The Future of Driving, Part II: Life after driving

In a world where cars can drive

In part one of our "Future of Driving" series, we looked at the current state of self-driving car technology and tried to predict how that technology would progress in the coming decades. Now we're going to assume that the technical problems we discussed can be solved and explore how self-driving cars could change society.

Some benefits of self-driving cars are obvious—less time spent behind the wheel and fewer accidents—but the consequences are likely to be much broader than that. Among the most intriguing are much greater use of taxis, more widespread use of smaller, more energy-efficient cars, the virtual elimination of parking lots, and a dramatic transformation of the retail sector.

Throughout this article we'll be linking to essays by Brad Templeton, a Silicon Valley entrepreneur who is currently the chairman of the Electronic Frontier Foundation. In recent months, Templeton has become an evangelist for self-driving vehicle technology, speaking and writing extensively about the topic. Many of the predictions we make in this story are based on ideas sketched out in Templeton's writings. If you're interested in more discussion of the topic, Templeton's web site is the place to go.

An important caveat before we get started: predicting the future is hard, and self-driving technology is still young enough that we're guaranteed to get some of the details wrong. One only has to glance through past predictions about the future to see how difficult it is to predict the social effects of new technologies. Nevertheless, we think it's worthwhile to spend some time thinking about the promise of this technology. We don't have all the answers, but we hope that talking about these benefits will inspire the next generation of engineers and entrepreneurs to turn the dream into a reality.

The deadly human driver

Image credit: evelynishere

Highway safety has improved steadily over the last half century. In the United States, five people died for every 100 million vehicle miles traveled in 1960. By 1980, cars were killing 3.3 people per 100 million vehicle miles. In 2000, the rate was down to 1.5. But progress has slowed since the turn of the century, and this may be because most of the low-hanging fruit—seatbelts, anti-lock brakes, stronger drunk-driving enforcement—have already been plucked. The introduction of advanced collision-prevention software, which we discussed in our first installment, will help to push accident rates lower. And we can expect the introduction of fully self-driving cars to push accident rates lower still.

That's important because for all our progress, we still lose far too many people on our highways. Here in the United States, there were six million car crashes in 2006, injuring 2.5 million people and killing 42,000. Worldwide, according to World Health Organization figures, cars kill about 1.2 million people each year and injure 50 million. Many of these crashes are alcohol-related: the National Highway Traffic Safety Administration estimates that in 2004, 14,400 Americans died in crashes involving at least one driver with a blood-alcohol level of .08 or higher. Other crashes are caused by drivers who are fatigued, distracted, or reckless.

Self-driving cars will never be drunk, tired, or inexperienced. They should make designated drivers as anachronistic as linotype operators, freeing suburbanites from worrying about how they'll get home after an evening of drinking. Similarly, people on long road trips won't need to worry about falling asleep at the wheel. They'll be able to take naps while their cars drive for them. Hundreds of truckers die every year, and the automation of the trucking industry could eliminate the need for human truck drivers, saving hundreds of lives in the process. And far fewer teenagers will have their lives cut tragically short due to crashes caused by their lack of experience behind the wheel.

Image credit: bloomsberries

In short, a car that drives as well as the best human drivers would save tens of thousands of lives in the United States and hundreds of thousands of lives worldwide. And most likely, we'll be able to do even better than that. Computers have much faster reaction times than humans do, and they will be "looking" in all directions simultaneously. Self-driving cars may be able to avoid many of the mistakes that even experienced human drivers make. They won't have blind spots, they'll have better sensors, and they will be able to react almost instantaneously to unexpected problems, giving them the ability to recover from dangerous situations that no human driver could have handled.

Dramatically fewer accidents is the most obvious—and probably the most important—benefit of self-driving technology. But self-driving technologies will also bring significant changes to peoples' daily lives. Next we'll consider how self-driving technology could transform the transportation system, reducing congestion and sprawl and dramatically improving energy efficiency.

Human drivers are expensive

Cars came to dominate the transportation industry in the 20th century because they offered unparalleled convenience. Trains, buses, airplanes, and ships have a variety of advantages, but none of them will pick you up at your front door and transport you directly to the destination of your choice. Because time is scarce, most of us have been willing to pay a premium for the convenience and reliability offered by the automobile.

Our demands for convenience have led to some wasteful behaviors. People want the ability to haul cargo, so they buy SUVs and pickup trucks with cargo space they rarely use. People want convenient parking, so businesses have built vast parking lots for their employees and customers that sit empty much of the time. The landscape is littered with gas stations so that people never have to drive too far to find one. And despite the concerns of environmentalists, internal combustion engines dominate the market because no other technology offers a comparable range at a reasonable cost.

The situation is different in the densest cities. Visit Manhattan, and you'll see hundreds of yellow taxi cabs and few privately-owned cars. Taxis avoid many of the inefficiencies that otherwise flow from automobile ownership, because customers use taxicabs for exactly as long as they need them and then release them for others to use. Because cars aren't waiting around for their owners, the need for parking is dramatically reduced. And taxi drivers rarely drive SUVs or pickup trucks, preferring cheaper and more fuel-efficient cars.

