Autonomous race cars crash at 155 mph during A2RL event, exposing limits of self-driving tech

The Abu Dhabi Autonomous Racing League (A2RL) held its first international race at Italy’s Autodromo Internazionale Enzo e Dino Ferrari in Imola, a notoriously difficult circuit where even Formula 1 drivers have crashed.

Only two of the five participating autonomous vehicles completed the race, with PoliMove’s entrant sustaining damage after colliding with Unimore’s vehicle in the Rivazza section, a final-sequence of challenging corners.

Unimore’s car, Gianna, initiated a safety stop after losing all lidar and radar data, relying on GPS alone for localization. PoliMove’s Eva detected Gianna ahead but could not avoid impact due to perception, decision-making, and actuator delays, even though the system identified the hazard.

TUM withdrew its entrant, Hailey, after a mechanical brake failure during the formation lap, highlighting that autonomous systems cannot compensate for hardware faults.

A2RL’s organizers framed the event as a controlled laboratory to stress-test perception, prediction, and vehicle control at extreme speeds without endangering human drivers.

Teams had only nine days of physical testing, including in rain and hail, and the league’s next race may shorten testing further or restrict GPS access to increase difficulty.

Organizers and participants suggested the data and edge cases—such as sensor loss, compromised localization, and sudden braking—could inform improvements in civilian autonomous vehicle safety and hardware resilience.