A widespread power outage across SanFrancisco over the weekend exposed a surprising vulnerability in the city’s fleet of autonomous vehicles operated by Waymo, Alphabet’s self-driving car unit. As traffic lights went dark and infrastructure faltered, dozens of Waymo robotaxis stalled at intersections and clogged roadways, forcing the company to suspend service temporarily and prompting questions about how autonomous systems handle real-world disruptions.
The blackout triggered by a fire at a Pacific Gas & Electric (PG&E) substation left about 130,000 homes and businesses without electricity, knocking out traffic signals and plunging large swaths of the city into darkness.
The outage, which began Saturday morning and persisted to varying degrees across multiple days, created widespread disruption across San Francisco’s transportation network. With traffic lights out at key intersections, gridlock quickly unfolded as drivers attempted to navigate intersections manually. In the midst of these challenges, Waymo’s autonomous vehicles struggled to adapt, often stopping abruptly sometimes in the middle of intersections rather than pulling aside.
Videos shared on social media showed Waymo cars with their hazard lights blinking, halted in places that created choke points and complicated traffic flow. In some cases, human drivers were forced to maneuver around the robotaxis to reach their destinations.
Many of Waymo’s robotaxis are designed to treat a non-functioning traffic signal as a four-way stop, a standard fallback behavior in autonomous driving logic. However, the scale and widespread nature of the blackout appear to have overwhelmed those programming assumptions, resulting in situations that left the cars stationary far longer than anticipated.
Service Suspended and Resumed
Facing a mounting traffic problem, Waymo temporarily suspended its San Francisco service Saturday evening, pulling cars from active service as the outage continued. By Sunday afternoon, the company announced that services had resumed after a pause of approximately one day.
Waymo says it completed most active trips safely before directing vehicles back to depots or pulling them over as conditions deteriorated. A company spokesperson emphasized that the blackout was a significant infrastructure failure, but insisted Waymo was already working to enhance how its technology adapts to such events.
Passenger and Public Frustration
Residents and commuters in the city found themselves caught up in the disruption. One local who had intended to leave for the airport described a scene where multiple Waymo cars were clustered at a traffic light outside his apartment, blocking his route and forcing rideshare drivers to weave through the stalled fleet.
The incident revived long-standing conversations about the readiness of autonomous vehicles to handle unpredictable conditions and raised fresh concerns among some San Franciscans who have previously questioned the safety and adaptability of robotaxi technology. (L
The California Public Utilities Commission (CPUC) which regulates Waymo’s autonomous taxi operations along with the California Department of Motor Vehicles confirmed that it is reviewing the incidents in greater detail. Officials are examining how and why so many robotaxis stalled simultaneously and whether the company’s contingency planning for infrastructure failures was adequate.
This scrutiny comes as autonomous vehicle operators across the state are under heightened public and regulatory attention, with safety incidents, unexpected operational behaviors, and debates over liability and readiness often making headlines.
Transportation experts point out that autonomy logic typically errs on the side of caution: when a self-driving car encounters confusion or uncertainty such as failing traffic lights, it is programmed to slow down or stop until it can verify the situation safely. But in large-scale power outages, that logic can inadvertently create hazards of its own by leaving vehicles immobilized where they disrupt intersections and emergency routes.
One expert described the situation not as a pure software failure, but as an “operational management failure”, where the number of vehicles needing assistance exceeded Waymo’s ability to centrally intervene or reroute them effectively.
City officials, including San Francisco Mayor Daniel Lurie, reportedly contacted Waymo leadership directly to request the removal of robotaxis from key intersections during the crisis, underscoring the severity of the situation.
Critics argue that this episode highlights a significant limitation of current autonomous systems: their reliance on functioning infrastructure like traffic signals, GPS systems, and cellular connectivity. When those systems fail — whether due to a blackout, natural disaster, or other disruptions self-driving cars may lack the flexibility that human drivers apply instinctively.
Public safety advocates and some local leaders are urging greater oversight, improved contingency features, and transparent reporting from autonomous vehicle companies to ensure that these services do not inadvertently endanger other road users or impede emergency response teams during crises.
Waymo’s Response and Next Steps
Waymo has stated that it will integrate lessons learned from the blackout into its technology and procedures, and that it maintains 24/7 emergency communication channels to assist in unusual circumstances. The company has also reiterated its commitment to enhancing how its vehicles interpret and respond to complex traffic scenarios.
While the power outage was a disruptive event for many San Franciscans, no serious injuries related to the robotaxis were reported. Still, the incident serves as a stress test for autonomous system resilience, underscoring the challenges that lie ahead as self-driving services scale up in urban environments dependent on robust infrastructure.
The San Francisco blackout and its impact on Waymo’s robotaxis have raised important questions about how autonomous vehicles behave when the unexpected happens. From technical limitations to emergency readiness and policy oversight, the episode illustrates the complexity of integrating autonomous systems into everyday life and how far they still have to go before they can fully navigate the unpredictable realities of urban transportation.
