Self-Driving Cars

Background

Self-driving cars are autonomous automobiles that can operate without human intervention. However, some people also use the term “self-driving cars” to refer to cars with advanced driver assistance systems.

Cars with ADAS can perform some driving functions—such as maintaining speed or following a route (see overview)—but still require a driver to be ready and able to take over the controls. Because this may cause confusion, this page solely focuses on truly autonomous self-driving cars.

How Self-Driving Cars Work

While self-driving cars are not yet on the market, proponents say that widespread adoption of autonomous vehicles would have a host of theoretical benefits. These include improved safety (meaning fewer crashes), economic benefits associated with fewer vehicle accidents, improved mobility for people unable to drive, fewer traffic delays, and environmental benefits associated with less traffic and more efficient driving practices.

Self-driving cars rely on a suite of sensors—including radar, lidar (which uses lasers), and photographic technologies—to perceive the world around the vehicle. Data collected by these sensors are used to create a “map” of the vehicle’s environment that identifies everything from street signs and traffic signals to other vehicles and pedestrians.

That map is constantly being updated, allowing an artificial intelligence program to make decisions about when to accelerate, brake, or turn in order to travel safely from Point A to Point B.

Some researchers are also working on concepts where self-driving cars are networked to each other and to roadside technologies, such as traffic signals. These “vehicle-to-everything” networks—or V2X—hold promise for substantially improving traffic congestion and reducing the likelihood of accidents. For example, a V2X network would allow traffic signals to know how many cars are coming from every direction and coordinate both signal changes and the speed of approaching autonomous vehicles to improve traffic flow through an intersection. However, the complex nature of V2X networks also means they are likely further from realization.

Current State

As of the fourth quarter of 2024, there are no fully autonomous self-driving cars on the market for consumers. Depending on where you live, their deployment may face legal hurdles—laws governing autonomous vehicles vary widely across the United States (see overview). Some states require a human driver behind the wheel of self-driving cars, some don’t, and some have yet to pass any regulations.

However, some companies are testing self-driving cars on public roads. More than three dozen companies—from Waymo to Mercedes-Benz—have secured licenses to test their autonomous (and ADAS) cars in California alone, with testing taking place from Sacramento to San Diego.

The tests are designed to help designers ensure their self-driving cars work the way they’re supposed to—and learn from cases where things don’t go as planned. For example, pilot projects involving “robotaxis” have already been flagged for causing accidents and committing traffic violations.

Future

While there is widespread agreement that self-driving cars will eventually make it to the marketplace, there is no consensus on when they will become available. And despite their best efforts, the autonomous vehicle industry has had to keep postponing target dates for widespread release.

In addition to technical and legislative hurdles, one question looming over the future of self-driving cars is whether consumers will accept them—a January 2024 survey of US adults found that 93% had concerns about autonomous vehicles. Given the scope of potential benefits associated with autonomous vehicles, the industry will almost certainly keep working to address these challenges.