This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
On February 6, 2024, a Waymo autonomous vehicle collided with a bicyclist at an intersection in San Francisco. Waymo stated that the cyclist was obscured by a large truck and moved into the vehicle's path, leading to a collision despite heavy braking. The cyclist sustained minor injuries and did not require hospitalization, and the incident is currently under review by local authorities and the California DMV.
A Waymo robotaxi in San Francisco reportedly failed to detect a cyclist obscured by a truck, resulting in a collision with minor injuries, at 17th and Mississippi Streets in Potrero Hill. The incident reportedly demonstrated a vulnerability in autonomous vehicles' ability to safely navigate complex urban environments.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
The incident represents a lack of capability or robustness in the AI system's perception and reaction under occlusion, as it failed to anticipate the cyclist behind the truck and avoid the collision.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The collision was caused by the decision of the deployed Waymo autonomous vehicle to proceed into the intersection, resulting in an unintentional accident post-deployment.
A driverless Waymo vehicle collided with a cyclist in San Francisco, causing minor injuries. The incident was a localized, unintentional traffic accident resulting from sensor occlusion. It has negligible national security implications, representing a standard public safety and regulatory issue managed by local authorities rather than a threat to national security.
The Waymo vehicle operates independently in real-time, responding to environmental inputs (like the truck and the cyclist) using learned patterns without explicit human intervention or complex planning. The report notes it was a 'driverless Waymo car' that 'moved into the intersection' and 'applied heavy braking' autonomously.
High Risk: The system is an autonomous vehicle operating in public transport and traffic infrastructure, which has significant implications for public safety and critical infrastructure.
The Waymo vehicle's primary function is to operate as a self-driving car, navigating public roads autonomously without a human driver.
The reports provide consistent, detailed accounts of the collision from multiple sources, including statements from Waymo, the police, and local officials. The sequence of events is clearly described, and the extent of the injuries is verified by first responders.
One could argue the cyclist was primarily at fault for turning into the vehicle's path while obscured, rather than it being a pure AI failure, but the system's inability to anticipate the hazard still points to a robustness limitation.
Technology
Sensor Occlusion by Large Truck
Vehicle sensors could not see the cyclist who was hidden behind a truck.
Inability to Predict Hidden Hazards
AV system failed to anticipate a hazard hidden behind the passing truck.
Sub-human Level Safety Systems
Academic research indicates AV safety is not yet at human-level capability.
Data Inputs
Blocked Sensor Field of View
Oncoming truck physically blocked sensor line of sight to the cyclist.
Human Factors
Cyclist Sudden Left Turn
The cyclist turned left into the AV path immediately after following the truck.
No Human Driver to Intervene
As a driverless vehicle, there was no human driver to anticipate the hazard.
Process and Methods
Inadequate Defensive Driving Logic
AV did not assume a hidden road user could emerge from behind the truck.
Management
Premature Service Expansion
Waymo seeking service expansion before solving complex occlusion scenarios.
Regulatory Environment
Unresolved AV Safety Standards
Regulators still reviewing safety risks of autonomous vehicles on public roads.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
