Two of five driverless cars crashed while reaching speeds of 155 mph during a race at Imola’s Treacherous Rivazza corners, underscoring the current technological barriers in autonomous vehicle perception, decision-making, and control under extreme conditions.
- Two cars crashed at the difficult Rivazza corners
- Sensor failures and limited fallback localization revealed
- Race highlights gaps between perception and physical response
What happened
During the 2026 autonomous race at Imola’s Autodromo Internazionale Enzo e Dino Ferrari, two driverless cars crashed in the final sequence of corners, known as Rivazza. These corners are notably difficult even for professional Formula 1 drivers. The race consisted of five competing cars, but only two finished due to these incidents and mechanical issues. The race was conducted under challenging weather, including rain and hail, with only nine days of prior testing.
One car experienced a total sensor failure, losing lidar and radar data and relying solely on GPS, which was insufficient for accurate localization at racing speeds. This forced the vehicle to initiate a safety stop. Another car was unable to avoid collision despite detecting the slowed vehicle ahead, as split-second decision and physical braking limits made avoidance impossible. These crashes illustrated the inherent difficulties autonomous systems face on complex circuits.
Why it matters
The Imola event demonstrated that while autonomous driving technology has advanced, there are critical vulnerabilities when operating in high-speed, unpredictable environments. Perception systems can fail or provide incomplete information, leading to compromised decision-making. Furthermore, the physical constraints of vehicle dynamics limit reaction options even if hazards are recognized promptly.
This race acts as a real-world laboratory, pushing autonomous systems beyond simulation and controlled tests to understand practical limitations. The findings have implications not just for motorsport, but also for the future deployment of driverless cars on public roads, where rapid and reliable hazard response is vital for safety.
What to watch next
Future autonomous racing events will likely focus on improving sensor redundancy and fallback strategies, such as integrating additional camera and inertial measurement units to maintain localization during sensor outages. Teams will also explore enhancements in computational speed, prediction algorithms, and actuator responses to narrow the gap between hazard recognition and physical action.
Monitoring developments in the A2RL series and similar autonomous motorsport endeavors will provide valuable insights into how close driverless technology is to meeting the demands of dynamic real-world conditions—progress that could ultimately accelerate safer autonomous vehicle adoption at scale.