Chris Urmson, the former Google executive who once led the company’s self-driving car project, said “one of the great things about American innovation” was that if the law “doesn’t say you can’t do it, then you can.” (source) Google is not alone in this perspective and, in fact, many of the technology companies have been lobbying the government to minimize regulations so as not to slow technological advancement. It’s no coincidence that Uber is testing its driverless vehicles in Pittsburgh and Google is testing a fleet of driverless vehicles in Austin. Pennsylvania and Texas are two states that have permissive regulations and have purposely minimized any regulatory barriers to testing driverless vehicles.
As stated in a Washington Post article, “Pittsburgh might be the exact environment that innovators love to leap into — a legal void that can be defined by technologists, not bureaucrats. The question is how fast, and under what conditions, should the testing of a life-changing technology occur.”
While I am extremely supportive of technological advancement, I also believe in government regulation advancement. Technology is being developed quickly and, not surprisingly, government regulations are not keeping up. That doesn’t mean government regulation isn’t needed….it just means it needs to be figured out! While the bureaucrats may not know how to regulate driverless technology yet, I think it’s important that they establish some regulations that keep them well-educated on the advancing technologies and well-connected to the technology companies. I believe this is the only way that we can ensure the appropriate government regulations are in place when the technology is actually public-ready.
Michigan is working to maintain itself as an autonomous vehicle-friendly state by passing legislation that supports both the technology development and deployment. The chamber unanimously approved a package of four bills (S.B. 995, S.B. 996, S.B. 997, and S.B. 998), which will allow for the testing of driverless vehicles with no steering wheels or pedals, establishing an AV research center, and creating a liability shield for AV mechanics, amongst other rules. As stated in this article, “if these bills pass the state House of Representatives without substantial amendment, it’s not unthinkable that this new industry’s center of gravity may continue to shift away from Silicon Valley and back toward the original Motor City.”
Differing state regulations, of course, do not address the issue of having a patchwork of requirements state-by-state. I know I’m not alone in being anxious to see the federal guidelines for driverless vehicles, which are expected to come out any day!
Good insight, Lauren. I suppose the challenge for lawmakers with this or any other new technology is the balance of setting objectives (must operate safely) versus being prescriptive (e.g. must have a steering wheel). It sounds like the Michigan legislature might be ahead of California in this regard; at least we have the weather.
I think this is a combination of technologists getting ahead of themselves as much as regulators lagging behind. There are a host of unanswered questions regarding the technologies’ performance under real world road conditions over the 10-20 year life of a vehicle. All the sensor technologies leave significant gaps in performance and no automated system has yet been run through adequate testing. The only way to reach full confidence in the real-world impact of automated driving on road safety will be through deployment of millions of vehicles. Real traffic is too complex and diverse to model in a lab or capture through one or two hundred vehicles traveling 100 million miles (Americans alone drive 250 million vehicles three trillion miles each year).
Like you, I would like to see rapid deployment, but also prudent regulation. These technologies will save lives because they go to the root cause of most accidents. But they also will open the door to new accident scenarios if we do not carefully think through possible consequences and eventual failure modes. The best way to approach this would be gradual introduction which seems to be what we are actually doing. Equip vehicles with automatic braking and limited corrective steering to bring down accident rates while conducting more ambitious limited experiments with highly automated vehicles to build our experience with their capabilities, limitations, and impact on traffic. This will allow us to reduce the risks and develop appropriate regulations to govern their design and performance.