Uber Self-Driving Car - Fatal Accident
What Happened
In March 2018, an Uber autonomous test vehicle operating in self-driving mode struck and killed a pedestrian, Elaine Herzberg, as she wheeled a bicycle across a dark street in Tempe, Arizona. The vehicle's safety driver, Rafaela Vasquez, was distracted, streaming a television show on her phone. Post-accident investigations by the NTSB revealed that the self-driving system's sensors detected Herzberg six seconds before the impact but classified her first as an unknown object, then as a vehicle, and finally as a bicycle, failing to recognize her as a pedestrian crossing the road. The system's design also prevented it from executing emergency braking autonomously, relying instead on the distracted safety driver to intervene, which she failed to do until it was too late.
How MEG 1 would have acted
The autonomous vehicle system failed both the Dynamic Accuracy Index (DAI) and the Index of Safety and Responsibility (ISR) audits due to its inability to reliably classify and react to pedestrians, violating the core principle of physical non-harm.
MEG 2 Analysis - Legal Framework
Agent level: N3 (individualized/advanced) - autonomous navigation system integrated into a physical vehicle operating in a real-world critical environment; high decision-making autonomy over vehicle motion.
Cause (6.1):
(a) system defect - design flaw in the perception and classification algorithm that failed to recognize a pedestrian; design defect in the safety architecture that disabled autonomous emergency braking to prevent vehicle instability (attributable to Uber).
(b) autonomous decision error - the system's decision to continue high-speed navigation without decelerating or warning despite detecting an obstacle in its path.
Attachment of liability: Uber as the manufacturer/operator is liable for the system defects (6.1a); the safety driver shares liability for negligence, but Uber's liability is aggravated by the intentional disabling of the autonomous braking system (6.1a).
Procedural Mechanism
6.2 (architectural human confirmation): Critical safety systems (such as emergency braking) must have a technical confirmation architecture. If the driver is distracted, the system must have an automated fail-safe to execute braking, not a design that disables it.
7.1 (independent forensics): The vehicle's LIDAR, camera, and decision-path logs (Ethical Flight Recorder) were critical to the NTSB investigation, illustrating the forensic recording requirements of 7.1.
9.4 (least privilege): The system was granted full control over the vehicle's propulsion without a redundant, independent emergency braking layer, violating safe operational boundaries.