I felt a strange blend of fascination and unease watching dozens of Waymo robotaxis frozen in place across San Francisco last Tuesday. The citywide power outage that plunged neighborhoods into darkness had an unexpected casualty: autonomous vehicle operations. As I navigated around a white Jaguar I-PACE with its distinctive dome sensor completely unresponsive at the corner of Market and Van Ness, I couldn’t help but wonder if we were glimpsing a preview of the vulnerabilities in our increasingly automated transportation infrastructure.
The December 15th blackout affected nearly 60,000 PG&E customers across San Francisco when a substation failure cascaded through the grid around 2:15 PM. While conventional vehicles continued operating, approximately 37 Waymo vehicles simultaneously ceased operations, creating what local residents quickly dubbed “robotaxi graveyards” at major intersections.
“We’ve never seen a failure at this scale,” said Dr. Eleanor Hwang, autonomous systems researcher at Stanford University, when I spoke with her yesterday. “It reveals how centralized command structures remain vulnerable to infrastructure disruptions despite redundancy planning.”
According to Waymo’s official statement, their vehicles “appropriately entered safe mode during the unexpected power interruption.” The company maintained that safety protocols functioned as designed, preventing any accidents. However, this assertion rings hollow for commuters like Marcus Chen, who found himself stranded when his Waymo ride suddenly pulled to the curb and stopped responding to app commands.
“I was already running late for a meeting when everything just shut down,” Chen told me. “The doors stayed locked for about four minutes before I could exit. No information, no alternative – just stuck.”
What makes this incident particularly troubling is that these vehicles should theoretically have substantial backup systems. Modern autonomous vehicles contain multiple redundant power supplies and local decision-making capabilities. Yet something in Waymo’s network architecture created a critical dependency on external infrastructure.
Transportation officials and tech observers have raised questions about why the outage triggered such a widespread operational failure. Most concerning was the revelation that Waymo’s remote monitoring system experienced a simultaneous disruption, leaving the company unable to remotely access or direct vehicles for approximately 35 minutes.
City Supervisor Catherine Rivera has called for an investigation and public hearing. “When we permit autonomous vehicles on our streets, we need absolute confidence they won’t become obstacles during emergencies,” she said at yesterday’s press conference. “This raises serious questions about disaster readiness.”
The timing couldn’t be worse for Waymo, coming just days after the company announced plans to expand its San Francisco fleet by an additional 150 vehicles in early 2026. The California Public Utilities Commission had previously approved this expansion based on Waymo’s safety record and infrastructure resilience claims.
This incident highlights the complex interdependencies between our physical and digital infrastructure. While Tesla vehicles continued operating during the same outage thanks to their more distributed architecture, Waymo’s more centralized approach proved vulnerable.
“There’s a fundamental tension between centralized control for safety oversight and distributed autonomy for resilience,” explained Dr. Raj Patel, transportation systems engineer at UC Berkeley. “Companies make different design choices that reflect their safety philosophies, and we’re witnessing the real-world testing of these approaches.”
For San Francisco residents, the sight of immobilized robotaxis blocking traffic during an already stressful citywide power outage has eroded some confidence in autonomous technology. A flash poll conducted by the San Francisco Chronicle found 63% of respondents expressing decreased trust in autonomous vehicles following the incident.
The broader implications extend beyond one company’s reputation. As cities increasingly incorporate autonomous vehicles into their transportation planning, the resilience of these systems during infrastructure disruptions becomes critical. The incident raises questions about how autonomous vehicle networks should be designed to degrade gracefully during emergencies rather than failing catastrophically.
Waymo has promised a full technical review, with spokesperson Melissa Chen telling me, “We’re committed to learning from this unprecedented event and implementing additional redundancies to ensure continuous service during infrastructure disruptions.”
Whether that will be enough to restore public confidence remains to be seen. As I watched city workers and Waymo engineers manually moving vehicles from intersections, I couldn’t shake the feeling that we’re still in the fragile early days of autonomous transportation – a technology promising reliability while demonstrating just how vulnerable our new mobility systems can be when the lights go out.
