"We may be on the cusp of a safety innovation revolution," said Mark Rosekind, who runs the National Highway Traffic Safety Administration (NHTSA), kicking off the federal government's first public hearing on self-driving cars. Over 200 people attended the hearing Friday at the US Department of Transportation's headquarters in Washington, DC. For almost seven hours, agency officials heard testimony from automakers, engineers, consumer watchdogs, and disability advocates on the hopes and fears surrounding autonomous vehicles. They're comments ran the gamut from "this is the best thing ever" to "ban self-driving cars before they kill us all."
"this is the best thing ever" vs "they'll kill us all"
And now it's the federal government's job to distill all these feelings into a set of policy proposals that address the serious technical, safety, and ethical issues surrounding self-driving cars, without killing innovation. Last January, DOT Secretary Anthony Foxx said his agency would take six months to issue guidance and model state policy on autonomous vehicles in order to avoid a "patchwork" system that could create conflicts. California is already fighting with Google over whether self-driving cars should be required to have a human being behind the wheel at all times. In February, NHTSA said the computer in Google's self-driving car could be legally considered "the driver."
"That is why we are taking a deliberative approach."
"I believe that too often we talk about a tension, or striking a balance, between safety and innovation, as if there is a trade-off between the two," Rosekind said in his opening remarks. "It is true that there are real and significant questions about the safety of new technology. At the same time that new technology is bringing safety advancements, it can also create new vulnerabilities and risks. That is why we are taking a deliberative approach, making sure we get the safety advancements we know work to market, while also making sure that new automated technology is safe."
In some ways, it was an illustration of how slowly the gears of government tend to grind, especially in the face of fast-paced technological innovations. NHTSA thinks it's acting fast by promising to release a comprehensive take in a speedy six months. However Rosekind clarified that NHTSA wasn't going to release "thousands of pages" of new regulations on this issue, but viewed this process as a starting-off point for a broader discussion about the possibilities and dangers posed by self-driving cars. The release of the agency's guidance later this year will trigger another round of public comments and rule-making. And so on and so forth.
Rosekind said data security and privacy will feature prominently in his agency's forthcoming proposals, citing the Wired article from July 2015 about white-hat hackers who remotely took over a connected Jeep's controls while it was speeding down the highway. "We don't know the potential technology stuff that could be coming," he said, noting that today everyone is talking about fully autonomous cars, but that tomorrow "it could be something completely different."
"We don't know the potential technology stuff that could be coming."
Representatives from Ford, Honda, and General Motors testified at the hearing, but other major players were notably absent. No one from Google, Tesla, Uber, Lyft, Apple, Volvo, or any other major companies that are working on autonomous technology, were in attendance. They will have another chance to have their voices heard in a few weeks, when NHTSA convenes a second hearing at Stanford University on April 27th.
DOT is working with many of these players in developing its guidance, including GM, Ford, Tesla, Volvo, Fiat Chrysler, Delphi, and Google — a fact which caused some consternation during the hearing. Joan Claybrook, who ran NHTSA during the Carter administration, criticized the agency for allowing automakers to provide too much input its regulatory process.
"NHTSA should put more of its energy into serious evaluation of driverless vehicles and less into efforts to avoid the effective regulatory process that save hundreds of thousands of lives," Claybrook said. "The safety of the American public will not be best protected with a kumbaya between the federal agency charged with issuing safety rules and the industry constantly seeking to avoid safety regulation."
One speaker said he was concerned that self-driving cars could be use to conduct terrorist attacks. James Niles, from Orbit City Labs, said that "autonomous vehicles could be used as weapons — like drones on wheels." He cited a 2014 report from the Federal Bureau of Investigation, that warned about future where robot cars served as "lethal weapons." Niles recommended that carmakers install something like a "sniffer" that could detect the presence of hazardous materials and automatically disable the car.
Consumer advocates accused the government of "rushing" to develop rules on self-driving cars before the technology could even be understood. John Simpson, privacy project director for Consumer Watchdog, was sharply critical of Google's autonomous vehicle program, raising questions about the ethics behind this technology. "While [Google] wants to take the human driver out of the equation, its programmers will be making life and death decisions in the algorithms they create to navigate the vehicle," he said. "What ethical choices will these humans program into the robot car? Will the vehicle protect the safety of its occupants over the safety of pedestrians and cyclists?"
"What ethical choices will these humans program into the robot car?"
Many of the speakers called on the government to issue strict regulations, not voluntary standards or guidance, with regard to self-driving cars. "You cannot rely on the industry to do the right thing," said Clarence Ditlow of the Center for Auto Safety. "Otherwise you're going have the American public serving as guinea pigs in driverless vehicles." Meanwhile, automakers pled for flexibility and leniency. Paul Scullion from the Association of Global Automakers said the new rules should not be "the be-all-end-all" of autonomous driving.
Some of the most poignant comments were made by advocates for people with disabilities, who see self-driving cars as way to gain independence in their lives. Anthony Stephens from the American Council of the Blind, spoke of reading Robert Heinlein's science-fiction classic "Stranger in a Strange Land" as a blind child, and being fascinated by the concept of "a car that would come and pick people up at their apartment and drive them to their destination... That to me was science fiction."
"that to me was science fiction."
As the hearing progressed, it was clear that proponents of fully autonomous vehicles were outnumbered by the skeptics. Of course, companies like Google and Tesla are multi-billion dollar entities that can afford armies of lobbyists to influence the regulatory process, so it's unlikely their voices will be unheard. Case in point, the Senate recently held a hearing on self-driving cars, during which each committee member spoke glowingly about Tesla's Autopilot feature. They liked it because Tesla gave them all rides in its shiny new cars. Questions about data security and ethics tend to melt away when you're behind the wheel of a Model X that's driving itself.