Regulating Self-Driving Cars For Safety Even Before They're Built
The U.S. government wants to help you take your hands off the wheel.
The Department of Transportation on Tuesday issued its Federal Automated Vehicle Policy, which outlines how manufacturers and developers can ensure safe design of driverless vehicles, tells states what responsibilities they will have and points out potential new tools for ensuring safety.
Regulators say they want to prepare for the transition to self-driving vehicles, which they say will save money, time and lives. Transportation Secretary Anthony Foxx tells NPR's Robert Siegel he hopes these guidelines will ensure that safety is a priority as the technology continues to be developed.
Asked to predict how soon self-driving vehicles will represent, say, 25 percent of the cars on the road, Foxx says, "I don't know about percentages, but it's very clear that there is a growing interest in the marketplace to bring these vehicles into the lives of Americans. And it's incumbent upon us to get ahead of it and to make sure that safety is part of the thought process at the very beginning, and that's part of what our policy will set forth."
On why self-driving cars need federal regulation in the early stages
Well, I would say that there's not really a conflict between innovation and safety. That you can actually have innovation, you can have safety, and you can innovate in the safety arena if you take the right approach.
We did not have the National Highway Traffic Safety Administration back when the Model T was put on the assembly line, and if we had, we probably would have saved untold numbers of lives by having that kind of vigilance at the beginning. We have that opportunity today. This is a once-in-a-hundred-year moment to capture a technology while it's in its early stages and build a culture of safety within it, and that's what we intend to do.
On whether the guidelines require manufacturers to share research
We've had good experience with data sharing among highly competitive cohort of industry. I would say the Federal Aviation Administration is our best example of this, where data is shared anonymously, but it has helped us in many ways predict safety challenges and avert safety challenges within the industry.
We think this model could be used in the auto industry, particularly with a driverless environment where there is going to be so much more data available than we currently have today. We certainly want to encourage collaboration within the industry.
On the issue of liability with self-driving cars
I think that's a question that's going to need further conversation, and in the guidance we've laid out we expect the states to be engaged in that discussion as well. ...
The policy recognizes there are areas that we have deep knowledge today and can develop policy around, and there are areas that need to be discussed over a longer time frame and that's one of them.
On whether U.S. infrastructure is ready for self-driving vehicles
Well, I have questions about whether our streets are in a condition for human drivers today. That's why I went and argued so strenuously for a long-term surface bill. Obviously, our infrastructure needs to be kept at a good state of repair.
I also believe that over the next decade or so we're going to start integrating more technological capability into the infrastructure itself — much more sophisticated street signal alignment. The street lights networks that we have today actually communicating with cars, and turning off when there are no cars on the road, and turning on when there are. I think you're going to see a lot of that technology take root, and you're going to see it at the municipal level, at the state level.
Copyright 2020 NPR. To see more, visit https://www.npr.org.