Autonomous Car

An autonomous car, also known as a "robotic car" or "driverless car," is a vehicle designed to operate without the guidance or control of a human driver. Engineers began designing prototypes and control systems for autonomous vehicles as early as the 1920s, but the development of the modern autonomous vehicle began in the late 1980s.

Proponents of autonomous car technology believe that driverless vehicles will reduce the incidence of traffic accidents, reduce fuel consumption, alleviate parking issues, and reduce car theft, among other benefits. One of the most significant potential benefits of "fully autonomous" vehicles is to provide independent transportation to disabled individuals who are not able to operate a traditional motor vehicle. Potential complications or problems with autonomous vehicles include the difficulty in assessing liability in the case of accidents and a reduction in the number of driving-related occupations available to workers. Starting in the early 2020s, several highly-publicized accidents involving self-driving or partially autonomous vehicles provoked debate over the pros and cons of this technology.

Self-driving cars are ranked at five levels of autonomy. A level 0 car is a traditional vehicle that is not able to drive itself. Most of these ranks, including conditional autonomy (level 3), require some degree of assistance or intervention from a human driver. Only vehicles classified as level 5, which indicates full autonomy, are able to drive themselves at all times in any conditions.

Background

Autonomous car technology has its origins in the 1920s, when a few automobile manufacturers, inspired by science fiction, envisioned futuristic road systems embedded with guidance systems that could be used to power and navigate vehicles through the streets. For instance, the Futurama exhibit at the 1939 New York World’s Fair, planned by designer Norman Bel Geddes, envisioned a future where driverless cars would be guided along electrically charged roads.

Until the 1980s, proposals for autonomous vehicles involved modifying roads with the addition of radio, magnetic, or electrical control systems. During the 1980s, automobile manufacturers working with university engineering and computer science programs began designing autonomous vehicles that were self-navigating, rather than relying on modification of road infrastructure. Bundeswehr University in Munich, Germany produced an autonomous vehicle that navigated using cameras and computer vision. Similar designs were developed through collaboration between the US Defense Advanced Research Projects Agency (DARPA) and researchers from Carnegie Mellon University. Early prototypes developed by DARPA used LIDAR, a system that uses lasers to calculate distance and direction. In July 1995, the NavLab program at Carnegie Mellon University produced one of the first successful tests of an autonomous vehicle, known as "No Hands Across America."

The development of American autonomous vehicle technology accelerated quickly between 2004 and 2007 due to a series of research competitions, known as "Grand Challenges," sponsored by DARPA. The 2007 event, called the "Urban Challenge," drew eleven participating teams designing vehicles that could navigate through urban environments while avoiding obstacles and obeying traffic laws; six designs successfully navigated the course. Partnerships formed through the DARPA challenges resulted in the development of autonomous car technology for public use. Carnegie Mellon University and General Motors partnered to create the Autonomous Driving Collaborative Research Lab, while rival automaker Volkswagen partnered with Stanford University on a similar project.

Stanford University artificial intelligence expert Sebastien Thrun, a member of the winning team at the 2005 DARPA Grand Challenge, was a founder of technology company Google’s "Self-Driving Car Project" in 2009, which is considered the beginning of the commercial phase of autonomous vehicle development. Thrun and researcher Anthony Levandowski helped to develop "Google Chauffeur," a specialized software program designed to navigate using laser, satellite, and computer vision systems. Other car manufacturers, including Audi, Toyota, Nissan, and Mercedes, also began developing autonomous cars for the consumer market in the early 2010s. In 2011, Nevada became the first state to legalize testing autonomous cars on public roads, followed by Florida, California, the District of Colombia, and Michigan by 2013.

In May 2013, the US Department of Transportation’s National Highway Traffic Safety Administration (NHTSA) released an updated set of guidelines to help guide legal policy regarding autonomous vehicles. The NHTSA guidelines classify autonomous vehicles based on a five-level scale of automation, from zero, indicating complete driver control, to four, indicating complete automation with no driver control.

Between 2011 and 2016, several major manufacturers released partially automated vehicles for the consumer market, including Tesla, Mercedes-Benz, BMW, and Infiniti. The Mercedes S-Class, which featured automated systems options including parking assistance, lane correction, and a system to detect when the driver may be at risk of fatigue.

According to a November 2014 article in the New York Times, most manufacturers at that time were developing vehicles that would require "able drivers" to sit behind the wheel, even though the vehicle’s automatic systems would operate and navigate the car. Google’s "second generation" autonomous vehicles were an exception, as the vehicles lacked steering wheels or other controls, therefore making human intervention impossible. According to Google, complete automation reduced the possibility that human intervention will lead to driving errors and accidents, although a number of critics voiced skepticism toward this opinion. Google argued further that fully autonomous vehicles could open the possibility of independent travel to the blind and individuals living with a variety of other disabilities that make it more difficult to drive a conventional car. In September 2016 Uber, a ridesharing company, launched a test group of automated cars in Pittsburgh. They started with four cars that had two engineers in the front seats to correct errors. The company rushed to be the first to market and planned to add additional cars to the fleet and have them fully automated.

During the 2010s, electronic vehicle (EV) manufacturer Tesla Inc. emerged as a major player in the self-driving car industry, and Tesla CEO Elon Musk became known for his promotion of the technology. In 2014 the company began including its Autopilot technology, which provided driver assistance but not full automation, in its cars. The system included features such as autosteering to keep the vehicle from swerving out of its lane and warnings when no hands were on the steering wheel.

