Self-driving car#Level 5

{{Main|History of self-driving cars}}

Experiments have been conducted on advanced driver assistance systems (ADAS) since at least the 1920s.{{cite news |date=1926 |title='Phantom Auto' will tour city |work= Milwaukee Sentinel |page=4 }} Cited in {{cite conference |doi=10.1145/3290589.3290599 |isbn=9781450366205 |s2cid=58534759

|last1=Munir|first1= Farzeen |last2= Azam |first2= Shoaib |last3= Hussain |first3= Muhammad Ishfaq |last4= Sheri |first4= Ahmed Muqeem |last5= Jeon|first5= Moongu

|title = Autonomous Vehicle: The Architecture Aspect of Self Driving Car |date=2018| publisher = Association for Computing Machinery| conference=Proceedings of the 2018 International Conference on Sensors, Signal and Image Processing

}} The first ADAS system was cruise control, which was invented in 1948 by Ralph Teetor.

Trials began in the 1950s. The first semi-autonomous car was developed in 1977, by Japan's Tsukuba Mechanical Engineering Laboratory.{{Cite journal

|title= Review on self-driving cars using neural network architectures

|last1=Srinivas |first1=Rao P|author2=Rohan Gudla |author3= Vijay Shankar Telidevulapalli |author4= Jayasree Sarada Kota |author5= Gayathri Mandha |journal= World Journal of Advanced Research and Reviews|date =2022 |volume=16 |number=2|pages= 736–746 |doi= 10.30574/wjarr.2022.16.2.1240 |url=https://www.researchgate.net/figure/Tsukuba-Mechanical-Engineering-Lab-Japan-1977-computerized-driverless-car-achieved-spe_fig2_365874855|doi-access=free }} It required specially marked streets that were interpreted by two cameras on the vehicle and an analog computer. The vehicle reached speeds of {{convert|30|km/h|abbr=on}} with the support of an elevated rail.{{Cite magazine|url=https://www.wired.com/2012/02/autonomous-vehicle-history/|title=Autonomous Cars Through The Ages|magazine=Wired|first1=Tom |last1=Vanderbilt|access-date=26 July 2018|date=6 February 2012}}{{cite web|url=http://www.computerhistory.org/atchm/where-to-a-history-of-autonomous-vehicles/|title=Where to? A History of Autonomous Vehicles|website=Computer History Museum|first=Marc|last=Weber|access-date=26 July 2018|date=8 May 2014}}

Carnegie Mellon University's Navlab{{cite web|url=https://www.cs.cmu.edu/afs/cs/project/alv/www/index.html|title=Carnegie Mellon|website=Navlab: The Carnegie Mellon University Navigation Laboratory|publisher=The Robotics Institute|access-date=20 December 2014}} and ALV{{Cite conference |last=Kanade |first=Takeo |title=Proceedings of the 1986 ACM fourteenth annual conference on Computer science - CSC '86 |date=February 1986 |chapter=Autonomous land vehicle project at CMU |url=http://dl.acm.org/citation.cfm?id=325197 |pages=71–80 |doi=10.1145/324634.325197|isbn=9780897911771 |s2cid=2308303 }}{{Cite conference |last=Wallace |first=Richard |year=1985 |title=First results in robot road-following |url=http://www.ijcai.org/Past%20Proceedings/IJCAI-85-VOL2/PDF/086.pdf |url-status=dead |conference=JCAI'85 Proceedings of the 9th International Joint Conference on Artificial Intelligence |archive-url=https://web.archive.org/web/20140806093746/http://ijcai.org/Past%20Proceedings/IJCAI-85-VOL2/PDF/086.pdf |archive-date=6 August 2014}} semi-autonomous projects launched in the 1980s, funded by the United States' Defense Advanced Research Projects Agency (DARPA) starting in 1984 and Mercedes-Benz and Bundeswehr University Munich's EUREKA Prometheus Project in 1987.{{cite web|url=http://www.idsia.ch/~juergen/robotcars.html|title=Prof. Schmidhuber's highlights of robot car history|last=Schmidhuber|first=Jürgen|year=2009|access-date=15 July 2011}} By 1985, ALV had reached {{convert|31|km/h|abbr=on}}, on two-lane roads. Obstacle avoidance came in 1986, and day and night off-road driving by 1987.{{Cite journal |last1=Turk |first1=M.A. |last2=Morgenthaler |first2=D.G. |last3=Gremban |first3=K.D. |last4=Marra |first4=M. |date=May 1988 |title=VITS-a vision system for automated land vehicle navigation |journal=IEEE Transactions on Pattern Analysis and Machine Intelligence|volume=10 |issue=3 |pages=342–361 |doi=10.1109/34.3899 |issn=0162-8828}} In 1995 Navlab 5 completed the first autonomous US coast-to-coast journey. Traveling from Pittsburgh, Pennsylvania and San Diego, California, 98.2% of the trip was autonomous. It completed the trip at an average speed of {{convert|63.8|mph|abbr=on}}.{{cite web|url=https://www.cmu.edu/news/stories/archives/2015/july/look-ma-no-hands.html|title=Look, Ma, No Hands |publisher= Carnegie Mellon University|access-date=2 March 2017}}{{cite web|url=https://www.cs.cmu.edu/~tjochem/nhaa/navlab5_details.html|title=Navlab 5 Details|website=cs.cmu.edu|access-date=2 March 2017}}{{cite web|url=http://www.roboticstrends.com/article/back_to_the_future_autonomous_driving_in_1995|title=Back to the Future: Autonomous Driving in 1995|work=Robotics Trends|last=Crowe|first=Steve|access-date=2 March 2017|date=3 April 2015|archive-date=29 December 2017|archive-url=https://web.archive.org/web/20171229081126/http://www.roboticstrends.com/article/back_to_the_future_autonomous_driving_in_1995|url-status=dead}}{{cite web|url=https://www.cs.cmu.edu/afs/cs/usr/tjochem/www/nhaa/Journal.html|title=NHAA Journal|website=cs.cmu.edu|access-date=5 March 2017}} Until the second DARPA Grand Challenge in 2005, automated vehicle research in the United States was primarily funded by DARPA, the US Army, and the US Navy, yielding incremental advances in speeds, driving competence, controls, and sensor systems.{{Cite book|publisher=National Research Council |date=2002|title=Technology Development for Army Unmanned Ground Vehicles|url=http://www.nap.edu/catalog/10592|doi=10.17226/10592|isbn=9780309086202}}

