Driverless Cars: The Next Transportation Revolution

By Randal O'Toole -- February 3, 2015

“Imagine going to work and then sending your car home to pick up your kids and take them to school. Self-driving cars won’t require able-bodied drivers, extending the benefits of mobility to many more people….  Personally, I am looking forward to sending my dogs to the vet in a self-driving car so I don’t have to go myself.”

“Though technical challenges remain, the real challenges are legal and institutional. Even they are not insurmountable. Instead of individuals insuring their cars, automakers might buy insurance against liability suits and pass the cost onto auto buyers. Since self-driving cars can record every event leading up to an accident, it will be easy to determine who is at fault in any crash.”

There’s a self-driving car in your future. Most experts believe that cars capable of driving themselves in almost any situation will be on the market by 2020, and cars that don’t even have the option of being driven by humans will be on the market by 2030.

This technology will have as profound an impact on the way we live as the moving assembly lines that allowed Henry Ford to cut the price of his Model T in half and double worker pay. Affordable cars changed the shape of twentieth-century cities and provided almost everyone with access to affordable housing, low-cost consumer goods, and a wide range of social and recreational opportunities.

Since Ford’s time, cars have required licensed drivers devoting all of their time to watching the road and controlling their vehicles. Self-driving cars will free us of that requirement, changing the way we live in the twenty-first century.

Imagine being able to work, read, or watch entertainment on the way to work. More than half of Americans say the time required for travel is a bigger limiting factor than the cost of travel. Self-driving cars will reduce that time cost and change the way we think about travel since we’ll be able to be as productive in motion as we are in the office.

Imagine going to work and then sending your car home to pick up your kids and take them to school. Self-driving cars won’t require able-bodied drivers, extending the benefits of mobility to many more people. Senior citizens will be able to live in their homes longer as they can take their self-driving cars shopping without having to worry about accidents. Personally, I am looking forward to sending my dogs to the vet in a self-driving car so I don’t have to go myself.

Imagine grocery shopping on line and then sending your self-driving car to a robo-supermarket to pick up your order. People may still want to visit farmers’ markets to choose their own produce and other fresh foods, but self-driving cars will eliminate the need to walk up and down grocery aisles looking for canned and package goods.

Imagine not even owning a car but using an app on your smart phone to call up a car at any time that can take you from your door to the front door of your destination. Self-driving car sharing will probably reduce auto ownership rates without greatly reducing miles of driving. Such car sharing, however, will render obsolete most of our heavily subsidized urban transit systems.

Self-driving cars will reduce traffic congestion, making it possible for more people to travel without building a lot of new roads. Self-driving cars will also be safer: Google has tested its self-driving cars in California over more than 700,000 miles of travel with only one accident, which was when someone rear-ended one of Google’s cars at a stop light.

Google’s software can deal with pedestrians, bicyclists, construction detours, and just about any other obstacle on the roads. The one technical challenge that remains is dealing with snow or ice that is covering road stripes, but eventually cars will probably use GPS to determine the exact locations of highway lanes.

Though technical challenges remain, the real challenges are legal and institutional. Even they are not insurmountable. Instead of individuals insuring their cars, automakers might buy insurance against liability suits and pass the cost onto auto buyers. Since self-driving cars can record every event leading up to an accident, it will be easy to determine who is at fault in any crash.

You can already buy cars that contain much of the hardware needed to make them self-driving. Mid-priced cars such as Fords, Mazdas, and Subarus come with adaptive cruise control, which keeps the cars a fixed distance behind the cars in front; lane assist, which alerts the drivers if they drift out of their lanes; and collision avoidance, which detects cars that might be in the drivers’ blind spots and brakes or otherwise overrides the driver’s control to prevent accidents.

More expensive cars, such as some Mercedes and Lincolns, are capable of actually following highway stripes and keeping the vehicle in the center of the lane. Worries about liability, however, have led manufacturers to disable this technology in the United States if drivers take their hands off the steering wheel for more than 15 seconds.

Turning any of these cars into completely self-driving cars would require little more than the addition of a few sensors and a software upgrade costing about the same as a laptop computer. This means that when self-driving cars hit the market in 2020 or so, the owners of a lot of existing cars may decide to retrofit their cars, which would accelerate the adaptation of this new technology.

One of the first impacts we will see is a reduction in congestion. We often say people driving at 60 miles per hour should leave six car lengths in front of them. If the average car is 18 feet long, this should allow a single freeway lane to move 2,500 vehicles per hour. In fact, actual measurements show freeway lanes rarely move more than 2,000 vehicles per hour, and never more than 2,200.

The reason actual capacity is below 2,500 is slow human reflexes: if one person slows down 1 mph, the next person may slow down 2 mph, and so forth. If there are enough cars in line, this can bring traffic to a complete stop. If as few as 25 percent of cars use adaptive cruise control, these slow-downs would disappear, effectively allowing roads to operate at full capacity during more hours of the day.

The Subaru adaptive cruise control, for example, allows cars to follow barely a second behind the car in front. At 60 miles per hour, this could boost lane capacities to 2,500 to 2,800 cars per hour. Eventually, fully self-driving cars should double or triple road capacities.

The resulting reductions in congestion will save energy. As cars become safer, it may be possible to build lighter-weight vehicles, saving even more energy. However, this energy savings could be offset by increases in driving as more people (and my dogs) take advantage of their increased mobility.

Most state laws require that a licensed driver be behind the wheel of a car, but they don’t require that a licensed driver actually be in control. This means people can use self-driving cars so long as they have a license and are ready to take over if need be. California, Nevada, and a few other states have legalized the experimental and, eventually, routine use of cars without licensed drivers, and Google and other companies are conducting most of their tests in these states.

The federal government funded early research on self-driving cars in the 1990s, culminating in a successful test in which eight self-driving cars drove at speed on a California freeway one car-length apart from one another. At the conclusion of that test, the Department of Transportation cancelled all federal funding. Since then, all research has been funded either by auto makers, auto suppliers, or software companies like Google, and the lesson these companies learned was not to rely on the government for any of their systems to work.

Since all of the brains behind a self-driving car are on board the car itself, there will be no central control monitoring where people go. However, the Obama administration wants to mandate that all new cars be built with vehicle-to-infrastructure communication systems that would allow the government to monitor individual cars, give hackers opportunities to disrupt travel flows (perhaps by putting out false signals that some roads have been closed due to accidents), and even open the possibility that government bureaucrats could remotely turn off your car if they decide you’ve driven too much.

Contrary to claims by some, vehicle-to-infrastructure communication is not essential for self-driving cars, and Google and other companies have specifically said they don’t need it for their cars to drive themselves.

Vehicle-to-infrastructure communications would require building a huge communications network into our transportation infrastructure. This seems like an absurd goal when we can’t even maintain the infrastructure we have. Rather than try to build such a network, engineers designing self-driving cars say they simply want transportation agencies to fill the potholes, make sure lane stripes are clearly visible, and perhaps develop highway signs that are consistent across the nation.

While we can speculate how self-driving cars will change the way we live, most of the effects will be totally unpredictable. How will car sharing influence auto ownership rates? How much more travel will take place because more people can drive without a license? How much less travel will take place because people who rent cars will have to pay the average cost of driving rather than the marginal cost? How will self-driving cars affect parking requirements? Instead of demonizing single-occupant vehicles, will environmentalists begin to demonize zero-occupant vehicles?

No one knows the answers to these questions. Yet federal law requires that every state and urban area in America write long-range transportation plans based on speculation and guesses. Instead of wasting time on such plans, transportation agencies should concentrate on solving today’s problems and let the future take care of itself.

———–

Randal O’Toole (rotoole@cato.org) is a Cato Institute Senior Fellow working on urban growth, public land, and transportation issues. His books include Reforming the Forest Service (1988);The Vanishing Automobile and Other Urban Myths (2001), and The Best-Laid Plans: How Government Planning Harms Your Quality of Life, Your Pocketbook, and Your Future (2007), which calls for repealing federal, state, and local planning laws and proposes reforms that can help solve social and environmental problems without heavy-handed government regulation, and Gridlock: Why We’re Stuck in Traffic and What to Do About It (2010).

O’Toole (rotoole@cato.org) travels extensively and has spoken about free-market environmental issues in dozens of cities. An Oregon native, O’Toole was educated in forestry at Oregon State University and in economics at the University of Oregon.