Close is not good enough—except when it pretty much is. That's the case for Stanford engineers, when the racetrack performance of their robotically guided car is matched against the results of a race car driver in the same vehicle.
J. Christian Gerdes, professor of mechanical engineering and director of the Revs Program, reports that computer-charted comparisons from late 2014 show differences of just tenths of a second between autonomous laps by Stanford's Audi TTS Coupe (named Shelley) and the times recorded with racetrack CEO David Vodden at the wheel. Depending on which point in a lap is picked as the start and finish, the driverless car either wins by four-tenths of a second or loses by five-tenths of a second.
And that, says Gerdes, represents a rapid advance from previous tests at Vodden's Thunderhill Raceway Park in Willows, Calif. "We were a few full seconds behind David the last time we did a comparison," he notes. The ultimate goal of these tests is to understand automated control in ways that further driving safety.