The US allocated US$650 million in 1991 for research on the National Automated Highway System,{{Cite web |title=The Automated Highway System: An Idea Whose Time Has Come {{!}} FHWA |url=https://highways.dot.gov/public-roads/summer-1994/automated-highway-system-idea-whose-time-has-come |access-date=2023-08-30 |website=highways.dot.gov}} which demonstrated automated driving, combining highway-embedded automation with vehicle technology, and cooperative networking between the vehicles and highway infrastructure. The programme concluded with a successful demonstration in 1997.{{Cite news|url=https://www.smithsonianmag.com/history/the-national-automated-highway-system-that-almost-was-63027245/|title=The National Automated Highway System That Almost Was|last=Novak|first=Matt|work=Smithsonian|access-date=8 June 2018}} Partly funded by the National Automated Highway System and DARPA, Navlab drove {{convert|4584|km|abbr=on}} across the US in 1995, {{convert|4501|km|abbr=on}} or 98% autonomously.{{Cite news|url=https://www.roboticsbusinessreview.com/slideshow/back_to_the_future_autonomous_driving_in_1995/|title=Back to the Future: Autonomous Driving in 1995|date=3 April 2015|work=Robotics Business Review|access-date=8 June 2018|archive-date=12 June 2018|archive-url=https://web.archive.org/web/20180612140201/https://www.roboticsbusinessreview.com/slideshow/back_to_the_future_autonomous_driving_in_1995/|url-status=dead}} In 2015, Delphi piloted a Delphi technology-based Audi, over {{convert|5472|km|abbr=on}} through 15 states, 99% autonomously.{{Cite magazine|url=https://www.wired.com/2015/04/delphi-autonomous-car-cross-country/|title=This Is Big: A Robo-Car Just Drove Across the Country|magazine=WIRED|access-date=8 June 2018}} In 2015, Nevada, Florida, California, Virginia, Michigan, and Washington DC allowed autonomous car testing on public roads.{{cite news|url=http://www.richmond.com/news/article_b1168b67-3b2b-5274-8914-8a3304f2e417.html|title=Self-driving cars to be tested on Virginia highways|last=Ramsey|first=John|date=1 June 2015|work=Richmond Times-Dispatch|access-date=4 June 2015}}

From 2016 to 2018, the European Commission funded development for connected and automated driving through Coordination Actions CARTRE and SCOUT programs.{{Cite book |last=Meyer |first=Gereon |date=2018 |chapter=European Roadmaps, Programs, and Projects for Innovation in Connected and Automated Road Transport |editor1= G. Meyer |editor2= S. Beiker |title= Road Vehicle Automation |series=Lecture Notes in Mobility |pages=27–39 |publisher= Springer |doi=10.1007/978-3-319-94896-6_3 |isbn=978-3-319-94895-9 |s2cid=169808153 }} The Strategic Transport Research and Innovation Agenda (STRIA) Roadmap for Connected and Automated Transport was published in 2019.{{Cite book|publisher=European Commission|date=2019|title=STRIA Roadmap Connected and Automated Transport: Road, Rail and Waterborne|url=https://trimis.ec.europa.eu/sites/default/files/roadmaps/stria_roadmap_2019-connected_and_automated_transport.pdf|access-date=10 November 2019|archive-date=16 October 2022|archive-url=https://web.archive.org/web/20221016163601/https://trimis.ec.europa.eu/sites/default/files/roadmaps/stria_roadmap_2019-connected_and_automated_transport.pdf|url-status=dead}}

In November 2017, Waymo announced testing of autonomous cars without a safety driver.{{cite web|url=https://www.theverge.com/2017/11/7/16615290/waymo-self-driving-safety-driver-chandler-autonomous|title=Waymo is first to put fully self-driving cars on US roads without a safety driver|website=The Verge|first=Andrew J.|last=Hawkins|date=7 November 2017|access-date=7 November 2017}} However, an employee was in the car to handle emergencies.{{cite web|url=https://waymo.com/apply/faq/|title=FAQ – Early Rider Program |website=Waymo|access-date=30 November 2018}}

In March 2018, Elaine Herzberg became the first reported pedestrian killed by a self-driving car, an Uber test vehicle with a human backup driver; prosecutors did not charge Uber, while the human driver was sentenced to probation.{{cite news|last1=Billeaud|first1=Jacques|last2=Snow|first2=Anita|title=The backup driver in the 1st death by a fully autonomous car pleads guilty to endangerment|url=https://apnews.com/article/autonomous-vehicle-death-uber-charge-backup-driver-1c711426a9cf020d3662c47c0dd64e35|work=Associated Press|date=2023-07-28|access-date=2024-09-01}}

In December 2018, Waymo was the first to commercialize a robotaxi service, in Phoenix, Arizona.{{Cite news|url=https://www.washingtonpost.com/local/trafficandcommuting/waymo-launches-nations-first-commercial-self-driving-taxi-service-in-arizona/2018/12/04/8a8cd58a-f7ba-11e8-8c9a-860ce2a8148f_story.html|title=Waymo launches nation's first commercial self-driving taxi service in Arizona|newspaper=The Washington Post|access-date=6 December 2018}} In October 2020, Waymo launched a robotaxi service in a (geofenced) part of the area.{{Cite news|date=21 January 2021|title=Waymo's Self-Driving Future Looks Real Now That the Hype Is Fading|work=Bloomberg.com|url=https://www.bloomberg.com/news/articles/2021-01-21/waymo-self-driving-taxis-are-coming-to-more-u-s-cities|access-date=5 March 2021}}{{cite web|last=Ackerman|first=Evan|date=4 March 2021|title=What Full Autonomy Means for the Waymo Driver|url=https://spectrum.ieee.org/full-autonomy-waymo-driver|access-date=8 March 2021|website=IEEE Spectrum: Technology, Engineering, and Science News}} The cars were monitored in real-time, and remote engineers intervened to handle exceptional conditions.{{cite web|last=Hawkins|first=Andrew J.|date=8 October 2020|title=Waymo will allow more people to ride in its fully driverless vehicles in Phoenix|url=https://www.theverge.com/2020/10/8/21507814/waymo-driverless-cars-allow-more-customers-phoenix|access-date=5 March 2021|website=The Verge}}

In March 2019, ahead of Roborace, Robocar set the Guinness World Record as the world's fastest autonomous car. Robocar reached 282.42 km/h (175.49 mph).{{cite web|date=17 October 2019|title=Robocar: Watch the world's fastest autonomous car reach its record-breaking 282 km/h|first = Connie |last = Suggitt |url=https://www.guinnessworldrecords.com/news/2019/10/robocar-watch-the-worlds-fastest-autonomous-car-reach-its-record-breaking-282-k|access-date=|website=Guinness World Records}}

