- Accident He starts imitating a Formula 1 with a Citroën AMI and this happens
- New Porsche 911 Dakar: flying over the dunes
The Indy Autonomous Challenge (IAC) wants to attract the best university talent in algorithms and programming to participate in the teams that develop the cars that line up in this competition, which started more than two years ago with an experimental race on the oval circuit of Indianapolis. What's new? There is no pilot on board, these are autonomous cars that are operated remotely.
That October 2021 event involved 21 universities, half of which were from North America and 10 other countries around the world (including South Korea, Germany, Canada, UAE, etc.)
Engine of 450 hp and... Many sensors and cameras
The car was the same for all: a Dallara IL-15 IACV rear-wheel drive, powered by a 450 hp internal combustion engine with 6-speed sequential transmission. The vehicle was equipped with six single cameras, four radars, three Lidar and one RTK GPS. As the main computing unit, commercial hardware consisting of an Intel Xeon 3.30 GHz, an Nvidia Quadro RTX 8000 and 64 GB of RAM was used.
The victory went to the Technical University of Munich (TUM), which brought a hefty check to Bavaria, for nothing more than one million dollars. This is possible thanks to the financial support of major global companies, as well as suppliers/sponsors such as Bridgestone (tires), Cisco (communications and networks), Continental (radars), AWS (cloud computing/data) and Luminar (Lidar, the most powerful of radars).
Universities that "train pilots"
The cars (similar to those used in the Indy Light Series, the promotional formula of Formula Indy) are all the same and no changes can be made to the engine, chassis, software, sensor set (radars, lidars, cameras and sensors) or their integration, which is periodically improved.
Then, college students work on driverless race cars to improve Artificial Intelligence (AI) algorithms, improving the "driver" as much as possible. Between the action of this "driver" and the vehicle supplied to each of the teams, it is expected that the Dallara IL15 will be as fast as possible and that they can avoid colliding with each other or against the walls of the oval.
Speed record for autonomous vehicles
The announcement of the start of this initiative was made at the Consumer Electronics Fair (CES) in 2021, but only in January 2022 was the first race held in which the cars were on track simultaneously. And again in January 2023, for another race on the final day of CES this year, at the Las Vegas circuit.
Meanwhile, two races have been held in 2022 (one of them, in November, debuting at Texas Motor Speedway with very challenging weather conditions, such as high winds, rain and very low temperatures) and a new speed record for autonomous vehicles was achieved, on April 22 of last year at the Kennedy Space Center in Florida.
At almost 310 km/h
With 30% more power compared to the first generation, the AV21 reached an incredible speed of 2022.309 km/h over a distance of 3 meters in this straight-line test at the end of 1000. The same technical basis helped the PolimOVE team (of Politecnico di Milano) win the race at Las Vegas Motor Speedway as part of the CES 2023 program, during which it set another new record for driverless vehicles: 289 km / h, lower therefore, but on the circuit.
And, by the end of 2023, all the major suppliers mentioned will help execute the transition from AV21 to AV23, with improvements to Dallara's chassis, computing power, data analytics, more advanced sensor generation, etc.
The European premiere will be in Monza in June
But one of the big news announced at CES this year was that the IAC will come to Europe for two races at the legendary Monza circuit in Italy. In 2023 (in June) the event will have the exhibition costume, but in 2024 a real race will already be held, which will pose greater challenges to the teams due to the much greater diversity of the track, compared to the American ovals.
Paul Mitchell, president of IAC: A race against a human? Maybe in 2025
What are the main objectives of robot car racing? I would say there are three main ones. Attract to this project the brightest minds in the world in Artificial Intelligence (AI), who will then be able to participate more easily in other initiatives also based on AI, Machine Learning (ML), Computer Science, Sensor Fusion, etc., whose importance in the future will go far beyond the automotive sector, touching sectors such as digital health, aerospace, etc. It is the path of human progress. On the other hand, we want to gain the trust of consumers in autonomous driving technology so that they can accept it in their day to day. And finally, to evolve autonomous driving technology to the point that it is possible to use it at very high speeds: it is one thing to see a robotaxi drop someone off at the airport on an experimental trip, and another to control a car at more than 300 km / h. Two of the most famous car racing circuits in the world... The two oldest and two of the most famous, yes. Not surprisingly, this Italian circuit is known as the "temple of speed". And in addition to the new challenges that will make the teams make a qualitative leap in terms of technology (to be able to face a circuit much more varied than the ovals), the automotive industry will also be very close to us because the race will take place simultaneously with the MIMO motor show (a new fair in an open space), where almost fifty global brands will be present. In Monza, which F1 visits every year, the average speed is around 250 km / h, being able to exceed 300 km / h on the longest straight and then there are curves where it is not possible to enter more than 60 km / h and all this with accelerations / decelerations of up to 3G.One of the ways to increase the notoriety of the IAC would be to promote a duel between the best car-robot and an F1 driver first level. Is that initiative in your plans? At the moment we are going to launch a competition between sim-racers and our autonomous cars, in a virtual racing environment, but merging these two worlds. It will be important to capture the interest of many university students who usually have this type of highly developed skills. Naturally, a race against a real good driver is a dream, but you have to wait until the technology is mature enough. Perhaps around 2025 this will be possible. Do you think there will be an audience willing to pay to see a wheeled robot race? I think so, as long as it's a different kind of spectator than the one who traditionally follows motorsport. People more connected to technology than petrol heads. And in some cases the same person may have these two interests that we can help reconcile at events organized during traditional career programs. In any case, today the cars make autonomous laps at a high average speed and face "one on one" in the qualifiers. But we are still more than a decade away from the day when we will have 10 or 20 autonomous single-seaters fighting for victory in a race of 200 kilometers or more.
According to The Trust Project criteria