San Francisco Blackout Exposes the Limits of Driverless Mobility

A widespread power outage in San Francisco has highlighted a critical vulnerability in today’s autonomous transport systems. As traffic lights failed and connectivity faltered across large parts of the city, Waymo temporarily suspended its robotaxi operations after several driverless vehicles came to a halt in active traffic lanes and near major intersections.

The outage, triggered by a fire at infrastructure operated by Pacific Gas and Electric, reportedly left around a third of residents without electricity. With signalised junctions suddenly rendered inoperative, normal traffic flow gave way to confusion—an environment in which human drivers adapted quickly, but autonomous vehicles struggled to proceed.

Driverless Cars Meet an Unscripted City

Images and videos shared by residents show clusters of empty robotaxis stopped at critical junctions, some obstructing crosswalks and exits. The pattern was consistent: where traffic control became informal, through hand signals from police officers or improvised four-way stops, autonomous vehicles defaulted to caution.

Human drivers rely on eye contact, gestures, and experience to negotiate such moments. Driverless systems do not. Designed to prioritise safety above all else, they operate within strict confidence thresholds. When those thresholds are breached, stopping is often the only permissible response.

A Stress Test for Urban Autonomy

The San Francisco blackout did not prove that autonomous vehicles are unsafe. It demonstrated something more nuanced: that autonomy performs best in environments that remain predictable. Cities, however, rarely are.

As driverless technology continues to expand, resilience—not novelty—will increasingly define success. The ability to respond to disorder, rather than simply to rules, may be the true benchmark of readiness.

For now, the blackout stands as a reminder that even the most advanced mobility systems are only as robust as the cities they operate in.