Image credit: Tomás Fano

Taxis have the additional advantage of saving customers the hassles of automobile ownership. Taxi passengers don't have to worry about filling up the gas tank, taking the car to the shop for maintenance, or shopping for auto insurance. All of those details are handled by the taxi company and bundled with the cost of cab fare.

So if taxis are so great, why aren't they popular everywhere? The problem is that when you rent a taxi, you're not only renting a car, but you're hiring a driver as well. And human labor is expensive. So taxis only make sense financially in places where parking is so expensive or hard to find that driving your own car isn't worth the trouble. Everywhere else, the cost of the driver is high enough that driving and parking your own car is a better deal.

Self-driving taxis

Self-driving cars offer all the benefits of taxis for the cost of a traditional car. A self-driving vehicle will be able to show up on demand, transport passengers to a destination, and then drive off to pick up more passengers, refuel, or find a parking space. When self-driving taxis are readily available, many people—even far from dense urban areas—will find renting both cheaper and more convenient than owning a vehicle.

Image credit: FaceMePLS

One way customers might hail self-driving taxis may be using an application on their smartphones. Many phones already know their current locations, and such phones will be ubiquitous within a few years. A customer could enter his destination, number of passengers, and other preferences. The taxi company would then look for the nearest vehicle that meets the customer's range and capacity needs. It should be feasible to pre-position vehicles around a metropolitan area so that any hail can be served within seconds in high-density areas and within a few minutes in the suburbs. And because there's no human driver waiting inside the cab, it will cost very little to hail a vehicle well in advance and have it wait nearby until the customer is ready to leave.

Should people still choose to own cars in the new world of automated taxis, they could make more efficient use of them than they currently do. Today, the cars people drive to work sit idle all day, taking up space in a parking lot. If cars could drive themselves, they could spend the workday ferrying other passengers around downtown, re-charging their batteries, delivering packages, or performing other useful work. That has twin benefits: it helps earn revenue for the car owner, and it avoids having to pay for parking. The result will be that in most circumstances, hiring a cab (or allowing your car to be used as a cab when you're not in it) will be dramatically cheaper than owning a car and having it parked all day.

Image credit: faceless b

A shift toward widespread taxi use would mean a dramatic reduction in the acreage devoted to parking. If most people rent rather than own their cars, then most cars will be on the road serving passengers most of the time, and may never require a parking space at all. Moreover, as Brad Templeton points out, parking lots would work differently if cars could move themselves at a moment's notice. We can only park today's cars in places where we're sure they won't be in anyone's way. But cars that can move themselves could park in lots of places—in front of driveways or fire hydrants, in strangers' driveways—that ordinary cars cannot. If they find themselves in someone's way, they can quickly move and find somewhere else to park. The same point applies in parking lots. Self-driving cars can safely double- or triple-park, dramatically improving space utilization. When a car needs to leave, it can signal its neighbors to move out of the way and let it get through.

As the demand for parking drops, most of the land now devoted to parking will be put to more productive use. In the long run, as the distance between buildings is no longer inflated by the presence of parking lots between them, suburbs are likely to become more compact and pedestrian-friendly. And that means less sprawl and congestion.

Self-driving cars are likely to reduce congestion in other ways too. Thanks to shorter reaction times, self-driving cars will be able to drive closer together, enhancing throughput. Self-driving vehicles won't be prone to rubber-necking and other erratic behavior by human drivers that can snarl traffic. And self-driving cars will probably be able to automatically poll real-time traffic databases in order to plot routes that will avoid the most congested roads.

Smaller, lighter, cheaper

Self-driving technologies are likely to make small, light, and efficient electric vehicles practical for everyday use. The limited storage capacity of batteries has been a major stumbling block for all-electric vehicles like the EV-1, which took more than an hour to charge its battery and only went about 50 miles on a typical charge. In contrast, cars with traditional internal combustion engines can go hundreds of miles on a tank of gas, and refilling the tank takes only a few minutes.

Electric car manufacturers can add more batteries, but this increases the cost of the vehicle (limiting its market appeal) and its weight (limiting the energy-efficiency benefits). Plug-in hybrids deal with this problem by adding an internal combustion engine, but this too adds weight and costs, as well as reducing their environmental benefits.

As Brad Templeton has noted, small, light, and cheap electric vehicles would be much more appealing if they could go off and charge themselves between trips. The vast majority of urban trips are shorter than the 50-mile range of the EV-1. The problem is that recharging electric vehicles between trips is a major hassle. But that ceases to be a problem if cars can recharge themselves while their passengers are busy doing something else.

