A Waymo autonomous vehicle was involved in a crash with a bicyclist in downtown San Francisco on February 6th, the latest incident for the self-driving company operating in the city.

According to Waymo’s statement, their vehicle had come to a complete stop at a 4-way intersection. A large truck passed through the intersection, blocking the view of the bicyclist following closely behind. When the bicyclist suddenly came into view, Waymo’s vehicle braked but was unable to avoid a collision.

The cyclist reportedly suffered only minor abrasions and left the scene on their own accord. Waymo says they contacted authorities about the event.

The incident highlights the challenges autonomous vehicles still face with detecting vulnerable road users like cyclists and anticipating their movements. Over 130,000 bicyclists are injured every year in the U.S., underscoring the need for advanced detection and careful driving around them.

Waymo says its vehicles are trained to drive safely around cyclists, using 360-degree sensors and software to understand the surroundings. But real-world unpredictability continues to test the limits of self-driving systems.

The crash also comes amid turbulence for autonomous vehicle operators in San Francisco. Competitor Cruise had its permits revoked last October after failing to promptly share footage with regulators of a crash where their vehicle dragged a pedestrian.

And the city recently filed a lawsuit aiming to reduce the number of robotaxis, pending review of the state regulators’ decision to grant Waymo and Cruise expanded driverless permits.

While AV companies tout progress in structured testing, chaotic urban environments like San Francisco continue to expose flaws. As software and sensors improve, it still may be years until self-driving cars can match human perception and decision-making across all driving scenarios. This latest Waymo crash serves as another reminder of the work left to reach that goal.