Experts: Accident Would Shut Down Google's Driverless Car Experiment

People "won't tolerate machines killing humans," experts say.


California Gov. Jerry Brown, front left, rides in a Google driverless car to sign the bill that opened the way for driverless cars in the state.

By + More

In the future, you'll be able to text while you drive all you want, watch TV, maybe even read a book. But the dawn of that future, experts say, is all riding on Google.

[PHOTOS: Make Way For Google's Driverless Car]

The search giant thinks the future of the car is driverless, and it has the testing, and backing from three states to prove it. Nevada, Florida and California have legalized testing of the car on its public roads, and the car has intrigued lawmakers in Washington, D.C. and Texas.

Through several hundred thousand hours of testing, the car has yet to crash while in automated mode (it had one accident while a human driver was operating it, the company says).

The company is reportedly talking to Tesla Motors about implementing its driverless technology into Elon Musk's electric cars.

But when a driverless car inevitably crashes, and it will—all the good will the company has built up could be lost, experts warned Tuesday at an event discussing the future of unmanned vehicles in the United States.

"There's been all this public outcry about [drones], but the Google car has slipped in under the radar," Missy Cummings, a former fighter pilot turned MIT drone researcher said Tuesday. "You're more likely to be run over by the Google car far before a [drone] would ever fall on your head."

[READ: Washington, D.C. Could Soon Pass Driverless Cars Bill]

Whether that's true or not, and considering that the Federal Aviation Administration is still concerned about a drone's ability to "sense and avoid" other aircraft, there's little doubt a driverless car will hit and injure someone at some point.

Even though Google's car can detect obstacles and brake faster than a human can, it can't defy the laws of physics. Early tests suggest that, at 40 miles per hour, an automated car can stop within 9 feet; the average human (who is paying attention), will stop within 12 feet.

"If a child steps out at 10 feet, the human kills the child, the automated car doesn't," Michael Toscano, head of the Association for Unmanned Vehicles International said Tuesday. "At 8 feet, either one will kill the child. We accept humans to be faulty, but we don't accept machines killing human beings."

Last week, an art collective in New York City caused a bit of a stir when it staged accidents in a fake Google driverless car, hitting a cyclist.

When a child, or a biker or a car inevitably jumps in front of an automated car and causes an accident, the public reaction would be "catastrophic," Cummings says.

[ALSO: Google, Apple Prepare to Square Off in Smartwatch War]

"If in this early stage of development, the Google car kills a child, it will shut the robotic car industry down and might have an effect on the air as well," she said. "No technology is every 100 percent perfect. The public is not comfortable with machines taking over our lives. I think if in the next few years someone is killed by a driverless car, it would be a setback for the community."

Each year, more than 32,000 Americans are killed in auto accidents. People drive drunk, distracted and tired, making them inferior to machine drivers, Peter Stone, a computer scientist at the University of Texas at Austin told U.S. News last year. He's currently working on a new traffic management system that driverless cars could use.

"Can we do better than human drivers? That's not really a high bar to clear," he said. "I believe they will be significantly safer than human drivers. They won't drive drunk, suffer from road rage or text while driving."

More News: