Cars that can drive safely and reliably without human supervision are no longer science fiction. Indeed, cars are becoming smarter—they’re equipped with novel sensors for navigation and environment perception, wireless networking devices to communicate with other vehicles and roadside infrastructure, and computers for running sophisticated real-time planning and control algorithms (see “The Internet of Cars Is Approaching a Crossroads”). Several companies and universities have demonstrated vehicles that can drive autonomously in traffic, in the process building social awareness and pushing the boundaries of current regulations and risk management practices (see “Driverless Cars Are Further Away Than You Think”).
At this point, a natural question to ask is: what is the point of autonomous cars? To demonstrate that autonomy is indeed a transformative technology with the potential to drastically redefine mobility, let’s consider some of the foreseen benefits of autonomous cars, and put them in financial context.
Perhaps the most obvious benefit is increased safety. Human error is the leading cause of traffic accidents. Computers, on the other hand, do not get distracted, fatigued, or drunk. They’re not subject to road rage, and they have faster reaction times than humans. Hence, autonomous cars could dramatically reduce accident rates. Avoiding injuries and loss of life is clearly an extremely desirable, almost priceless, outcome for those affected and their families and friends. Financially speaking, the cost of traffic accidents in 2009 has been evaluated to be about $300 billion a year (pdf).
Another benefit of autonomy is the potential increase in traffic efficiency, and the ensuing reduction in congestion, as autonomous vehicles can precisely monitor one another’s position and coӧrdinate their motion to an extent impossible for human drivers. Furthermore, autonomous vehicles can travel at speeds that minimize emissions. The cost of congestion, including lost productivity and public health consequences of the pollution, has been estimated by the Texas Transportation Institute (pdf) and by the Harvard School of Public Health to be about $150 billion each year.
A fully autonomous car, which doesn’t require its driver to pay attention and be ready to take control, could also relieve people from the most burdensome aspects of driving. What is the value of the time that people can get back from driving to pursue a more desirable or productive activity, such as relaxing, watching a movie, working, or texting to their heart’s content? There are about 210 million licensed drivers in the U.S., spending on average 465 hours per year driving a car; assuming a value of time of one half the average hourly wages, i.e., $12 per hour, the societal benefit due to increased productivity and leisure time is roughly $1.2 trillion a year.
Finally, another potential major benefit of autonomous vehicles is the enabling of widespread car sharing. Car-sharing services are growing worldwide, but typically do not offer one-way, door-to-door rental options. Or, if they do, they often suffer from limited car availability. If shared cars could drive to a customer’s pickup location and return to a parking station by themselves upon drop off, they could offer an unprecedented level of convenience. Financially, car sharing distributes the cost (in terms of dollars and of time) of purchasing, maintaining, and insuring autonomous vehicles across a large user base. In addition, autonomous shared cars will eliminate the need to look and pay for a parking spot, and will reduce the time needed to walk to and from the parked vehicle. An estimate of the average benefit of autonomous vehicle sharing, assuming a sharing factor of four (i.e., each autonomous car replaces four private cars), is about $8,700 a year, which at the aggregate level amounts to $1.8 trillion each year in the U.S. More details on this analysis are available in a research study conducted by myself and colleagues.
Within the limits of the approximations and assumptions made, the above estimates suggest that autonomous driving is indeed a transformative technology, with a potential financial benefit to U.S. on the order of more than $3 trillion per year. It is interesting that the benefits due to increased safety and reduced congestion pale in comparison with those due to sharing and increased productivity. In particular, the synergy between autonomy and car sharing is readily apparent.
The economics seem huge, but these estimates must be taken with a grain of salt. For example, they don’t take into account latent demand—in other words, should a new mode of transportation arise, offering the convenience of a “robotic chauffeur” at an affordable price, will the demand for mobility increase? More generally, what would be the effect of autonomous cars on, say, congestion? Autonomous shared cars could end up traveling more miles than private cars because of the empty trips to pick up customers. On the other hand, it may be the case that empty trips will take place predominantly along routes with a low traffic volume, thus not increasing congestion significantly. These are unresolved questions, and they’re the focus of ongoing research.
Yet the analysis indicates that autonomy could transform, in a very essential way, the current model of car ownership and usage, enabling a paradigm shift toward a service-centered model of personal mobility, with possibly profound consequences on our lives, our cities, and the automotive industry.
Emilio Frazzoli is a professor of aeronautics and astronautics at the Massachusetts Institute of Technology. His current research focuses on autonomous vehicles, mobile robotics, and transportation systems.
How AI is reinventing what computers are
Three key ways artificial intelligence is changing what it means to compute.
These weird virtual creatures evolve their bodies to solve problems
They show how intelligence and body plans are closely linked—and could unlock AI for robots.
Our brains exist in a state of “controlled hallucination”
Three new books lay bare the weirdness of how our brains process the world around us.
We reviewed three at-home covid tests. The results were mixed.
Over-the-counter coronavirus tests are finally available in the US. Some are more accurate and easier to use than others.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.