Powering full-steam ahead, the Obama administration proposed a $4 billion spending bill the other day that would put self-driving cars in the fast lane. Looking to accelerate “the acceptance of driverless cars on U.S. roads and curb traffic fatalities and travel delays,” the Wall Street Journal reports that this spending measure would likely require a 10 year plan and congressional approval. If passed, the plan would require federal regulators to work closely with automakers and third parties so that a laundry list of policies and rules could then be officially proposed. Once underway it would also usher in a pilot program for testing “connected vehicles,” which would share data with one another in order to avoid crashes and group vehicles together based upon who is going where.
While Obama is pushing to see this bill pass, regulators say a lack of clear guidance on their end is the primary obstacle keeping driverless cars from reaching the mainstream. According to the report, “car makers prefer to have a clear national road map for approving autonomous vehicles rather than a state-by-state patchwork of rules,” an issue that raises more red tape than one might think. Don’t get us wrong, regulators want to encourage advancements in the technologies in order to boost vehicle safety, but this is going to be one hell of an expensive uphill battle.
“Automated vehicles open up opportunities for saving time, saving lives and saving fuel,” said U.S. Transportation Secretary Anthony Foxx last Thursday at the North American International Auto Show. If all goes well, the National Highway Traffic Safety Administration plans to exempt certain autonomous vehicles from the existing rules if they can “provide significant safety benefits,” thus giving automakers a loophole that could allow up to 2,500 driverless cars on the road for up to two years.
Regarding the execution of this potential loophole, Mark Reuss, General Motors’s product chief was quite specific. “The clarity is important,” he says. “Vehicles as they get this technology are going to have to operate in a lot of different scenarios. We have to be precise when we do this.”
Meanwhile, regulators acknowledge that they are racing to keep pace with companies that are already putting driverless cars on the road for tests and other limited uses. As of now, federal automobile safety regulations were passed way before driverless cars became a reality, and even though these rules are getting updated all the time, they typically tend to focus on seat belts, brakes, air bags, and vehicle crash safety ratings.
But companies are sick of waiting, and Google, Tesla, and GM want to see fully-autonomous vehicles become a reality as soon as possible, even as safety features like automatic braking, lane-keep assist, collision mitigation, and adaptive cruise control continue to make the cars of today semi-autonomous.
Having said that, there remain a lot of issues surrounding this technological advancement, starting with whether a driver should be liable for damage, injuries, or deaths resulting from a vehicle’s decision. Let’s say you are autonomously driving along a canyon road one day, when a stranded traveler suddenly jumps onto the tarmac in front of you. It’s decision time, but it is not yours to make. Will the car’s computer plow onward, hitting the traveler at full speed, or will it save them by swerving, and instead send you careening off a cliff?
There’s no clear-cut answer. States like California aim to hold motorists responsible for any violations, even if they are not in control of the car at the time. But companies like Google, Volvo, and Mercedes-Benz have pledged to take full responsibility should their cars do the unthinkable. What’s more, Google claims its vehicles have driven 1.3 million miles, the equivalent of 90 years of human driving, and while turning around in a cul-de-sac it avoided hitting a woman in an electric wheelchair chasing a duck with a broom. This may be a far cry from barreling around a mountainside at 75 miles-per hour, but it’s still a step in the right direction, something Google is quick to point out. “There was no single line of code in our algorithm that specifically predicted that event, but our car still knew what to do.”
According to Mike Jackson, chief executive at AutoNation Inc., America’s largest dealership chain, replacing the nation’s car fleet with autonomous vehicles would take about two decades, and millions of Americans would still likely keep driving older cars regardless. “The benefit of autonomous cars is misunderstood,” Jackson says. “It’s not the luxury of having this computer drive the car for you. If the computer realizes you’re about to do something completely stupid, it will intervene.”
As part of an effort to gain control of autonomous vehicles, U.S. regulators say that within six months they plan to issue recommendations to state authorities on the most crucial guidelines for self-driving cars. As of now, California, Nevada, Florida, and Michigan are the only states with laws allowing the testing of autonomous cars, and Google argues that being limited to these states inhibits real advancements. It also believes autonomous vehicles “shouldn’t require additional regulations” and only need government certification since “they follow traffic laws better than humans.”
The government is beginning to agree. Last week, Foxx told reporters that the widespread usage of driverless cars could have saved more than 25,000 lives last year merely by eliminating human error. “We’re entering a new world and we know it,” he says. “We’re doing our level best to get ahead of it.” With over 32,000 reported traffic deaths last year alone we can totally understand why this is one hot button topic on capital hill right now.