Driverless cars, the pinnacle of automotive innovation and the potential future of safe driving, have recently proven to be anything but: according to a study from the University of Michigan’s Transportation Research Institute in Ann Arbor, auto accident rates are twice as high for driverless cars as they are for your average error-prone human driver.
With rates like that, any self-driving car owner can expect to pay a number of visits to her lawyer, if the car doesn’t crash itself along the way.
What’s causing this auto accident discrepancy? After all, anyone would think human drivers in the kind of rush-hour traffic that backs up miles outside Farmington, Utah would have a harder time navigating the highway than a cool, calculating robot.
The catch? Self-driving cars are programmed to obey all traffic laws, regardless of the situation. So whether it’s merging onto high-speed traffic on the highway or rolling into a four-way intersection in Farmington, a self-driving car makes no concessions.
Human drivers, meanwhile, bend the rules of traffic law with abandon. Most drivers are guilty of rolling through the occasional stop sign, speeding through that yellow light or driving “with the flow of traffic” on the highway—even if traffic’s running 15 over the speed limit.
In the interest of preventing an auto accident and a subsequent trip to the local personal injury lawyer, should self-driving cars bend to the will of human error?
It’s a sticky situation, to be sure. If Google programs its cars to disobey traffic laws, the next question is: how much? If self-driving cars start deliberately breaking the law, the search engine giant will be sure to face an onslaught of government and lawyer inquiries.
In the meantime, Google is working to program its cars to be more “aggressive” while still adhering to all traffic laws. Driving is a complex social practice, whether you’re driving on the interstate or around the shops of downtown Farmington.
For driverless cars, the game is still very much a human one, law breaking and all.