Driverless cars will make us much safer ... if we let them
We shouldn’t overreact to fatalities or overly regulate the rapidly evolving technology of autonomous vehicles, argues Pittsburgh attorney PAT SOREK
Car accidents happen thousands of times a day and seldom make the news. This one was reported around the world. On March 18, Elaine Herzberg was struck by an Uber vehicle in Tempe, Ariz., and tragically lost her life. But the involvement of an Uber autonomous vehicle (AV) turned the event into frontpage news.
On the same day of the Uber collision, a 47-year-old Pittsburgh woman also was killed by a car. This got a four-sentence mention in the Post-Gazette.
In a rational world, and in a nation that professes to place high value on human life, the Pittsburgh woman’s death would be considered just as terrible as Ms. Herzberg’s. The fact that she got a blurb illustrates the double standard that has grown up around the AV industry. The problem this poses is that, when early fatalities are given such high-profile coverage, AV experiments get shut down or suspended, which threatens to delay the development of AV technologies that ultimately will save thousands of lives a year.
Pittsburgh is one of the hubs for AV companies. Argo, Aurora, Aptiv and Uber have operations here. The AV technology they are pioneering now faces a test that is perhaps as daunting as laser-mapping roads — that is, navigating the quirks of human judgment.
For all their promise, AVs will not deliver perfect safety, especially in their early stages. They need to be considered within the context of driving as it is now — with humans at the wheel — so we can make them as flawless as possible as quickly as is feasible. We should want them to start saving lives and preventing injuries and improving this everyday activity of modern life ASAP.
Humans are dangerous
Human error while driving happens in a fraction of a second, with often grievous results. Nothing else we do almost daily involves this kind of threat.
In 2015, there were 6.3 million accidents in the United States, causing nearly 33,000 deaths and 2.3 million injuries. As The Economist recently reported, “even with modern safety features, more Americans have died on the road since 2000 than were killed in all the wars of the 20th century.” Federal studies have shown that human error causes 94 percent of accidents.
Aside from death and injury, more than $500 billion is wasted each year when loss of life, decreased quality of life and pain are included in calculations.
Congestion is worse than ever, too. Drivers in 68 urban areas lost $72 billion in wasted time and fuel, about $755 per driver, in 2014. Each rush-hour commuter lost 42 hours to congestion. AV vehicles will be able to talk with one another to avoid traffic jams.
Traffic’s emotional toll cannot be overlooked. It can produce emotions strong enough to cause road rage, which also can result in death and injury. AVs won’t cut off each other and can’t get mad.
To get perspective on what an outlier traffic carnage is when it comes to public acceptance, let’s compare it with other health risks.
Start with lettuce. When lettuce carries germs that make even a dozen people sick, public institutions leap to prevent further harm to the public.
Smoking? You’re free to smoke, at least outside, because
lighting up creates a tiny chance of harming others, yet your smoking is rigidly regulated and controlled by social norms.
Opioid abuse: Deaths from overdoses eclipsed those from car accidents in 2016, climbing above 42,000. This is an epidemic of people mostly hurting themselves, but massive federal, state and local initiatives have been launched to cut off the supply of drugs, provide life-saving antidotes and offer treatment for addiction.
If we mobilize so actively to prevent death or injury from lettuce, smoking and drugs, why are we so tolerant of the possibility of being killed or maimed on the road? Why does life suddenly get so cheap when you step into a car?
We should expect better.
The AV revolution
Yes, driving is, in theory, highly regulated and bound by rules designed to keep us safe, such as the 1,012 provisions in the Pennsylvania Motor Vehicle Code. Yet when the state sends registration forms to every vehicle owner each year, the single traffic law required to be mentioned isn’t about, for example, distracted driving. It’s about littering.
We could improve traffic safety now by beefing up what we know will work, such as education and enforcement. The campaign to promote seat-belt use has been remarkably effective. Tougher enforcement, starting with stiff warnings and costly fines, also would improve safety.
But the system is too fragmented to be fully effective. Traffic laws are written by 50 different states and then enforced at the state, county and local levels. Add in the fact that police answer to elected officials, who in turn answer to citizens, who hate getting tickets, and the reasons for the sad state of traffic safety in the United States become clearer.
So, who will we have to depend on to dramatically improve the traffic system with focused intelligence, energy and investment? Mostly, it’s the AV companies.
This does not mean they are or should be blindly trusted, by any stretch. AV companies have plenty of safety rules to follow already.
But we also should remember that there are other effective, non-regulatory ways to protect life and limb. Front and center is our time-tested legal system. If a product hurts someone and the maker is found at fault, the injured person is compensated and punitive damages can be levied in cases of gross negligence. We the people, assembled as jurors, decide who is to blame.
If regulators never wrote another word about AV activity — and they shouldn’t write too many, given how difficult it is to keep up with rapidly evolving technology and practices — AV companies still would be held accountable by the legal system.
As the U.S. Supreme Court once put it, “The obligation to pay compensation can be, indeed is designed to be, a potent method of governing conduct and controlling policy.” This method is invoked more than 200,000 times a year for vehicle accidents.
The legal system is pretty good at dealing with technical complexity, too. In recent years, Toyota faced some 500 claims that its cars were prone to unintended acceleration, resulting in payouts of billions of dollars. Volkswagen rigged its engine software in about 11 million cars to defeat emissions tests. The consequences were severe: The company was hit with a criminal fine of $2.8 billion and had to spend additional billions on lawsuits, recalls and retrofitting.
The market also can punish companies that screw up. AV companies, in particular, know the spotlight is on them. Bad publicity can shrink their market share and tank their stock prices.
By contrast, a blanket program of governmental regulation under current conditions is unlikely to solve anything.
AVs pose novel technical issues about which regulators know little. Many have candidly voiced their uncertainty over what standards to set. Moreover, we’re already swimming in traffic regulations, which have not stopped tens of thousands of deaths and injuries each year. Even precise, science-based regulations can be defeated or evaded, as in the case of Volkswagen.
Federal, state and local officials and regulators should not be mere bystanders. They should closely monitor the progress of AV technologies, investigate and set standards where it makes sense, and press for common platforms. But the expertise is being developed in the private sector. And the private sector has everything to lose and nothing to gain if its products do not prove to the public that they work better than what’s already out there.
It’s important to remember that we Americans have tolerated egregious levels of death, injury and waste as our driving culture has evolved. AVs do not offer the prospect of perfect safety on the road, but it won’t be long before they get pretty close — that is, if we don’t overreact to early tragedies and stunt their growth.