Robotaxis in San Francisco. Justin Sullivan/Getty Images

 

By David Zipper

 

Autonomous vehicle companies claim that “humans are terrible drivers” and their tech is needed to save lives. Don’t buy it.

 

Traffic deaths have been tumbling across the rich world, with Japan and Norway among the countries recently reaching postwar lows. The notable outlier is the United States.

 

American crash fatalities hit a 16-year high in 2021 before barely budging last year. An American is now two to five times more likely to die in a collision than citizens of peer nations. Those expressing concern about this trend include Transportation Secretary Pete Buttigieg (who has called it “a national crisis”), roadway safety advocates, and newspaper editorial pages. Another worried party is Cruise, the self-driving car company that is majority-owned by General Motors.

 

“Humans are terrible drivers,” reads a full-page advertisement that Cruise ran in multiple newspapers last week, including the New York Times. “People cause millions of accidents every year in the US. Cruise driverless vehicles are designed to save lives.” Other autonomous-vehicle companies and investors have likewise pitched their technology as a solution to America’s road-safety struggles. Two days before Cruise ran its newspaper ads, the Google-owned Waymo published a blog post about the dangers of speeding, concluding that “by respecting the speed limit, the Waymo Driver will contribute to safer streets in places where we operate.” The website of Safer Roads for All, an AV-funded lobbying group, bluntly declares that “autonomous vehicles will take human error off the road, reduce crashes, and save lives.”

 

The AV industry is only partially correct. The soaring number of U.S. roadway deaths is absolutely a national emergency that demands action, but its companies and supporters are utterly wrong when they suggest that self-driving cars offer a clear solution, let alone the best one. Claims that autonomous vehicles present a deus ex machina that would liberate Americans from deadly collisions—claims that the industry has made for years—should be treated as the self-serving deceptions that they are.

 

Boosters of self-driving technology have been citing safety as their motivating goal for well over a decade: In 2010, Google, which had recently launched the AV venture that would turn into Waymo, published a blog post claiming that “our goal is to help people prevent traffic accidents.” Once AV tech is sufficiently reliable and ubiquitous, the story goes, car crashes can be relegated to history, alongside pestilences like smallpox and polio.

 

Are AVs safer than those driven by humans? It’s a question without a clear answer. Cruise’s recent advertisement claimed, accurately, that its robotaxis “never drive distracted, drowsy or drunk,” and that they will not commit certain mistakes that human drivers do. But that’s only part of AVs’ safety story.

 

Computer-powered vehicles can—and do—make other mistakes in situations that a person sitting behind a wheel could easily handle. Residents of San Francisco, the national epicenter of robotaxi deployment, have seen this firsthand. The city has documented myriad traffic and safety problems caused by confused or malfunctioning robotaxis, including an incident in March in which a Cruise vehicle got tangled in streetcars’ overhead wires, dragging them a full block before smashing through caution tape blocking a street.

 

Self-driving technology will certainly improve as companies learn and adapt. Eventually, maybe, a self-driving vehicle will be objectively safer than one driven by a human. But that’s just one car. A hundred or even a million autonomous vehicles would have a small impact in a country with 278 million personal vehicles that are an average of 12 years old. Even under the most optimistic of scenarios (which also assume a dramatic decline in the cost of AV software and hardware), the U.S. is decades away from self-driving tech being sufficiently advanced and widespread to fulfill boosters’ visions of a nation without crashes. Hundreds of thousands of Americans will die while we wait.

 

But why wait? There are plenty of surefire strategies to reduce crash deaths that are ready for adoption right now. The most obvious one: Redesigning roads and reducing speed limits to slow down motor vehicles, which provides more time to avoid a collision and lessens the impact of those that still occur. Cities and states could also install more automatic enforcement cameras, which have been shown to reduce dangerous traffic violations. (Such cameras, often loathed by drivers, are currently illegal in states like Massachusetts and Texas.) Officials might also follow the lead of New York City, which is using its own fleet vehicles to pilot “intelligent speed assistance,” technology that adjusts a vehicle’s top speed to reflect the roadway’s speed limit. (Just a handful of cars with speed limiters can have an outsized safety impact.) Or they could create vehicle weight fees that disincentivize big SUVs and trucks that endanger pedestrians.

 

Better yet, safety-motivated public officials could accomplish a world of good by investing in public transportation. As I’ve argued previously in Slate, riding the bus or train is orders of magnitude safer than driving a car; improved transit service would encourage more car owners to leave their keys at home. Notably, AV boosters have cited its future potential as a reason to avoid funding transit expansions.

 

If you’re still unconvinced about AVs’ wholly optional role in roadway safety, just look across the Atlantic. With nary a self-driving car in sight, the European Union has seen a steady decline in roadway deaths. Last fall I visited Helsinki, a city with so few annual traffic deaths that one can count them on a single hand (for comparison, Portland, Oregon, a city with an equal number of residents, had 63 such fatalities last year). I asked a Helsinki official what role technology had played in his city’s safety success: “Zero,” he told me. “We simply slowed down the cars.”

 

Why, then, are Americans being told that self-driving cars are essential to fixing the crisis on our roadways, when they so clearly aren’t? For one thing, robotaxi operators like Cruise are hemorrhaging money at a time when tech companies are tightening their belts. While these companies could expand deployments at a slower pace over the last decade, setting up pilots in cities like San Francisco and Phoenix, they now face enormous pressure to scale their service in order to keep investors at bay. (The recent shuttering of Argo.ai, which had raised billions from investors, casts a long shadow.)

 

Self-driving companies are also heavily regulated, which makes public and government perceptions critical to their future plans in places like California, where state officials have twice this summer postponed a major decision about robotaxi expansions in San Francisco. It’s a lot easier to win support for an argument of “Please let us expand so we can save lives” than “Please let us expand so we can keep our investors from shutting us down.”

 

It is a cynical but savvy move for AV companies to focus on safety as their core lobbying pitch, instead of on other options like convenience or access for those unable to drive. Unlike more realistic road-safety strategies like slowing down urban traffic, self-driving technology does not threaten the primacy of the automobile in American life, which many public officials are wary of challenging. In fact, overhyping the safety benefits of self-driving cars allows the auto industry to concurrently fulfill two key objectives: It positions car companies as a solution to a American safety crisis they themselves helped create, and it serves as a distraction from proven tactics (like road diets or transit expansions) that make their cars and tech less useful in urban areas. And they have little to lose by exaggerating AV benefits; past promises of car-dominated utopias have repeatedly come to naught without inspiring a regulatory smackdown or popular backlash.

 

So by all means, do take seriously the chorus of voices declaring that the American approach to road safety has failed, and that the country urgently needs a course correction to save lives. But claims that self-driving cars are necessary to reduce the carnage should be seen as the hollow sales pitch that they are.

 

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.

 

https://slate.com/technology/2023/07/cruise-autonomous-vehicles-safety-waymo-self-driving-cars-ad-new-york-times.html