Self-driving, autonomous vehicles — a topic of speculation, if not science fiction, just five years ago — are suddenly going mainstream. Every major car manufacturer is offering prototypes or early models of cars that do most of the driving automatically, including luxury models from Tesla with a built-in “autopilot” feature. Startups and major tech companies, notably Alphabet’s Google X division, are investing heavily in smart car technology, as are network ride-sharing companies such as Uber and Lyft.
This means self-driving cars have shifted from a period of wild experimentation directly to market adoption — what Paul Nunes and I describe in our 2013 HBR article as “big bang” disruption.
Before long, we’ll stop referring to the underlying technologies involved — including lasers, radar, cameras, embedded sensors, and advanced machine learning software — in terms of what they’re replacing. “Self-driving” or “smart” cars will simply become whatever we call the next generation of transportation technology. The change will become invisible.
But typical of disruptive transformation in other industries, the U.S. legal system is already having trouble keeping up with the pace of developments in transportation. When Google first began testing its almost comical-looking prototype vehicles on California roads in 2009, lawmakers didn’t even have a vocabulary to talk about the new technology, let alone any understanding of whether driving rules and accident liability laws dating back 100 years or more would need to be adapted or completely rewritten.
It’s not just the rules of the road that will be affected by driverless cars. Autonomous vehicles will profoundly affect insurance, road design and construction, traffic management, taxi and limousine services, the materials and safety equipment in vehicles, and asset ownership (who needs to own a car when one can simply be summoned from the most efficient location?). All of these are heavily regulated, often by federal, state, and local agencies.
While cars have been getting smarter and smarter, the removal of human operators is what will dramatically change the law. For centuries, the principal assumption behind our transportation supply chain, one baked deeply into the legal system, is that cars are dangerous — or, to be more specific, that drivers are.
Get behind the wheel, and suddenly you are calculating thousands of precise time and motion trajectories and contingency scenarios, and the effects of other drivers of widely varying skill levels making the same calculations. Errors compound, causing unnecessary traffic slowdowns at best and multiple-vehicle pileups at worst. Plus, the rise of mobile phones is being blamed for a new spike in U.S. traffic deaths.
From the beginning, self-driving vehicles have proven to be better drivers than most human operators, since software is able to learn as it goes. Networked smart vehicles, for example, can safely travel much closer together — a technique known as platooning.
The legal system is playing an urgent game of catch-up, focused on the perhaps lengthy interim period when autonomous vehicles share the roads with distracted, drunk, and error-prone humans.
Lawmakers face a critical choice. History affords plenty of examples of the right ways and the wrong ways to manage the regulatory transition from one paradigm to the next. Done correctly, an evolving legal system can encourage optimal investment in technologies that will increase social welfare, public safety, and sustainable energy consumption, as well as positively impact labor markets, land use, public health, and more.
If lawmakers don’t handle this correctly — well, consider Red Flag laws. At the dawn of the first automotive age, laws were passed in some areas that required a person carrying a red flag to warn people that a “horseless carriage” was coming. Pennsylvania went further, requiring that motorists “stop, disassemble their vehicle, and conceal the parts in bushes if the car frightened a passing horse.” (Only a veto by the governor kept the law, passed unanimously, from taking effect.)
Even in the last few years, we have dozens of examples of both the right and wrong kind of regulation for autonomous vehicles, where individual states are experimenting with everything from largely open legal frameworks to outright bans. Given its proximity to much of the innovation, California has given the most thought to the transition, already working through several draft regulation revisions since 2012. (The Stanford Center for Internet and Society keeps an active list of laws proposed, passed, and defeated across the U.S.)
In an effort to impose some consistency, the National Highway Transportation Safety Authority (NHTSA) belatedly released its own guidance in September for both federal and state policies, aimed at encouraging safe but accelerated deployment of autonomous vehicle technologies.
NHTSA recognizes the danger of untested new technologies but tries to balance that risk against the existing dangers of human drivers, who kill over 30,000 people a year in the U.S. alone. The sooner we get law-abiding robots on the road, the report concludes, the sooner many of those lives — along with billions of gallons of wasted fuel and an equal number of hours of productivity lost in traffic — can be saved.
For now, developers will be asked to report on 15 distinct safety assessments, including obvious topics such as data privacy and crash-worthiness, as well as existential questions about how vehicle software will resolve ethical problems such as when to break the law to get around traffic obstacles or when to prioritize the safety of one vehicle over another.
The NHTSA guidance has been well-received by developers, with the exception of a suggestion that some “pre-market approval authority” might be required, similar to design reviews of commercial aircraft overseen by the Federal Aviation Authority. According to Adam Thierer, author of Permissionless Innovation, in an op-ed: “This would require any new driverless car design to be approved by NHTSA before reaching market and could even be extended to pre-approval for new software updates on existing models. The main problem with such a pre-market approval regime is that it significantly increases the time and cost of deploying driverless cars onto the road.”
But the law of smart vehicles isn’t all reactive. As with the original introduction of the automobile, governments play a key role in building and funding new infrastructure. Smart roads and sensor-driven traffic management systems, for example, will need to replace our badly aging highway system, especially in urban areas.
Our new smart infrastructure will also require real-time, high-reliability, low-latency communications networks. These systems, known as 5G, are already being tested in the U.S. and Asia but require coordinated allocation of radio frequencies and reduced local bureaucracy in approving permits for underground cabling and overhead antennae.
To its credit, the FCC is already on the case, as is the Department of Commerce, the Department of Transportation, and the White House, along with state counterparts in many, but certainly not all, states. The Department of Transportation last month announced the winners of its Smart City challenge, which will provide $65 million in grants “to become the country’s first city to fully integrate innovative technologies — self-driving cars, connected vehicles, and smart sensors — into their transportation network.”
First place went to Columbus, Ohio, but other cities are working to get smarter sooner. According to Brent Skorup, research fellow at the Mercatus Center, Atlanta has taken an early lead in integrating smart vehicles. “Atlanta has built a dedicated fiber and electrical network in a downtown corridor in order to support roadside sensors and cameras for driverless cars,” Skorup said. “The city is currently soliciting proposals from tech companies to test and develop driverless car and roadside sensor technology.”
As these examples suggest, getting the right policies in place at the right time isn’t just a matter of allowing the sometimes chaotic transformation of affected industries. The states and the countries that provide the most-balanced regulatory environment, offering predictability and encouraging innovation with needed safeguards, will gain significant competitive advantage in the development and manufacture not just of smart cars but of the industries that support them.
And if the claims of self-driving vehicle entrepreneurs even come close to proving accurate, the secondary benefits of fully autonomous transportation will be even more substantial. More-efficient vehicles will vastly improve human productivity, air quality, and overall livability, allowing for efficiently-designed urban and suburban communities.
The cities, states, and countries that get there first will attract new investment in industries having nothing to do with transportation. That’s where the real disruptive innovation will come from.