In March 2021, Honda began leasing in Japan a limited edition of 100 Legend Hybrid EX sedans equipped with Level 3 "Traffic Jam Pilot" driving technology, which legally allowed drivers to take their eyes off the road when the car was travelling under {{Convert|30|kph|mph|abbr=out}}.{{cite web|url=https://global.honda/newsroom/news/2021/4210304eng-legend.html |date=4 March 2021 |title=Honda to Begin Sales of Legend with New Honda SENSING Elite |website=Honda |access-date=6 March 2021 }}{{cite news |url=https://mainichi.jp/english/articles/20210304/p2g/00m/0bu/109000c |date=4 March 2021 |title=Honda to start selling world's 1st level-3 autonomous car for $103K on Fri. |agency=Kyodo News |access-date=6 March 2021 |archive-date=5 March 2021 |archive-url=https://web.archive.org/web/20210305144526/https://mainichi.jp/english/articles/20210304/p2g/00m/0bu/109000c |url-status=dead }}{{cite web|url=https://www.mlit.go.jp/report/press/jidosha08_hh_003888.html |date=11 November 2020 |title=世界初！ 自動運転車（レベル3）の型式指定を行いました |trans-title=The world's first! approval of level-3 type designation for certification |language=ja |website=MLIT, Japan |access-date=6 March 2021 }}{{cite web|last=Beresford|first=Colin|date=4 March 2021|title=Honda Legend Sedan with Level 3 Autonomy Available for Lease in Japan|url=https://www.caranddriver.com/news/a35729591/honda-legend-level-3-autonomy-leases-japan/|access-date=6 March 2021|website=Car and Driver}}

In December 2020, Waymo became the first service provider to offer driverless taxi rides to the general public, in a part of Phoenix, Arizona. Nuro began autonomous commercial delivery operations in California in 2021.{{Cite news |date=24 December 2020 |title=Nuro set to be California's first driverless delivery service |publisher=BBC News |url=https://www.bbc.com/news/technology-55438969 |access-date=27 December 2020}} DeepRoute.ai launched robotaxi service in Shenzhen in July 2021.{{Cite web |last=Staff |first=The Robot Report |date=14 September 2021 |title=DeepRoute.ai closes $300M Series B funding round |url=https://www.therobotreport.com/deeproute-ai-closes-300-million-series-b-funding-round/ |website=The Robot Report}} In December 2021, Mercedes-Benz received approval for a Level 3 car.{{cite web |date=9 December 2021 |title=Mercedes-Benz self-driving car technology approved for use |url=https://www.fleetnews.co.uk/news/manufacturer-news/2021/12/09/mercedes-benz-self-driving-car-technology-approved-for-use |url-status=dead |archive-url=https://web.archive.org/web/20211209192401/https://www.fleetnews.co.uk/news/manufacturer-news/2021/12/09/mercedes-benz-self-driving-car-technology-approved-for-use |archive-date=9 December 2021 |access-date=10 December 2021 |website=Feet News}} In February 2022, Cruise became the second service provider to offer driverless taxi rides to the general public, in San Francisco. In December 2022, several manufacturers scaled back plans for self-driving technology, including Ford and Volkswagen.{{cite news |last= |first= |date=28 November 2022 |title=Slow Self-Driving Car Progress Tests Investors' Patience |work=The Wall Street Journal |url=https://www.wsj.com/articles/investors-are-losing-patience-with-slow-pace-of-driverless-cars-11669576382 |access-date=14 December 2022}} In 2023, Cruise suspended its robotaxi service.{{Cite news |last1=Shepardson |first1=David |last2=Klayman |first2=Ben |date=November 14, 2023 |title=GM's Cruise suspends supervised and manual car trips, expands probes |url=https://www.reuters.com/business/autos-transportation/gms-cruise-suspends-supervised-manual-car-trips-expands-probes-2023-11-15/}} Nuro was approved for Level 4 in Palo Alto in August, 2023.{{Cite web |title=nuro/status/1688965912165265408 |url=https://twitter.com/nuro/status/1688965912165265408 |access-date=2023-08-10 |website=Twitter |language=en}}

{{As of|2023|August}}, vehicles operating at Level 3 and above were an insignificant market factor;{{Citation needed|reason=This is a significant claim that needs evidence|date=February 2024}} as of early 2024, Honda leases a Level 3 car in Japan, and Mercedes sells two Level 3 cars in Germany, California and Nevada.{{Cite web |last=AUTOCRYPT |date=2023-01-13 |title=The State of Level 3 Autonomous Driving in 2023 |url=https://autocrypt.io/the-state-of-level-3-autonomous-driving-in-2023/ |access-date=2024-04-21 |website=AUTOCRYPT |language=en-US}}{{Cite web |last=Tucker |first=Sean |date=2024-01-09 |title=Self-Driving Cars: Everything You Need To Know |url=https://www.kbb.com/car-advice/self-driving-cars/ |access-date=2024-04-21 |website=Kelley Blue Book |language=en-US}} BMW also sells its Level 3 Personal Pilot in Germany.{{Cite web |last=Carlson |first=Cody |date=2023-11-14 |title=BMW’s $6,000 ’Personal Pilot’ Level 3 self-driving system is fast approaching |url=https://www.the-express.com/lifestyle/cars/118336/bmw-confirms-arrival-personal-pilot-self-driving |access-date=2025-05-14 |website=Daily Express US |language=en}}

Organizations such as SAE have proposed terminology standards. However, most terms have no standard definition and are employed variously by vendors and others. Proposals to adopt aviation automation terminology for cars have not prevailed.{{cite journal |last=Umar Zakir Abdul |first=Hamid |display-authors=etal |date=2021 |title=Adopting Aviation Safety Knowledge into the Discussions of Safe Implementation of Connected and Autonomous Road Vehicles |url=https://www.researchgate.net/publication/350669647 |journal=SAE Technical Papers (SAE WCX Digital Summit) |issue=2021–01–0074 |access-date=12 April 2021}}

Names such as AutonoDrive, PilotAssist, Full-Self Driving or DrivePilot are used even though the products offer an assortment of features that may not match the names.{{cite web |last=Morris |first=David |date=8 November 2020 |title=What's in a name? For Tesla's Full Self Driving, it may be danger |url=https://fortune.com/2020/11/08/tesla-full-self-driving-autonomous-vehicle-safety/ |access-date=8 March 2021 |website=Fortune}} Despite offering a system dubbed "Full Self-Driving", Tesla stated that its system did not autonomously handle all driving tasks.{{cite news |last=Boudette |first=Neal E. |date=23 March 2021 |title=Tesla's Autopilot Technology Faces Fresh Scrutiny |url=https://www.nytimes.com/2021/03/23/business/teslas-autopilot-safety-investigations.html |url-access=limited |archive-url=https://ghostarchive.org/archive/20211228/https://www.nytimes.com/2021/03/23/business/teslas-autopilot-safety-investigations.html |archive-date=28 December 2021 |access-date=15 June 2021 |newspaper=The New York Times}}

Tesla vehicles remain at Level 2 automation and are therefore not "fully self-driving" and require active driver supervision.

{{cbignore}} In the United Kingdom, a fully self-driving car is defined as a car so registered, rather than one that supports a specific feature set.{{cite news |last=Cellan-Jones |first=Rory |date=12 June 2018 |title=Insurers warning on "autonomous" cars |url=https://www.bbc.com/news/technology-44439523 |publisher=BBC News}} The Association of British Insurers claimed that the usage of the word autonomous in marketing was dangerous because car ads make motorists think "autonomous" and "autopilot" imply that the driver can rely on the car to control itself, even though they do not.

