Roborace tackled Goodwood's famous hill climb with its self-driving electric race car, known as the Robocar
The Goodwood Festival of Speed is well known for courting some of the world’s finest automobiles and high-performance machines. This year, something very different was in attendance.
The Robocar is no the same as the DevBot prototype we’ve seen in the past; Robocar is closer to a finished product and bears the design shown in the concept vehicle unveiled in 2017.
To maneuver the 1.16-mile course and make its way through hay bales, flint walls and forests found on the Goodwood estate, the driverless Robocar relied on an assortment of lidar, radar, GPS, ultrasonic, and camera sensors. An Nvidia Drive PX 2 computer processes all of the data collected and then specific algorithms calculate the correct steering, acceleration and braking steps — all within the fraction of a second.
As we saw in the video of a test session (the final run is shown below), the Robocar doesn’t simply crawl up the hill. Onboard are four electric motors working to deliver over 500 horsepower. Roborace itself simply provides the Robocar with an API (application programming interface) and then hands off the self-driving car to whomever to install their own driving algorithms.
That honor this time went to Arrival, another automotive technology company.
Roborace has big plans for the future with its self-driving racer. Eventually, the company will spawn a race series expected to coincide with Formula E. The cars will all feature identical hardware and electric powertrains, but teams will duke it out via algorithms to see who can create the quickest self-driving racer.
It’s unclear when the inaugural race will take place.
The Robocar wasn’t the only self-driving car to take the hill climb this year. A 1965 Ford Mustang with self-driving technology from Siemens and Cranfield University was also present, though the run was hardly what you’d call successful. It’s pace was more of a crawl and a human driver behind the wheel needed to make numerous corrections.