When the No. 4 Corvette C6.R beat out a field of Ferraris, Porsches and Vipers at the 12 Hours of Sebring earlier this month, the ‘Vette had a leg up on the competition. It wasn’t extra horsepower or super-sticky tires that gave the drivers an advantage, but a handful of sensors and a PC running Linux.
The Pratt & Miller team used a rear-facing radar system — the same technology your adaptive cruise control uses to avoid smacking the guy in front of you — to create an identification and collision-avoidance setup that identifies and tracks cars approaching from behind and relays the information to the driver through a 7-inch, cockpit-mounted display. The system can track up to 32 cars and tell the driver how many are behind him, how quickly they’re closing in, which classification they are and — if they’re fast enough — which side of the Corvette they’ll pass on.
“So far the drivers really love it,” Chris Hammond, embedded systems engineer for Pratt & Miller, told Wired. “[Driver] Oliver Gavin said it was like having a second set of eyes.”
The system, developed over the course of six months, uses an off-the-shelf automotive radar sensor from Bosch with customized firmware. It’s linked to an Intel Core i3 computer that is mounted in a crash-resistant and watertight aluminum box and runs Linux with custom software. One of the biggest challenges was finding a camera that could relay information quickly enough to keep up with the pace on the track. Hammond turned to the robotics world, snagging a machine-vision camera with an ultra-low latency of around 50 milliseconds.
The video above provides a glimpse of what the driver sees when he glances at the display. Green chevrons indicate a vehicle that’s falling behind, while yellow chevrons transition to orange and eventually red as an approaching vehicle closes in and angles for a pass. More impressive, the system can detect if the approaching car is a P1 prototype class racer (which are faster than Corvette’s production-based GT class). By running a range of telemetry data through an algorithm that determines the approaching vehicle’s acceleration, cornering speed and other criteria, the system can tell which class of vehicle is approaching and place a strikethrough in the chevron to tell the driver it’s best to back off.
When a pass does occur, a flashing yellow arrow lets the driver know whether to close the hole or let a quicker car get by. More beneficial for the racer is the system’s ability to identify vehicles at night or in rain-soaked conditions, which makes racing safer and could eliminate many of the yellow flags that postpone a race.
For now, the Pratt & Miller Corvette is the only car equipped with the tech, but the company intends to make the system available to its competitors as early as this year.
“We’ve just got to figure out a price,” says Hammond.