Gearing Up for the Rise of Driverless Cars

The person in the driver's seat of this Google autonomous car is only there as a backup. Computers are controlling the car.
The person in the driver’s seat of this Google autonomous car is only there as a backup.

Would you be willing to give up the steering wheel of a car … and put your life in the hands of a computerized chauffeur?

While many people are likely to react with a resounding “No way!” to such a question now, the proposition has lots of reasons to recommend it. With big names from Google to recently, Newt Gingrich, hawking the idea of driverless (also called “autonomous” or “self-driving”) cars, the emerging technology offers the potential to be not just safer but greener. In this article we put forward a few of the reasons why.

Networked Artificial Intelligence

Certain higher-end cars today already have limited versions of computer-assisted driving, such as radar- or laser-enabled collision avoidance and lane departure warning.

That’s good enough when it comes to saving your own skin (and sheet metal) on the open road. But modern engineering can multiply that benefit to every other driver, too. Sufficiently equipped driverless cars could use wireless networking to coordinate their driving as a swarm. Maintaining a constant speed and distance from one another could have multiple benefits:

  • less congestion, since no vehicle would have to guess the intentions of another
  • more efficiency – the act of vehicles “drafting” in one another’s wake, minus the wastefulness of stop-and-go driving could yield up to a 30-percent savings in fuel or battery power
  • far fewer accidents; a widely cited KPMG and Center for Automotive Research report notes that human error is responsible for 93 percent of automobile accidents, and crashes cost U.S. citizens $299.5 billion a year

Computers Have Superior Reaction Time

Let’s face it, qualifying for a driver’s license isn’t exactly advanced physics. Although perhaps the actual driving is – which is why safety-enhancing features such as anti-lock braking and traction control systems are already handled by computers. Another noteworthy automated feature of modern cars: expanding airbags in an accident. A human could never react within the milliseconds of an unfolding accident to effectively deploy an airbag.

Radar, infrared, and 360-degree sensors would give a self-driving car a far more complete picture of its surroundings than someone relying on the limited field of vision afforded by human sight. And in case it needs repeating, machine reaction time is faster, much faster than humans’.

Fewer Accidents (and Deaths)

The terrorist attacks of 9/11 immediately killed nearly 3,000 people and sparked a radical overhaul of commercial aviation security. But in the United States alone, more than 10 times that number of people die every year in automobile accidents (though fatalities have been slowly trending downward). As previously noted, human operators were responsible in most of those accidents.

Clearly, there’s nowhere near enough data yet to say driverless cars will definitely eliminate most accidents. Take, for instance Google’s claim that its driverless cars have logged nearly half a million accident-free miles while under computerized control.

Bryant Walker Smith, a fellow at the Center for Internet and Society at Stanford Law School, has studied the legal implications of driverless cars in depth. He recently told Politifact Florida that driverless vehicles would have to:

“[D]rive themselves more than 725,000 representative miles without incident for us to say with 99 percent confidence that they crash less frequently than conventional cars. If we look only at fatal crashes, this minimum skyrockets to 300 million miles.”

Those numbers may seem like a high hurdle to clear. Plus the tech has other issues to work out, such as lowering the cost and learning to navigate snow and new, not-yet-catalogued roads. But it’s worth bearing in mind that the computational power and enabling technologies behind artificial intelligence have been increasing exponentially, rather than in the linear fashion most people are used to.

Perhaps the biggest challenge of all to autonomous cars will be the human factor – getting the public to be comfortable with them and convincing the labyrinth of regulators and policy makers that the vehicles equate to a change that will benefit society.

What do you think – are driverless cars the next big thing or a just a road to nowhere? Comment and let us know what you think.