RSS

The Bulletin

Human driver (barely) beats self-driving car in a race

Posting in Design

In a testament to just how far self-driving cars have come, an autonomous car raced a professional race car driver and lost. But just barely.

The scene was The Atlantic's Big Science Summit in California where the Center for Automotive Research at Stanford (CARS) raced their Audi TTS turned self-driving car against a professional driver. The pro won, but only by a few seconds.

Here's a video of self-driving car testing its abilities on the race track:

And there's a reason why it performs so flawlessly on the race track. The algorithms used to drive the car were built with the help of race car drivers, according to The Atlantic:

Using advanced neuro-monitoring technology, Gerdes placed electrodes on the scalps of drivers and monitored which sections of their brain fired as they performed different driving functions. Drivers make hundreds of decisions and tiny adjustments around every turn, and the goal of the car is to process all of the same data a driver sees and make decisions that are just as good -- or better. Using GPS, inertial sensors, and finely tuned robotics to turn the wheel, the robo-car is always adapting to conditions and asking itself what it can do to go faster.

Of course, when autonomous cars become used more widely they won't be driven around race tracks. But if you give a car the abilities of race car drivers (instead of the average driver) and combine them with "conservative software" used for standard driving then you can develop a safer driving experience.

In a Race Between a Self-Driving Car and a Pro Race-Car Driver, Who Wins? [The Atlantic]

[h/t Discovery News]

— By on November 1, 2012, 11:20 PM PST

Tyler Falk

Contributing Editor

Tyler Falk is a freelance journalist based in Washington, D.C. Previously, he was with Smart Growth America and Grist. He holds a degree from Goshen College. Follow him on Twitter. Disclosure