A TuSimple autonomous truck crashed into a concrete barrier on I-10 in Arizona when the AI system executed an outdated 2.5-minute-old left-turn command at 65mph due to improper system initialization.
On April 6, 2022, a TuSimple autonomous semi-trailer truck traveling at 65mph on I-10 in Tucson, Arizona suddenly veered left and crashed into a concrete divider. The truck was equipped with TuSimple's self-driving technology including 20 cameras and two lidar sensors, with a safety driver and test engineer aboard. According to internal company reports, the crash occurred because the autonomous driving system had not been properly rebooted before engagement, causing it to execute an outdated left-turn command that was 2.5 minutes old. TuSimple initially blamed human error, but autonomous vehicle researchers argued that proper safeguards should have prevented the system from responding to commands even hundredths of a second old, and should never have allowed sharp turns at highway speeds. The company reported no injuries or property damage from the incident. Following the crash, federal regulators including the Federal Motor Carrier Safety Administration and NHTSA launched safety compliance investigations. TuSimple modified its system after the incident to prevent engagement unless the computer system is fully functional, though former employees said this change was overdue.
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.
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