SAE identified 6 levels for driving automation from level 0 to level 5.{{Cite journal |date=2021-07-14 |title=An Integrated Approach for Predicting Consumer Acceptance of Self-Driving Vehicles in the United States |url=https://articlegateway.com/index.php/JMDC/article/view/4330 |journal=Journal of Marketing Development and Competitiveness |language=en |volume=15 |issue=2 |doi=10.33423/jmdc.v15i2.4330 |issn=2155-2843}} An ADS is an SAE J3016 level 3 or higher system.

{{Main|Advanced driver-assistance system}}

An ADAS is a system that automates specific driving features, such as Forward Collision Warning (FCW), Automatic Emergency Braking (AEB), Lane Departure Warning (LDW), Lane Keeping Assistance (LKA) or Blind Spot Warning (BSW).{{Cite journal |last1=Aleksa |first1=Michael |last2=Schaub |first2=Andrea |last3=Erdelean |first3=Isabela |last4=Wittmann |first4=Stephan |last5=Soteropoulos |first5=Aggelos |last6=Fürdös |first6=Alexander |date=2024-06-27 |title=Impact analysis of Advanced Driver Assistance Systems (ADAS) regarding road safety – computing reduction potentials |journal=European Transport Research Review |volume=16 |issue=1 |pages=39 |doi=10.1186/s12544-024-00654-0 |doi-access=free |bibcode=2024ETRR...16...39A |issn=1866-8887}} An ADAS requires a human driver to handle tasks that the ADAS does not support.

Autonomy implies that an automation system is under the control of the vehicle rather than a driver. Automation is function-specific, handling issues such as speed control, but leaves broader decision-making to the driver.{{cite journal |last1=Antsaklis |first1=Panos J. |last2=Passino |first2=Kevin M. |last3=Wang |first3=S.J. |year=1991 |title=An Introduction to Autonomous Control Systems |url=http://neuron-ai.tuke.sk/hudecm/PDF_PAPERS/Intro-Aut-Control.pdf |url-status=dead |journal=IEEE Control Systems Magazine |volume=11 |issue=4 |pages=5–13 |citeseerx=10.1.1.840.976 |doi=10.1109/37.88585 |archive-url=https://web.archive.org/web/20170516202116/http://neuron-ai.tuke.sk/hudecm/PDF_PAPERS/Intro-Aut-Control.pdf |archive-date=16 May 2017 |access-date=21 January 2019}}

Euro NCAP defined autonomous as "the system acts independently of the driver to avoid or mitigate the accident".{{cite web|url=https://www.euroncap.com/en/vehicle-safety/the-rewards-explained/autonomous-emergency-braking/|title=Autonomous Emergency Braking – Euro NCAP|website=euroncap.com}}

In Europe, the words automated and autonomous can be used together. For instance, Regulation (EU) 2019/2144 supplied:Regulation (EU) 2019/2144

• "automated vehicle" means a vehicle that can move without continuous driver supervision, but that driver intervention is still expected or required in the operational design domains (ODD);
• "fully automated vehicle" means a vehicle that can move entirely without driver supervision;

A remote driver is a driver that operates a vehicle at a distance, using a video and data connection.{{Cite journal |last1=Yu |first1=Yang |last2=Lee |first2=Sanghwan |date=16 June 2022 |title=Remote Driving Control With Real-Time Video Streaming Over Wireless Networks: Design and Evaluation |journal=IEEE Access|volume=10 |pages=64920–64932 |doi=10.1109/ACCESS.2022.3183758 |bibcode=2022IEEEA..1064920Y |doi-access=free }}

According to SAE J3016, {{blockquote|Some driving automation systems may indeed be autonomous if they perform all of their functions independently and self-sufficiently, but if they depend on communication and/or cooperation with outside entities, they should be considered cooperative rather than autonomous.}}

{{Excerpt|Operational design domain|paragraphs=1-2}}Vendors have taken a variety of approaches to the self-driving problem. Tesla's approach is to allow their "full self-driving" (FSD) system to be used in all ODDs as a Level 2 (hands/on, eyes/on) ADAS.{{Cite news |last=Lambert |first=Fred |date=March 8, 2023 |title=Tesla pushes new Full Self-Driving Beta v11 update as it slowly expands rollout |url=https://electrek.co/2023/03/08/tesla-full-self-driving-beta-v11-update-slowly-expands-rollout/ |work=electrek.co}} Waymo picked specific ODDs (city streets in Phoenix and San Francisco) for their Level 5 robotaxi service.{{Cite web |last=Ohnsman |first=Alan |title=Waymo's Robotaxis Are Hitting The Highway, A First For Self-Driving Cars |url=https://www.forbes.com/sites/alanohnsman/2024/01/08/waymos-robotaxis-are-hitting-the-highway-a-first-for-self-driving-cars/ |access-date=2024-02-13 |website=Forbes |language=en}} Mercedes Benz offers Level 3 service in Las Vegas in highway traffic jams at speeds up to {{Convert|40|mph}}.{{Cite web |last=Golson |first=Daniel |date=2023-09-27 |title=We put our blind faith in Mercedes-Benz's first-of-its-kind autonomous Drive Pilot feature |url=https://www.theverge.com/2023/9/27/23892154/mercedes-benz-drive-pilot-autonomous-level-3-test |access-date=2024-02-13 |website=The Verge |language=en}} Mobileye's SuperVision system offers hands-off/eyes-on driving on all road types at speeds up to {{Convert|130|kph}}.{{Cite web |title=Mobileye SuperVision™ {{!}} The Bridge from ADAS to Consumer AVs |url=https://www.mobileye.com/solutions/super-vision/ |access-date=2024-02-14 |website=Mobileye |language=en}} GM's hands-free Super Cruise operates on specific roads in specific conditions, stopping or returning control to the driver when ODD changes. In 2024 the company announced plans to expand road coverage from 400,000 miles to 750,000 miles.{{Cite news |last=HUNT |first=RHIAN |date=February 15, 2024 |title=GM Adding 350,000 Miles Of Super Cruise Road Coverage |url=https://gmauthority.com/blog/2024/02/gm-adding-350000-miles-of-super-cruise-road-coverage/ |work=GM Authority}} Ford's BlueCruise hands-off system operates on 130,000 miles of US divided highways.{{Cite web |last=Wardlaw |first=Christian |date=April 20, 2021 |title=What is Ford BlueCruise, and How Does It Work? |url=https://www.jdpower.com/cars/shopping-guides/what-is-ford-bluecruise-and-how-does-it-work |website=jdpower.com}}

The Union of Concerned Scientists defined self-driving as "cars or trucks in which human drivers are never required to take control to safely operate the vehicle. Also known as autonomous or 'driverless' cars, they combine sensors and software to control, navigate, and drive the vehicle."{{cite web|url=https://www.ucsusa.org/clean-vehicles/how-self-driving-cars-work|title=Self-Driving Cars Explained|website=Union of Concerned Scientists}}

