We talk a great deal about how cars congest our cities and pollute the atmosphere. We talk less about how they keep killing and injuring people simply trying to get from A to B on two feet.
Lately, our auto industry conversation about road safety has been dominated by visions, sold by Silicon Valley, of vehicles that minimize or even eliminate the need for input from a fallible human driver. Every year, more cars come armed with “pedestrian detection and avoidance” systems; soon, these systems will likely be standard issue. And not long after that, we are promised, sensors and self-improving algorithms will take over the driving process altogether, eliminating human error from roads and ushering in a new golden age of safety for all their users, whether or not they’re cocooned by a car’s steel frame. Since 2017, General Motors, the US’s largest car manufacturer, has claimed that it is developing self-driving vehicles in the service of a “triple-zero” world: zero crashes, zero emissions and zero congestion.
Car companies swear they are here to help – by selling us products that hardly ever hit anyone or anything. But the truth is that this promise is, at best, a distraction. In fact, much of our discourse around cars, self-driving or otherwise, is less about transforming the status quo than maintaining it, obscuring paths to progress exactly when we need them most, and leaving pedestrians right in the line of fire.
Ask a room full of road safety experts what is causing pedestrian fatalities to increase and most will admit that, well, they are not exactly sure. Every time a car hits a pedestrian, it represents the intersection of a vast number of variables. At the level of those involved, there is the question of who is distracted, reckless, drunk. Zooming out, there are factors such as the design and condition of the road, the quality (or absence) of a marked pedestrian crossing, the speed limit, the local lighting, the weight and height of the car involved. In a crash, all these variables and more converge at high speed in real-world, non-laboratory conditions that make it hard to isolate the influence of each variable.
Ask that same room of road safety experts a slightly different question – not exactly why US pedestrians fatalities have risen lately, but instead why the US has more of them than any other wealthy country – and the answers will come flooding out. In recent months, after conversations with more than a dozen such experts, I became familiar with a particular tone of voice: deep frustration at how obvious it all is, but wrapped in a package of professional cheeriness.
Here is what the frustrated safety experts will tell you: Americans are driving more than ever, more than residents of any other country. More of them than ever are living in cities and out in urban sprawl; a growing number of pedestrian fatalities occur on the fringes of cities, where high-volume, high-speed roads exist in close proximity to the places where people live, work, and shop.
Speed limits have increased across the country over the past 20 years, despite robust evidence that even slight increases in speed dramatically increase the likelihood of killing pedestrians (car passengers, too – but the increase is not as steep, thanks to improvements in the design of car frames, airbags and seatbelts). American road engineers tend to assume people will speed, and so design roads to accommodate speeding; this, in turn, facilitates more speeding, which soon enough makes higher speed limits feel reasonable. And more Americans than ever are zipping around in SUVs and pickup trucks, which, thanks to their height, weight and shape are between two and three times more likely to kill people they hit.
More fundamentally, the US is the country in the world most shaped, physically and culturally, by the presumption that the uninterrupted flow of car traffic is an obvious public good, one that deserves to trump all others in the road planning process. Many of its younger cities are designed almost entirely around planning paradigms in which pedestrians were either ignored or factored only as nuisances.
There is no greater symptom of this worldview than the recurring focus on mobile phones, especially smartphones and their tendency to monopolise our limited attention. Road signs warning against phone use while driving are so commonplace that they almost blend into the landscape. Parents make their kids promise they won’t use their phones while driving. Kids nod and promise they won’t. Phone-tracking studies indicate that most of them do it anyway and that their parents do, too.
In recent years, America’s fear of the distracted driver expanded to include the distracted walker. This is a replay of an old phenomenon: it was the US that invented the concept of the “jaywalker”, a “jay” being an unsophisticated person from the country who did not even know how to walk correctly. In the US, much like anywhere that cars have taken hold, drivers screaming at pedestrians (and cyclists) that they are doing it wrong is a fixture of national life. More recently, numerous states and cities, including San Francisco and New York, have launched public campaigns against inattentive walking, as has the US National Safety Council. Some jurisdictions have passed, or sought to pass, bills that would make using a smartphone while crossing the road an offence punishable by fine. On Twitter, the US National Highway Traffic Safety Administration has weighed in on the pedestrian safety crisis primarily by coaching pedestrians on how to protect themselves. “When you are walking, be predictable”, advises @NHTSAgov.
