Robotaxis Are Coming. So Why Are We Still So Unprepared?
Source: Wired, Ashley Nunes
Photo: Thomas Richter, Getty Images
IN A FEW weeks, Waymo will launch America’s first commercial robotaxi service. GM and Ford are both developing competing fleets. The grand vision? Fewer road fatalities, less congestion, and cleaner air, as human drivers cede control of their cars to automation. Vehicle ownership—pricey and inefficient—will ultimately drop, and Americans will opt instead for cheaper, autonomous taxis. A more productive continent will rise, the thinking goes, one where citizens are freed from the burdens of driving and can engage in other, more rewarding pursuits.
At least that’s the idea. Standing in the way, however, is government red tape—specifically, over 70 safety regulations (many outdated) that auto manufacturers must follow. But last week, Trump administration officials announced plans to “reconsider the necessity and appropriateness,” of some of these standards, the most notable of which requires that drivers always be human. As self-driving technology hits the road, officials want to do away with this requirement altogether. Their timing couldn’t be better.
Trump’s Department of Transportation argues that in the era of automation, these rules are irrelevant. Their solution, then, is to exempt manufacturers from following them. Yet while that’s the right sentiment to further the advancement of technology, it’s also one that overlooks the greater challenges posed by self-driving cars.
Understanding these challenges means accepting the fact that driverless does not mean humanless. Automation can alleviate some of the burdens of driving, but will never eliminate them entirely. Despite their best efforts, automakers can’t guarantee self-driving algorithms will work as intended all the time, every time. Nor can these algorithms be trusted to tackle road hazards (seagulls, snowflakes, and fake graffiti come to mind) as ably as a human driver.
So automakers have come up with a solution: remote control. The idea is simple. When a robotaxi malfunctions, a human steps in to help. Rather than sitting inside the car, however, these teleoperators will be stationed in a command center located miles away. Think of it as air traffic control for cars. Still, that system raises a vexing problem—how many robotaxis should one teleoperator watch?
Auto execs will say as many as possible. After all, paying one person to watch one robotaxi makes little sense. With that business model, the likes of Waymo are better off abandoning the concept altogether. But if teleoperator costs can be distributed over many robotaxis, the idea suddenly becomes more feasible. It also becomes far riskier.
The purpose of a teleoperator is to help riders in an emergency. That’s hard enough with one robotaxi at hand. But what happens when you have 10, or 20, or 50 vehicles under your purview? Can one person really be expected to help a distressed vehicle while keeping an eye on several others? What happens multiple robotaxis need help? Can safety be assured for each of those riders?
Another unaddressed concern is worker classification. A patchwork of collective bargaining agreements and state laws place a cap on the permissible number of working hours for employees. These caps are more often stringent for transportation workers, with good reason. Professions involving driving require focus and concentration, both of which fade over time. When that happens, the results can be catastrophic. Giving drivers more time off keeps them sharp at the wheel and keeps the public safe.
The rise of the robotaxis raises an important question: Should teleoperators be considered transportation workers? Doing so would grant them the same downtime as taxi drivers and, some say, ensure safer roads. But it would also cut into companies’ margins, as more staff would be required to keep their ventures going. That’s something auto execs are unlikely to be enthused about.
On these issues, the Trump administration has been silent. It avoids addressing the human component to automated cars at our peril. Self-driving technology can deliver considerable benefits to society, but realizing those benefits will require that safety and profitability go hand-in-hand. Our politicians need to craft laws that do just that.