The British Automated and Electric Vehicles Act 2018 law defines a vehicle as "driving itself" if the vehicle is "not being controlled, and does not need to be monitored, by an individual".{{cite web|title=Automated and Electric Vehicles Act 2018 becomes law|url=https://www.penningtonslaw.com/news-publications/latest-news/2018/automated-and-electric-vehicles-act-2018-becomes-law|access-date=24 March 2021|website=penningtonslaw.com}}

Another British government definition stated, "Self-driving vehicles are vehicles that can safely and lawfully drive themselves".{{cite web|url=https://www.gov.uk/guidance/self-driving-vehicles-listed-for-use-in-great-britain |title=Self-driving vehicles listed for use in Great Britain |publisher=GOV.UK |date=20 April 2022 |accessdate=19 July 2022}}

In British English, the word automated alone has several meanings, such as in the sentence: "Thatcham also found that the automated lane keeping systems could only meet two out of the twelve principles required to guarantee safety, going on to say they cannot, therefore, be classed as 'automated driving', preferring 'assisted driving'".{{cite web |last1=Hancocks |first1=Simon |date=26 October 2020 |title=The ABI and Thatcham warn against automated driving plans |url=https://www.visordown.com/news/industry/abi-and-thatcham-warn-against-automated-driving-plans |website=Visordown}} The first occurrence of the "automated" word refers to an Unece automated system, while the second refers to the British legal definition of an automated vehicle. British law interprets the meaning of "automated vehicle" based on the interpretation section related to a vehicle "driving itself" and an insured vehicle.Automated and Electric Vehicles Act 2018

In November 2023 the British Government introduced the Automated Vehicles Bill. It proposed definitions for related terms:{{Cite web|url=https://bills.parliament.uk/publications/52908/documents/3984|title=Automated Vehicle Bill}}

• Self-driving: "A vehicle “satisfies the self-driving test” if it is designed or adapted with the intention that a feature of the vehicle will allow it to travel autonomously, and it is capable of doing so, by means of that feature, safely and legally."
• Autonomy: A vehicle travels "autonomously" if it is controlled by the vehicle, and neither the vehicle nor its surroundings are monitored by a person who can intervene.
• Control: control of vehicle motion.
• Safe: a vehicle that conforms to an acceptably safe standard.
• Legal: a vehicle that offers an acceptably low risk of committing a traffic infraction.

File:Tesla Autopilot Engaged in Model X.jpg is classified as an SAE Level 2 system.{{cite web|url=https://www.tesla.com/support/autopilot |title=Support – Autopilot |work=Tesla |date=13 February 2019 |access-date=6 September 2019 |url-status=live |archive-url=https://web.archive.org/web/20190410153216/https://www.tesla.com/support/autopilot |archive-date=10 April 2019}}{{cite web|url=https://www.caranddriver.com/news/a35785277/tesla-fsd-california-self-driving/ |author=Roberto Baldwin |date=9 March 2021 |title=Tesla Tells California DMV that FSD Is Not Capable of Autonomous Driving |website=Car and Driver }}]]

A six-level classification system – ranging from fully manual to fully automated – was published in 2014 by SAE International as J3016, Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems; the details are revised occasionally.{{cite web |author=SAE International |date=30 April 2021 |title=Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles (SAE J3016) |url=https://www.sae.org/standards/content/j3016_202104/ |archive-url=https://web.archive.org/web/20211220101755/https://www.sae.org/standards/content/j3016_202104/ |archive-date=20 December 2021 |access-date=25 December 2021}} This classification is based on the role of the driver, rather than the vehicle's capabilities, although these are related. After SAE updated its classification in 2016, (J3016_201609),SAE International the National Highway Traffic Safety Administration (NHTSA) adopted the SAE standard.{{cite web|date=September 2016 |title=Federal Automated Vehicles Policy |url=https://www.nhtsa.gov/sites/nhtsa.gov/files/federal_automated_vehicles_policy.pdf |page=9 |website=NHTSA, U.S. |access-date=1 December 2021}}{{cite web |date=1 February 2018 |title=JASO TP 18004: 自動車用運転自動化システムのレベル分類及び定義 |trans-title=JASO TP 18004: Taxonomy and Definitions for Terms Related to Driving Automation Systems |url=https://www.jsae.or.jp/08std/data/DrivingAutomation/jaso_tp18004-18.pdf |website=JASO, Japan |access-date=1 December 2021 |archive-date=1 December 2021 |archive-url=https://web.archive.org/web/20211201121753/https://www.jsae.or.jp/08std/data/DrivingAutomation/jaso_tp18004-18.pdf |url-status=dead }} The classification is a topic of debate, with various revisions proposed.{{Citation |last1=Steckhan |first1=Lorenz |title=Beyond SAE J3016: New Design Spaces for Human-Centered Driving Automation |date=2022 |url=https://link.springer.com/10.1007/978-3-031-04987-3_28 |work=HCI in Mobility, Transport, and Automotive Systems |volume=13335 |pages=416–434 |editor-last=Krömker |editor-first=Heidi |place=Cham |publisher=Springer International Publishing |language=en |doi=10.1007/978-3-031-04987-3_28 |isbn=978-3-031-04986-6 |access-date=2023-01-24 |last2=Spiessl |first2=Wolfgang |last3=Quetschlich |first3=Nils |last4=Bengler |first4=Klaus|series=Lecture Notes in Computer Science }}{{Cite journal |last1=Inagaki |first1=Toshiyuki |last2=Sheridan |first2=Thomas B. |date=November 2019 |title=A critique of the SAE conditional driving automation definition, and analyses of options for improvement |url=http://link.springer.com/10.1007/s10111-018-0471-5 |journal=Cognition, Technology & Work |language=en |volume=21 |issue=4 |pages=569–578 |doi=10.1007/s10111-018-0471-5 |s2cid=254144879 |issn=1435-5558|hdl=1721.1/116231 |hdl-access=free }}

A "driving mode", aka driving scenario, combines an ODD with matched driving requirements (e.g., expressway merging, traffic jam).{{cite web|url=https://cdn.oemoffhighway.com/files/base/acbm/ooh/document/2016/03/automated_driving.pdf |title=Automated Driving – Levels of Driving Automation are Defined in New SAE International Standard J3016 |work=SAE International |year=2014 |archive-url=https://web.archive.org/web/20180701034327/https://cdn.oemoffhighway.com/files/base/acbm/ooh/document/2016/03/automated_driving.pdf |archive-date=1 July 2018 |url-status=live}} Cars may switch levels in accord with the driving mode.

Above Level 1, level differences are related to how responsibility for safe movement is divided/shared between ADAS and driver rather than specific driving features.

