Autopilot

A self driving race car participating in the Roborace. / Photo by: KAgamemnon via Wikimedia Commons

 

Adrenaline rush, strategic balance, and improvisations are elements experienced by human motorsport players. But can a self-driving racecar do the same thing? In the 2018 Goodwood Festival of Speed, the Roborace autonomous vehicle was able to make its first self-driving hill climb challenge.

On Friday at the 25th anniversary of the festival, Roborace successfully completed its first public attempt at a hill climb. Unlike the standard form of motorsport, a hill climb takes place in an uphill racetrack and drivers must complete the course before the time runs out. For a human, skills and training can make the feat easier to accomplish, but an autonomous vehicle requires high-tech systems to perform well.

"This environment is way more complex than a racetrack. The tracks that we've been on are very structured: track walls, Tarmac to the edges, so here it is a step beyond. It's actually closer to a road environment than ultimately what this technology will be applied to," explained Jen Horsey, CEO at Roborace.

The autonomous racing car has been equipped with multiple cameras and LIDAR sensors to monitor its surroundings. To be specific, the systems enabled it to create a map with a 360-degree vision of everything around it. After the mapping the area, the car checks any drivable surfaces, according to Autosport.

In the official website of Roborace, the design was made by Daniel Simon, who worked with vehicles in movies like Tron Legacy while the internal computing processors were provided by NVIDIA. The processor model used in Roborace was the Drive PX 2. NVIDIA DRIVE processors which are a part of the AI platform developed for AV manufacturers and startup companies.

The AV racing car will remain to complete its daily runs until the end of the festival.