As your workday is ending, you grab your smartphone and tap your driverless valet app to summon your electric vehicle from the parking lot. When it picks you up in front, the car door opens, as your key is now facial recognition technology. Your car’s interior is warmed or cooled based on the outside temperature and the seat is adjusted to your custom favorite position.
Your car is a Wi-fi hotspot—intelligent, connected, and continuously learning more about you. As you start your commute home, your vehicle’s audio system is playing your favorite after-work programs, which it has forecast based on your tastes. You opt to steer until you get into rush-hour gridlock, when you switch to over autopilot.
Your car is equipped with a dozen cameras and 30 radar sensors. With the help of the Internet, it will now take over the tedious task of stop-and-go driving. Freed from the wheel, you can take a nap, play a game, or watch something on your high-definition television.
Arriving home, you are dropped off at your door and put the car back into driverless valet mode, so it can go find a parking spot and wait for your next smartphone command. Never again will you waste time looking for parking spots or waste fossil fuels, now that you are driving your zero-emissions autonomous vehicle.
This may sound farfetched, but cars are available now that can do all of these things and have been road-tested on the West Coast with mostly passing marks. Tesla, Lucid Motors, and Faraday Future already manufacture a high-performing electric SUV with commuting ranges of 350 miles per charge. All have auto-driving features, such as traffic aware cruise control that detects the center of your lane with your hands off the wheel, and emergency braking if the car approaches an object in the road. Some of these features are sneaking into 2017 vehicles on Ford, Honda, and Nissan lots.
The biggest holdups for self-driving vehicles are regulation and infrastructure. Autonomous cars replace the driver with multiple sensors that detect other moving and non-moving objects and synthesize the vehicle’s location from constantly updating maps. What kinds of lighting and signage do they need? What happens when a stoplight goes out or a sign is obstructed? Is it possible to program a vehicle to respond to the complex situations encountered on the road? In the event of a crash, who is liable? Should they be programmed to automatically obey speed limits, or to keep pace with prevailing traffic?
So far, the only fatal crash involving a driverless car happened last May, when a Tesla struck a jack-knifed semi on a Florida highway. The Tesla couldn’t distinguish between the white truck and the bright sky, so it didn’t apply the brakes.
Driverless cars are much safer—at least in theory. They don’t drive drunk or tired. They don’t text and drive. They aren’t prone to road rage. They don’t make human mistakes. But they don’t apply human logic either, setting up a classic Trolley Problem in an emergency: If a driverless car encounters an unavoidable crash, how does it decide whether to hit the cement barricade or the pedestrian?
In many ways, driverless cars are still years away from the market. Current prototypes start at $150,000 and easily crawl up to $200,000 with fully loaded features. Faraday Futures (named after the inventor of the first electric motor in 1821) is touting its smartphone, keyless FF91 electric vehicle, whose interior turns into a plush theater.
But, in practice, a driver still has to be available to take over, and the distractions that make driverless cars so appealing are, ironically, the same features that could pose a hazard, as distracted drivers have a hard time shifting their attention back to driving quickly enough to react.
Last year, Uber rolled out a fleet of autonomous cars in Pittsburgh, complete with an iPad to entertain passengers and allow them to communicate with the vehicle. Uber’s Pittsburgh cars logged over 20,000 miles, but a human driver needed to intervene about once every mile.
If autonomous cars ever become widespread, they could change society in ways we haven’t even thought of yet. They could free up land currently needed for roads and parking, increase road capacity and minimize traffic congestion, and profoundly increase mobility for the young, the elderly, and the disabled. They could also displace workers in trucking and transportation, cut off city revenues from traffic fines and parking fees, and have unpredictable ramifications for public transit.
The whole face of commerce could change if it became cheaper to transport goods directly to individual consumers rather than to centralized markets. Cheaper, less stressful commutes could incentivize moving away from cities, increasing urban sprawl. Alternatively, since cars cause so many unpleasant aspects of city life, driverless vehicles might spark an urban renaissance.
It’s a brave new world out there—is anybody steering this thing?