Safety experts are not terribly excited about pedestrian avoidance technology. It wasn’t that they doubted it might save some pedestrian lives. Instead, their recurring concern was that it reflects an ongoing focus on individual shortcomings – on flawed drivers and walkers – and a neglect of flaws built in to the roads they are forced to use.
“Pedestrian detection will probably help a bit,” says Dan Albert, author of Are We There Yet?, a history of American car culture. “But at the same time, it’s pretty clear that these problems can be addressed without hi-tech solutions. And so what are the car companies up to? It’s not just about some altruistic desire for safety, or they would be including these systems on all of their cars, which very few companies are doing. It’s more about creating a range of products that allows them to maximise profit.”
Of course, people can learn to understand new tools. More troubling is the fact that very little robust evidence has been available as to pedestrian avoidance systems’ real-world benefits. The organisations rating these systems do so based on tests conducted in laboratories and on test tracks, but it has not yet been reliably established how well these tests predict performance on actual roads, with real live pedestrians instead of crash test dummies (not to mention variable light and rain conditions).
The self-driving car “space” is flooded in loose cash. This is why pedestrian avoidance systems are becoming ubiquitous. Investment dollars have both improved the required parts and programmes and pushed down their cost. Otherwise, they would not be affordable enough to sell to any but the richest car-buyers. Meanwhile, the race continues. The tech companies keep saying they are almost there (they know they said they were almost there before, but this time is different, they promise). Almost every major car company has partnered with a lab developing self-driving car technology, either because they think it might become a requirement to stay in business, or because enough of their shareholders think so that they need to make a show of playing along.
One night in March 2018, a 49-year-old woman named Elaine Herzberg was pushing a bicycle laden with grocery bags across a four-lane road in Tempe, Arizona when she was struck and killed by a Volvo X90 operating under the control of Uber’s self-driving software. As is standard practice, a backup safety driver, employed by Uber, was sitting in the car, the idea being that she would take over in the case of error. 1.3 seconds before impact, Uber’s software calculated that emergency braking was called for. According to Uber, self-braking had been turned off to reduce jerky and unpredictable behaviour. With less than a second to go, the safety driver tried––and failed––to swerve away. (The post-crash investigation indicated that she had probably been watching the singing contest The Voice on her phone.) News outlets around the world covered Herzberg’s death in obsessive detail, asking whether it indicated that autonomous vehicles were less “ready” than we had been led to believe.
But, just as the original quest for autonomous vehicles had nothing to do with pedestrian safety, concern over Herzberg’s death often felt curiously divorced from concern for pedestrians in general. Herzberg was killed in March 2018. Between January and June of that year, 124 pedestrians were killed in Arizona, more than all but four other states in absolute terms, and more than all but one on a per capita basis. It goes without saying that none of these pedestrians’ names are known around the world, or that any of them generated a slurry of commentary about whether the American transportation environment is “safe enough” or “as safe as it should be”.
In the US, meanwhile, it remains the case that pedestrian advocates have failed to engineer the cultural process that transforms a scattered mass of dead and injured bodies into a widely recognized problem. They have not come close. When two Boeing 737s went down, killing 346 people, it triggered multiple government investigations. Crash reconstruction and analysis experts showed up. Corporate spokespeople apologized, began handing out checks to victims’ families and swore to do better. Journalists searched for explanations. But cars kill a 737’s worth of American pedestrians every couple of weeks. Internationally, it is more than three 737s per day. And the news cycle barely stutters.
In 2017, for the first time, each US state was required to submit road fatality reduction targets to the federal government. Most states set extremely limited goals: Wisconsin, for example, aimed to have 342 pedestrian fatalities, instead of 361. Several set a rather fatalistic goal of no reduction at all. Eighteen states went a step further, setting as their target an increase in their pedestrian death count. It is not that they want more pedestrians to die. But they know that people are likely to be driving more, they know what their roads are like and they know the laws of physics. Unlike the unending stream of hype coming from the autonomous car sector, these dour projections received no coverage outside of traffic reform circles. Unfortunately, they are more likely to contain the truth.
Read the full article here.