- There’s a new AI-driven land speed record and it’s so close to 200 mph (320 km/h).
- The team responsible for the record held the previous one too and beat it by about 5 mph.
- Autonomous driving’s capability shines in pushing limits, evident in these high-speed tests.
High-speed autonomous driving tech isn’t a dream of the future. It’s happening right now and a Maserati MC20 just proved it by going 197.7 mph (318 km/h) without a driver. It’s the second time this car has set a speed record and the second time that this same team has set a new land speed record for AI driving tech. Here are all the details and why this is an important milestone.
Italy’s largest science-tech university, Politecnico di Milano, joined forces with the Indy Autonomous Challenge to set this record. They accomplished it at the Kennedy Space Center, and specifically, the Launch and Landing Facility where the Space Shuttles used when they returned from their mission. This runway was ideal because it provides some 15,000 feet to run a car to super-high speeds.
More: Dodge, Jeep, Alfa Could Get Level 3 Autonomy Before Tesla With New STLA AutoDrive
In this case, the MC20 leveraged software from the PoliMOVE-MSU team, which is part of the performance division of AIDA (Artificial Intelligence Driving Autonomous of Politecnico di Milano). Video captured during the test shows the Maserati flawlessly tracing a straight line down the runway and accelerating throughout.
It slowly but surely climbs to 197.7 mph before having to hit the brakes before the runway ran out. That beats the old record of 192.8 mph quite handily. Why take the time to make an autonomous car go so fast when Level 3, 4, and 5 autonomous are relatively rare today? The answer is simple: it’s all about learning.
“These world speed records are much more than just a showcase of future technology; we are pushing AI-driver software and robotics hardware to the absolute edge. Doing so with a streetcar is helping transition the learnings of autonomous racing to enable safe, secure, sustainable, high-speed autonomous mobility on highways”, said Paul Mitchell, the CEO of Indy Autonomous Challenge and Aidoptation BV.
Scientific Director of the project and Director of the Department of Electronics, Information, and Bioengineering at the Politecnico di Milano, Prof. Sergio Matteo Savaresi, elaborated on that point: “The goal of high-speed tests is to evaluate the behavior of robo-drivers in extreme conditions. These AI systems have been tested in production vehicles at legal road speeds in the Indy Autonomous Challenge races since 2021.
“The AIDA team used this test to push the boundaries of autonomous driving, improving safety and reliability. Conducted in controlled environments without a human driver, the test assesses the AI’s stability, robustness, and reaction time, ultimately enhancing safety for low-speed urban mobility situations.”
One might consider this sort of testing as similar to that of the way that racing technology sometimes ends up trickling down into real everyday street cars over time. We might not be using autonomous software to drive at triple digit speeds today, but these kinds of tests could help to enable it safely in the future.