Beyond Autopilot, Tesla also began equipping its cars with the hardware it claimed would eventually enable fully autonomous driving, including exterior-facing cameras and radar. In mid-October 2016 the company announced that all its vehicles would now carry such features, though the software controlling the self-driving ability would be withheld until it was fully developed and regulatory issues were settled.

Despite optimistic predictions by Musk and other key figures at Tesla, several crashes involving Tesla's Autopilot and self-driving test vehicles fueled doubts about the technology's practicality; notably, a Tesla car with Autopilot engaged was involved in a fatal crash in California in March 2018. In October 2018, Tesla removed "Full Self-Driving"—still meaning only that the car was equipped with hardware that Musk anticipated would eventually make self-driving possible—from the car customization options on its website, citing confusion among buyers.

Autonomous Cars Today

By the 2020s, modern autonomous vehicles utilized laser guidance systems, a modified form of LIDAR, as well as global positioning system (GPS) satellite tracking, visual computational technology, and software that allows for adaptive response to changing traffic conditions. Companies at the forefront of automated car technology had also begun experimenting with computer software designed to learn from experience, thereby making the vehicle’s onboard computer more responsive to driving situations following encounters. Advances in artificial intelligence (AI) during the early 2020s were also expected to have an impact on self-driving technologies.

A 2022 survey from Policygenius revealed that 76 percent of respondents did not believe that self-driving vehicles were accurate or safe, with 73 percent indicating that they felt less safe knowing that other people on the road are traveling in cars with self-driving features. More than 30 percent believed that a car with full self-driving capability would still require constant attention, and 80 percent said they would not pay more for a car with self-driving features. Another study by AAA in 2022 indicated that 85 percent of people at that time were fearful of self-driving technology and would not transport their loved ones in a car with self-driving features.

Part of the problem with the unavailability of these vehicles had to do with safety. Statistics released by US safety regulators by that time reported nearly four hundred crashes of vehicles with partially automated driver-assist systems. Two hundred and seventy-three of these crashes involved Teslas, resulting in the auto manufacturer facing a number of lawsuits, as well as a National Highway Traffic Safety Administration (NHTSA) investigation. In July 2022 a jury awarded a $10.5 million settlement to the family of one victim who died in a fatal accident involving a Tesla. In February 2023, after government regulators cited an increased risk of accidents, company recalled 362,000 vehicles with FSD technology. In April 2024 Tesla settled a different lawsuit brought by a man who had died in a different fatal accident in 2018 that involved the company's Autopilot technology.

By mid-2024, after investing billions of dollars, not one of the approximately fifteen manufacturers developing fully autonomous vehicles worldwide at that time had released one for commercial sale, although research, development, and occasional road tests continued in the US, Japan, China, Germany, and numerous other countries. By that time a number of companies, including Honda and Mercedes, offered consumers in some countries access to vehicles that had "conditional autonomy," meaning that they could operate themselves to a certain degree but still needed the intervention and supervision of a human driver.

Bibliography

Bogost, Ian. "The Secret History of the Robot Car." Atlantic. Atlantic Monthly Group, 14 Oct. 2014. Web. 16 Nov. 2014.

Garsten, Ed. "What are Self-Driving Cars? The Technology Explained." Forbes, 23 Jan. 2024, www.forbes.com/sites/technology/article/self-driving-cars/. Accessed 6 Aug. 2024.

Klender, Joey. "Survey shows Americans don't trust self-driving cars. They also don't know you can't buy one." Tesla Rati, 14 Sept. 2022, www.teslarati.com/americans-self-driving-cars-less-safe-dont-exist-survey-policygenius/. Accessed 28 Dec. 2022.

Moon, Mariella. "Now California's DMV Can Allow Fully Driverless Car Testing." Engadget, 3 Apr. 2018, www.engadget.com/2018/04/03/california-fully-driverless-car-testing/. Accessed 3 Apr. 2018.

Mussen, Melanie and Sara Routhier. "What States Allow Self-Driving Cars? (2022 Update)" AutoInsurance.org, 29 June 2022, www.autoinsurance.org/which-states-allow-automated-vehicles-to-drive-on-the-road/. Accessed 28 Dec. 2022.

"Nearly 400 Car Crashes in 11 Months Involved Automated Technology, Companies Tell Regulators." NPR, 15 June 2022, www.npr.org/2022/06/15/1105252793/nearly-400-car-crashes-in-11-months-involved-automated-tech-companies-tell-regul. Accessed 28 Dec. 2022.

"Self-Driving Cars." United States Environmental Protection Agency, 2 May 2024, www.epa.gov/greenvehicles/self-driving-vehicles. Accessed 6 Aug. 2024.

"Tesla Settles Lawsuit over Man’s Death in a Crash Involving its Semi-Autonomous Driving Software." AP News, 9 Apr. 2024, apnews.com/article/tesla-autopilot-lawsuit-settlement-f4c19ce05e17669de212fc262265e351. Accessed 6 Aug. 2024.

"US Department of Transportation Releases Policy on Autonomous Vehicle Development." NHTSA.gov. National Highway Traffic Safety Administration, 30 May 2013. Web. 16 Nov. 2014.

Wakabayashi, Daisuke. "Uber's Self-Driving Cars Were Struggling before Arizona Crash." The New York Times, 23 Mar. 2018, www.nytimes.com/2018/03/23/technology/uber-self-driving-cars-arizona.html. Accessed 3 Apr. 2018.