SAE Automation Levels have been criticized{{by whom|date=September 2024}} for their technological focus. It has been argued that the structure of the levels suggests that automation increases linearly and that more automation is better, which may not be the case.{{Cite journal |last1=Stayton |first1=E. |last2=Stilgoe |first2=J. |date=September 2020 |title=It's Time to Rethink Levels of Automation for Self-Driving Vehicles [Opinion] |journal=IEEE Technology and Society Magazine |volume=39 |issue=3 |pages=13–19 |doi=10.1109/MTS.2020.3012315 |issn=1937-416X |doi-access=free}} SAE Levels also do not account for changes that may be required to infrastructure{{cite web |date=6 July 2020 |title=Preparing the UK's motorways for self-driving vehicles: New £1m research project announced in partnership with Highways England |url=https://www.lboro.ac.uk/news-events/news/2020/july/preparing-motorways-for-autonomous-vehicles/ |access-date=13 April 2021 |website=Loughborough University}} and road user behavior.{{Cite journal |last1=Cavoli |first1=Clemence |last2=Phillips |first2=Brian |year=2017 |others=Tom Cohen |title=Social and behavioural questions associated with Automated Vehicles A Literature Review. |url=https://www.ucl.ac.uk/transport/sites/transport/files/social-and-behavioural-literature-review.pdf |journal=UCL Transport Institute}}{{Cite journal |last1=Parkin |first1=John |last2=Clark |first2=Benjamin |last3=Clayton |first3=William |last4=Ricci |first4=Miriam |last5=Parkhurst |first5=Graham |date=27 October 2017 |title=Autonomous vehicle interactions in the urban street environment: a research agenda |journal=Proceedings of the Institution of Civil Engineers - Municipal Engineer |volume=171 |issue=1 |pages=15–25 |doi=10.1680/jmuen.16.00062 |issn=0965-0903 |doi-access=free}}

thumb

Mobileye CEO Amnon Shashua and CTO Shai Shalev-Shwartz proposed an alternative taxonomy for autonomous driving systems, claiming that a more consumer-friendly approach was needed. Its categories reflect the amount of driver engagement that is required.{{Cite web |last=Hagman |first=Brian |date=2023-02-16 |title=Mobileye Proposes New Taxonomy and Requirements for Consumer Autonomous Vehicles to Ensure Clarity, Safety, and Scalability |url=https://selfdrivenews.com/mobileye-proposes-new-taxonomy-and-requirements-for-consumer-autonomous-vehicles-to-ensure-clarity-safety-and-scalability/ |access-date=2024-02-04 |website=Self Drive News |language=en-US}}{{Cite web |last1=Shashua |first1=Amnon |last2=Shalev-Shwartz |first2=Shai |date=February 5, 2023 |title=Defining a New Taxonomy for Consumer Autonomous Vehicles |work=Mobileye |url=https://www.mobileye.com/opinion/defining-a-new-taxonomy-for-consumer-autonomous-vehicles/}} Some vehicle makers have informally adopted some of the terminology involved, while not formally committing to it.{{Cite web |title=Ford BlueCruise {{!}} Consumer Reports Top-Rated Active Driving Assistance System {{!}} Ford.com |url=https://www.ford.com/technology/bluecruise/ |access-date=2024-02-08 |website=Ford Motor Company |language=en-US}}{{Cite web |title=Hands-Free, Eyes On |url=https://www.gm.com/commitments/av-safe-deployment |access-date=2024-02-08 |website=www.gm.com}}{{Cite web |title=Level 2 of autonomous driving - "EYES ON / HANDS OFF" |url=https://www.valeo.com/en/level-2-autonomous-driving-eyes-on-hands-off/ |access-date=2024-02-08 |website=Valeo |language=en-US}}{{Cite news |last=Dow |first=Jameson |date=September 27, 2023 |title=Hands-off with the first true hands-free car in the US, and it's not Tesla |url=https://electrek.co/2023/09/27/hands-off-with-the-first-true-hands-free-car-in-the-us-and-its-not-tesla/ |access-date=February 8, 2024 |work=Electrek.co}}

The first level, hands-on/eyes-on, implies that the driver is fully engaged in operating the vehicle, but is supervised by the system, which intervenes according to the features it supports (e.g., adaptive cruise control, automatic emergency braking). The driver is entirely responsible, with hands on the wheel and eyes on the road.

Eyes-on/hands-off allows the driver to let go of the wheel. The system drives, the driver monitors, and remains prepared to resume control as needed.

Eyes-off/hands-off means that the driver can stop monitoring the system, leaving the system in full control. Eyes-off requires that no errors be reproducible (not triggered by exotic transitory conditions) or frequent, that speeds are contextually appropriate (e.g., 80 mph on limited-access roads), and that the system handles typical maneuvers (e.g., getting cut off by another vehicle). The automation level could vary according to the road (e.g., eyes-off on freeways, eyes-on on side streets).

The highest level does not require a human driver in the car: monitoring is done either remotely (telepresence) or not at all.

A critical requirement for the higher two levels is that the vehicle be able to conduct a Minimum Risk Maneuver and stop safely out of traffic without driver intervention.

{{Main|Vehicular automation}}

The perception system processes visual and audio data from outside and inside the car to create a local model of the vehicle, the road, traffic, traffic controls and other observable objects, and their relative motion. The control system then takes actions to move the vehicle, considering the local model, road map, and driving regulations.{{Cite journal |last1=Hu |first1=J. |last2=Bhowmick |first2=P. |last3=Jang |first3=I. |last4=Arvin |first4=F. |last5=Lanzon |first5=A. |year=2021 |title=A Decentralized Cluster Formation Containment Framework for Multirobot Systems |url=https://ieeexplore.ieee.org/document/9423979 |journal=IEEE Transactions on Robotics |volume=37 |issue=6 |pages=1936–1955 |doi=10.1109/TRO.2021.3071615 |access-date=2024-02-02 |via=ieeexplore.ieee.org}}{{cite web |year=2015 |title=European Roadmap Smart Systems for Automated Driving |url=http://www.smart-systems-integration.org/public/documents/publications/EPoSS%20Roadmap_Smart%20Systems%20for%20Automated%20Driving_2015_V1.pdf |url-status=dead |archive-url=https://web.archive.org/web/20150212024339/http://www.smart-systems-integration.org/public/documents/publications/EPoSS%20Roadmap_Smart%20Systems%20for%20Automated%20Driving_2015_V1.pdf |archive-date=12 February 2015 |work=EPoSS}}{{Cite journal |last1=Lim |first1=THazel Si Min |last2=Taeihagh |first2=Araz |year=2019 |title=Algorithmic Decision-Making in AVs: Understanding Ethical and Technical Concerns for Smart Cities |journal=Sustainability |volume=11 |issue=20 |page=5791 |arxiv=1910.13122 |bibcode=2019arXiv191013122L |doi=10.3390/su11205791 |s2cid=204951009 |doi-access=free}}{{Cite journal |last=Matzliach |first=Barouch |year=2022 |title=Detection of Static and Mobile Targets by an Autonomous Agent with Deep Q-Learning Abilities |journal=Entropy |publisher=Entropy, 2022, 24, 1168 |volume=24 |issue=8 |page=1168 |bibcode=2022Entrp..24.1168M |doi=10.3390/e24081168 |pmc=9407070 |pmid=36010832 |doi-access=free}}