Moreover, in a world with ubiquitous self-driving taxis, passengers will have no reason to worry that their car won't be able to make it through a full day of driving. If a vehicle's batteries get close to empty, it can easily rendezvous with another self-driving taxi whose batteries are full. The passengers can switch vehicles, and the old vehicle can go off to recharge itself while the new vehicle carries the passengers to their destination with minimal disruption. That's a bit of a hassle, of course, but it's nowhere near as bad as running out of gas is today.

Image credit: Paul!!!

Of course, small electric vehicles won't be a good choice for every trip. People sometimes need to carry multiple passengers, haul large amounts of cargo, and go on long road trips. Small electric vehicles won't be suitable for these tasks. But rather than driving a larger, more expensive vehicle all the time, many people will find it makes sense to rent vans, trucks, or internal-combustion-engine cars only when they need the extra range or capacity. In most circumstances, such as commuting and running errands, the range and capacity of small electric vehicles will be more than adequate.

Saving time

Another important benefit of self-driving cars is that it will free up billions of person-hours that are currently devoted to driving. While some people like to drive, most people find it tedious and frustrating—especially during rush-hour commutes. When self-driving cars arrive on the scene, people will be able to spend their morning commutes being productive—eating breakfast, doing their hair and makeup, watching the news—while their cars worry about the traffic.

Eventually, self-driving technology may lead to a dramatic redesign of car interiors. Rather than demanding that one passenger sit in the front-left seat (or the front-right seat in countries that do things backwards) behind a steering wheel, car interiors may look more like the back of limousines, with passengers arranged to face one another and the interior space better optimized for comfort and productivity. Couches, desks, televisions, and other amenities could become commonplace. It might even be possible (and legal!) to have a beer on the drive home from work.

Image credit: timsamoff

Self-driving technology will be an even bigger timesaver for parents. Self-driving cars will be able to safely and automatically pick children up from school, day care, or violin practice and deliver them home, relieving parents of the need to play chauffeur. Self-driving technologies will dramatically improve the lives of the elderly, the blind, and others who lack the ability to drive themselves.

Self-driving technology will also make long road trips much more pleasant and convenient. Specialized "sleeper cars" could even provide comfortable overnight sleeping space, allowing people to travel hundreds of miles while they sleep. While it's unlikely that self-driving cars will ever be a practical way to get from New York to Los Angeles, sleeper cars could become a viable alternative for medium-range trips such as from San Diego to San Francisco, a trip that currently takes about eight hours by car. Families on vacation can spend their entire day seeing the sights, and leave the driving to their car (or cars, for larger families) while they're asleep.

A retail revolution

Self-driving technologies may also revolutionize the retail industry. We've already mentioned the rise of automated trucking, which would slash the costs of shipping while saving hundreds of truck drivers' lives every year. A smaller version of the same technology, which Brad Templeton dubs the deliverbot, could transform every retail store into a rapid delivery service. Rather than running to the grocery store for a gallon of milk, consumers will be able to visit a store's web site, select a gallon of milk, click "deliver it now," and have it appear at the customer's door in a matter of minutes. Every restaurant will gain the ability to deliver hot food directly to your door, and all manner of consumer products will be just a click and a few minutes away.

If it seems wasteful to dispatch a miniature car to deliver every customer's order, remember that in today's retail market, most customers come to the store in full-size cars. Because automated delivery vehicles wouldn't need to carry passengers, they could be far smaller, cheaper, and more energy-efficient than even the smallest passenger cars. If they avoided freeways, they wouldn't need to be very fast. Their interiors could be little more than a bare shell. (Obviously, stores that sold perishable goods might need temperature controls). And many stores may be able to economize by serving several adjacent customers with each trip. The cost and energy consumption of each delivery would be a small fraction of what it currently takes for a customer to drive to Wal-Mart and back, and the shipping costs would be much less than Amazon currently charges.

Image credit: dave_mcmt

Indeed, the rise of autonomous delivery vehicles may completely erase the distinction between online retailers like Amazon and traditional brick-and-mortar retailers like Wal-Mart. Wal-Mart and Amazon are already competitors in a limited sense, but the rise of rapid, cheap delivery would eliminate most of the remaining differences. Amazon will have to build more warehouses and offer faster delivery options to compete with Wal-Mart, while Wal-Mart will need to expand its selection and beef up its web site to keep stay up with Amazon. This may spell the doom of big-box retail store as we know it today. Wal-Mart has never had an especially pleasant ambiance. When customers can get the selection, low prices, and instant gratification of Wal-Mart from the comfort of their own homes, who is going to want to go to the actual store?

Many of today's retail outlets, then, may be transformed into Amazon-style warehouses that take orders via the web and send out deliveries using autonomous vehicles. Of course, there will still be room for specialized retailers that add value to their products through expert advice or other personal services. There are also industries, such as clothing, where customers want to see the merchandise before they make their selections. These businesses may look very much like today's retail stores. But for many routine purchases—groceries, consumer electronics, household supplies—autonomous delivery vehicles may render the brick and mortar retail store obsolete.