Autonomous racing is a curious and somewhat divisive topic. To many fans, removing the driver from the equation is anathema, literally stripping the soul out of the sport. However, the clue can be found in the name ‘motor’ sport. It is, and always has been, a competition between drivers and machines. As such, racing has always been a battle of engineers as much as drivers, and herein lies its appeal for the development of autonomous technologies.
Pivotal to achieving reliable and safe autonomous vehicle operations on public roads is being able to deal with what are known as edge cases, the terminology used to describe unexpected events. Edge cases can take a variety of forms, from handling adverse weather to erratic behaviour by pedestrians. Motorsport is a development accelerator and, if an autonomous vehicle’s AI can handle such decisions at the very limits of its abilities in the heat of competition, this should translate to improved performance on road. Enter the Indy Autonomous Challenge (IAC), a competition that pits autonomous vehicles against each other around some of the USA’s great oval tracks.
Paul Mitchell is president and CEO of IAC, and also of Energy Systems Network (ESN), an Indianapolis-based not-for-profit organisation working on developing an integrated energy ecosystem through cross-industry collaborations. ESN is also the lead backer of the competition. According to Mitchell, inspiration for the IAC stemmed from the DARPA Grand Challenges of 2004 and 2005, which saw autonomous vehicles attempting to traverse a desert course in California, competing for a prize put up by the US Defense Advanced Research Projects Agency.
Racecar covered these events in detail at the time and, in 2018, Mitchell and others at ESN felt a similar challenge, harnessing the appeal of motorsport, could help drive current autonomous vehicle development.
Brightest minds
With the kernel of an idea brewing, Mitchell invited Sebastian Thrun, who headed the original DARPA Grand Challenge-winning team from Stanford University, to visit the 2018 Indy 500. After DARPA, Thrun went on to found Google’s self-driving operation, which became Waymo, and now heads up Kitty Hawk, a company developing remote piloted air taxis.
If an autonomous vehicle’s AI can handle such decisions at the very limits of its abilities in the heat of competition, this should translate to improved performance on road
‘He was very insistent that we need something like that again. We needed to find a
