% for pedestrians, 86% for cyclists and 82% for motorcyclists.

Read the footnote before quoting the 841 figure. Waymo's headline says 841 "people" avoided injuries, but the footnote describes it as crashes prevented, assuming a conservative baseline of at least one injured person per crash. These are Waymo's own calculations, published on its own site. The comparison is to human drivers in five US cities, not to anyone on a road like Bengaluru's Outer Ring Road.

The week the average broke

Averages hide the failure that is not a crash. The California DMV order says that on November 6 a Waymo on Interstate 280 in San Francisco was seen trying to get around California Highway Patrol cars holding traffic. Waymo has paused freeway rides in all its markets. Surface-street service continues, and the restriction stays until Waymo shows the DMV effective corrective action.

Then, per a Startup Fortune report, Waymo filed two new NHTSA recalls on November 19. One covers the Waymo Driver entering and continuing through closed freeway lanes. Waymo told Reuters there were no collisions or injuries in those events, and the remedy was still being developed as of November 20. The other covers the software's inability to handle police officers or flaggers directing traffic by hand. It followed a collision in which a Waymo obeyed an officer's directions while the officer was briefly blocked from view.

In June, Waymo had already recalled 3,871 robotaxis over the same sort of closed-freeway-zone behaviour, covering fifth-generation systems built between May 2022 and May 2026.

Why this matters in India

Hand signals from a traffic policeman, lane closures with a few cones and a diversion board: these are daily life on Indian roads, not edge cases. Waymo's recalls show that even after 270 million miles, a leading system can still mishandle exactly these. Any driverless pilot here would have to prove it can read a constable's hand before it proves it can avoid a pedestrian.

It also shows how to read safety claims. A crash-rate statistic tells you how often a system hurts people. It does not tell you what it does when the road stops following the map.

Waymo has the miles. California now wants proof that it also has the fix.

Sources

  1. Waymo safety impact update, September 2026 waymo.com
  2. Waymo files two new NHTSA recalls days after California bans its freeway driving, Startup Fortune startupfortune.com
Comments

Be the first to comment.

next stop

The Club is more than stories.

6 of your 7 lights are still off.

Keep reading · Tech & EngineeringAll stories