Several classifications have been proposed to describe ADAS technology. One proposal is to adopt these categories: navigation, path planning, perception, and car control.{{Cite journal|last1=Zhao|first1=Jianfeng|last2=Liang|first2=Bodong|last3=Chen|first3=Qiuxia|date=2 January 2018|title=The key technology toward the self-driving car|journal=International Journal of Intelligent Unmanned Systems|volume=6|issue=1|pages=2–20|doi=10.1108/IJIUS-08-2017-0008|issn=2049-6427|doi-access=free}}

{{Main|Hybrid navigation}}

Navigation involves the use of maps to define a path between origin and destination. Hybrid navigation is the use of multiple navigation systems. Some systems use basic maps, relying on perception to deal with anomalies. Such a map understands which roads lead to which others, whether a road is a freeway, a highway, are one-way, etc. Other systems require highly detailed maps, including lane maps, obstacles, traffic controls, etc.

ACs need to be able to perceive the world around them. Supporting technologies include combinations of cameras, LiDAR, radar, audio, and ultrasound,{{cite web |date=20 February 2020 |title=2020 Autonomous Vehicle Technology Report |url=https://www.wevolver.com/article/2020.autonomous.vehicle.technology.report |access-date=11 April 2022 |website=Wevolver}} GPS, and inertial measurement.{{cite arXiv|last1=Huval|first1=Brody|last2=Wang|first2=Tao|last3=Tandon|first3=Sameep|last4=Kiske|first4=Jeff|last5=Song|first5=Will|last6=Pazhayampallil|first6=Joel|title=An Empirical Evaluation of Deep Learning on Highway Driving|eprint=1504.01716|class=cs.RO|year=2015}}{{Cite journal| title=An Introduction to Inertial and Visual Sensing |first1=Peter |last1=Corke |first2=Jorge |last2=Lobo |first3=Jorge |last3=Dias |date= 1 June 2007 | volume= 26|issue=6 |journal=The International Journal of Robotics Research|doi=10.1177/0278364907079279 |pages=519–535|citeseerx=10.1.1.93.5523|s2cid=206499861 }}{{Cite journal |last1=Ahangar |first1=M. Nadeem |last2=Ahmed |first2=Qasim Z. |last3=Khan |first3=Fahd A. |last4=Hafeez |first4=Maryam |date=January 2021 |title=A Survey of Autonomous Vehicles: Enabling Communication Technologies and Challenges |journal=Sensors |language=en |volume=21 |issue=3 |pages=706 |doi=10.3390/s21030706 |doi-access=free |pmid=33494191 |pmc=7864337 |bibcode=2021Senso..21..706A |issn=1424-8220}} Deep neural networks are used to analyse inputs from these sensors to detect and identify objects and their trajectories.{{cite conference |last1=Li |first1=Li |last2=Shum |first2=Hubert P. H. |last3=Breckon |first3=Toby P. |title=2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) |chapter=Less is More: Reducing Task and Model Complexity for 3D Point Cloud Semantic Segmentation |date=2023 |pages=9361–9371 |doi=10.1109/CVPR52729.2023.00903 |publisher=IEEE/CVF |arxiv=2303.11203 |isbn=979-8-3503-0129-8 }} Some systems use Bayesian simultaneous localization and mapping (SLAM) algorithms. Another technique is detection and tracking of other moving objects (DATMO), used to handle potential obstacles.{{Cite journal|journal= IEEE Robotics & Automation Magazine |title=Simultaneous localization and mapping | volume= 13|issue= 2 |pages= 99–110 |date=5 June 2006 |issn=1070-9932 |doi=10.1109/mra.2006.1638022 | last1 = Durrant-Whyte | first1 = H. | last2 = Bailey | first2 = T.|citeseerx=10.1.1.135.9810 |s2cid=8061430 }}{{Cite web|url=https://www.researchgate.net/publication/329873195|title=A Brief Survey on SLAM Methods in Autonomous Vehicle}} Other systems use roadside real-time locating system (RTLS) technologies to aid localization. Tesla's "vision only" system uses eight cameras, without LIDAR or radar, to create its bird's-eye view of the environment.{{Cite web |title=Tesla Vision Update: Replacing Ultrasonic Sensors with Tesla Vision {{!}} Tesla Support |url=https://www.tesla.com/support/transitioning-tesla-vision |access-date=2023-08-31 |website=Tesla |language=en}}

Path planning finds a sequence of segments that a vehicle can use to move from origin to destination. Techniques used for path planning include graph-based search and variational-based optimization techniques. Graph-based techniques can make harder decisions such as how to pass another vehicle/obstacle. Variational-based optimization techniques require more stringent restrictions on the vehicle's path to prevent collisions.{{Cite web |last1=Althoff |first1=Matthias |last2=Sontges |first2=Sebastian |date=June 2017 |title=Computing possible driving corridors for automated vehicles |url=https://www.researchgate.net/publication/318805881}} The large scale path of the vehicle can be determined by using a voronoi diagram, an occupancy grid mapping, or a driving corridor algorithm. The latter allows the vehicle to locate and drive within open space that is bounded by lanes or barriers.{{cite web |author=Deepshikha Shukla |date=16 August 2019 |title=Design Considerations For Autonomous Vehicles |url=https://www.electronicsforu.com/market-verticals/automotive/design-considerations-autonomous-vehicles |access-date=18 April 2018}}

Maps are necessary for navigation. Map sophistication varies from simple graphs that show which roads connect to each other, with details such as one-way vs two-way, to those that are highly detailed, with information about lanes, traffic controls, roadworks, and more. Researchers at the MITComputer Science and Artificial Intelligence Laboratory (CSAIL) developed a system called MapLite, which allows self-driving cars to drive with simple maps. The system combines the GPS position of the vehicle, a "sparse topological map" such as OpenStreetMap (which has only 2D road features), with sensors that observe road conditions.{{cite web |last1=Connor-Simons |first1=Adam |last2=Gordon |first2=Rachel |date=7 May 2018 |title=Self-driving cars for country roads: Today's automated vehicles require hand-labeled 3-D maps, but CSAIL's MapLite system enables navigation with just GPS and sensors. |url=https://news.mit.edu/2018/self-driving-cars-for-country-roads-mit-csail-0507 |access-date=14 May 2018}} One issue with highly-detailed maps is updating them as the world changes. Vehicles that can operate with less-detailed maps do not require frequent updates or geo-fencing.

