SAN FRANCISCO — On Christmas Day, as a powerful Pacific storm issued warnings of heavy rain and flash flooding across the Bay Area, snow giant Waymo temporarily suspended its fully autonomous robotaxi service in San Francisco and surrounding communities.
The pause, communicated to customers via the Waymo app with a terse warning citing a National Weather Service flash flood warning, comes after growing pains and public safety questions facing autonomous vehicles as they expand into complex urban environments.
The service disruption occurred amid a broader atmospheric river system that has lashed Northern California with heavy rain, damaging winds, and ongoing flood and tornado warnings — weather conditions that have already led to localized flooding, downed trees, and travel delays across the region on one of the busiest travel weekends of the year.
Second Interruption in a Week
Image Credit: Daniel Ramirez from Honolulu, USA, CC BY 2.0, Wikimedia.
This latest outage is not an isolated glitch: earlier in the week, a major power outage caused by a fire at a Pacific Gas & Electric substation knocked out power to about a third of San Francisco, leaving tens of thousands of homes in the dark and hundreds of Waymo vehicles stalled in traffic.
During that blackout, with traffic signals around the city dead, Waymo’s autonomous cars treated the lights that didn’t work as four-way stops (their standard programming) but in many cases requested remote “confirmation checks” from fleet operators before proceeding.
Videos shared widely on social media showed robotaxis motionless at intersections, hazard lights flashing, contributing to congestion as frustrated human drivers maneuver around them. City officials, including San Francisco Mayor Daniel Lurie, have urged the company to clean up its vehicles so emergency responders can more effectively navigate the outage.
Waymo’s response to that incident was to temporarily suspend service Saturday night until power was restored and traffic signals came back online. The company said it was working closely with government and emergency management officials on the outage but acknowledged the sheer scale of the outage had overwhelmed its systems.
In the days since, Waymo has begun rolling out fleet-wide software updates intended to give its autonomous vehicles a better context for handling major infrastructure failures, enabling more decisive navigation at dark intersections without over-reliance on remote human-confirmation controls that can hamper the system’s response in a mass outage.
Technology Meets Nature
Image Credit: Jonathan Hutchins, CC BY-S 2.0, Wikimedia.
The intersection of extreme weather and autonomous technology is now at the center of news circuit debates about how robotaxi systems handle real-world emergencies. Severe weather, blackouts, and other atypical road conditions are precisely the kind of edge cases that still challenge even the most advanced artificial intelligence driving systems.
“Malfunctioning infrastructure such as dead traffic lights during a blackout should be a routine problem for self-driving cars, but the incident exposed the potential safety risks when robotaxis lose contact with their remote human operators,” wrote analysts covering the December outage (as reported by Axios), noting that a major natural disaster such as an earthquake or flood could hamper an autonomous response.
San Francisco regulators, including the California Public Utilities Commission and the Department of Motor Vehicles, said they were reviewing the blackout incident for implications on future robotaxi operations.
Public Trust and Political Pressure
Public sentiment in the city was mixed. While autonomous taxis promise reduced crashes and expanded mobility — and Waymo notes that its vehicles have logged millions of autonomous miles with improved safety compared to human drivers — the prospect of robotaxis stalled during a major outage has reignited skepticism about their readiness for widespread urban deployment.
City officials and transit advocates are now asking pointed questions: What systems are in place to proactively move vehicles off busy roads? How should autonomous fleets behave during severe weather warnings? And what regulatory conditions should govern the density and timing of robotaxi operations in dense, infrastructure-dependent urban cores?
One transit expert quoted on LinkedIn suggested that regulators and industry partners reevaluate how many autonomous vehicles are allowed on city streets and ensure robust protocols for weather and emergency scenarios — possibly including backup human intervention — before full-scale deployment. “I think we need to ask ‘what is a reasonable number of [autonomous vehicles] to have in the city streets, according to the time of day, with the geography and the weather?'” they said.
As the winter storm continues to impact the Bay Area, Waymo’s Christmas Day break highlights a critical truth about autonomous systems: despite billions of miles of development and machine learning, they still operate within (and rely on) imperfect human infrastructure. The technology may be autonomous, but the environment it must navigate is anything but predictable.