Sensors are necessary for the vehicle to properly respond to the driving environment. Sensor types include cameras, LiDAR, ultrasound, and radar. Control systems typically combine data from multiple sensors.{{cite web |date=14 December 2017 |title=How Self-Driving Cars Work |work=Medium |url=https://medium.com/udacity/how-self-driving-cars-work-f77c49dca47e |access-date=18 April 2018 |last1=Silver |first1=David }} Multiple sensors can provide a more complete view of the surroundings and can be used to cross-check each other to correct errors.{{Cite journal |last1=Yeong |first1=De Jong |last2=Velasco-Hernandez |first2=Gustavo |last3=Barry |first3=John |last4=Walsh |first4=Joseph |date=2021 |title=Sensor and Sensor Fusion Technology in Autonomous Vehicles: A Review |journal=Sensors |language=en |volume=21 |issue=6 |pages=2140 |bibcode=2021Senso..21.2140Y |doi=10.3390/s21062140 |issn=1424-8220 |pmc=8003231 |pmid=33803889 |doi-access=free}} For example, radar can image a scene in, e.g., a nighttime snowstorm, that defeats cameras and LiDAR, albeit at reduced precision. After experimenting with radar and ultrasound, Tesla adopted a vision-only approach, asserting that humans drive using only vision, and that cars should be able to do the same, while citing the lower cost of cameras versus other sensor types.{{Cite web |last=Tara |first=Roopinder |date=October 2, 2023 |title=Now Revealed: Why Teslas Have Only Camera-Based Vision |url=https://www.engineering.com/story/now-revealed-why-teslas-have-only-camera-based-vision |access-date=2024-02-13 |website=Engineering.com}} By contrast, Waymo makes use of the higher resolution of LiDAR sensors and cites the declining cost of that technology.{{Cite web |date=September 21, 2022 |title=Informing smarter lidar solutions for the future |url=https://waymo.com/blog/2022/09/informing-smarter-lidar-solutions- |access-date=2024-02-13 |website=Waymo |language=en}}

{{main|Drive by wire}}

Drive by wire is the use of electrical or electro-mechanical systems for performing vehicle functions such as steering or speed control that are traditionally achieved by mechanical linkages.

{{main|Driver monitoring system}}

Driver monitoring is used to assess the driver's attention and alertness. Techniques in use include eye monitoring, and requiring the driver to maintain torque on the steering wheel.{{cite web|author=Alain Dunoyer |title=Why driver monitoring will be critical to next-generation autonomous vehicles |url=https://www.sbdautomotive.com/en/news-insight-driver-monitoring |website=SBD Automotive |date=27 January 2022 |access-date=13 May 2022 }} It attempts to understand driver status and identify dangerous driving behaviors.{{Cite web |title=How road rage really affects your driving -- and the self-driving cars of the future |url=https://www.sciencedaily.com/releases/2023/04/230411150507.htm |access-date=2023-11-25 |website=ScienceDaily |language=en}}

{{main|Vehicular communication systems}}

Vehicles can potentially benefit from communicating with others to share information about traffic, road obstacles, to receive map and software updates, etc.{{cite web|date=11 April 2019 |author=Mike Beevor |title=Driving autonomous vehicles forward with intelligent infrastructure |url=https://www.smartcitiesworld.net/opinions/opinions/driving-autonomous-vehicles-forward-with-intelligent-infrastructure

|website=Smart Cities World |access-date=27 April 2022 }}{{cite web |date=October 2010 |title=Frequency of Target Crashes for IntelliDrive Safety Systems |url=http://www.nhtsa.gov/DOT/NHTSA/NVS/Crash%20Avoidance/Technical%20Publications/2010/811381.pdf |website=NHTSA |access-date=27 April 2022 |archive-date=5 April 2021 |archive-url=https://web.archive.org/web/20210405114215/https://www.nhtsa.gov/DOT/NHTSA/NVS/Crash%20Avoidance/Technical%20Publications/2010/811381.pdf |url-status=dead }}

ISO/TC 22 specifies in-vehicle transport information and control systems,{{cite web |date=2 November 2016 |title=ISO/TC 22: Road vehicles |url=https://www.iso.org/committee/46706.html |access-date=11 May 2022 |website=ISO}} while ISO/TC 204 specifies information, communication and control systems in surface transport.{{cite web |date=7 July 2021 |title=ISO/TC 204: Intelligent transport systems |url=https://www.iso.org/committee/54706.html |access-date=11 May 2022 |website=ISO}} International standards have been developed for ADAS functions, connectivity, human interaction, in-vehicle systems, management/engineering, dynamic map and positioning, privacy and security.{{cite web |date=18 June 2019 |title=Standards Collection |url=https://www.connectedautomateddriving.eu/standards/standards-collection/ |access-date=23 November 2021 |website=connected automated driving.eu}}

Rather than communicating among vehicles, they can communicate with road-based systems to receive similar information.

{{See also|Over-the-air programming}}

Software controls the vehicle, and can provide entertainment and other services. Over-the-air updates can deliver bug fixes and additional features over the internet. Software updates are one way to accomplish recalls that in the past required a visit to a service center. In March 2021, the UNECE regulation on software update and software update management systems was published.{{cite web|date=4 March 2021 |title=UN Regulation No. 156 – Software update and software update management system |url=https://unece.org/transport/documents/2021/03/standards/un-regulation-no-156-software-update-and-software-update |website=UNECE |access-date=20 March 2022 }}

A safety model is software that attempts to formalize rules that ensure that ACs operate safely.{{cite arXiv|year=2017 |first1= Shai |last1=Shalev-Shwartz |first2= Shaked |last2=Shammah |first3= Amnon |last3=Shashua |title=On a Formal Model of Safe and Scalable Self-driving Cars |class= cs.RO |eprint= 1708.06374 }}

IEEE is attempting to forge a standard for safety models as "IEEE P2846: A Formal Model for Safety Considerations in Automated Vehicle Decision Making".{{cite web|title=WG: VT/ITS/AV Decision Making |url=https://sagroups.ieee.org/2846/ |website=IEEE Standards Association |access-date= 18 July 2022}} In 2022, a research group at National Institute of Informatics (NII, Japan) enhanced Mobileye's Reliable Safety System as "Goal-Aware RSS" to enable RSS rules to deal with complex scenarios via program logic.{{Cite journal|first1=Ichiro |last1= Hasuo |first2= Clovis |last2=Eberhart |first3=James |last3=Haydon |first4=Jérémy |last4=Dubut |first5=Brandon |last5=Bohrer |first6=Tsutomu |last6=Kobayashi |first7= Sasinee |last7=Pruekprasert |first8= Xiao-Yi |last8=Zhang |first9= Erik |last9=Andre Pallas |first10=Akihisa |last10=Yamada |first11= Kohei |last11=Suenaga |first12= Fuyuki |last12=Ishikawa |first13= Kenji |last13=Kamijo |first14= Yoshiyuki |last14=Shinya |first15= Takamasa |last15=Suetomi |date=5 July 2022 |title=Goal-Aware RSS for Complex Scenarios Via Program Logic |url=https://ieeexplore.ieee.org/document/9815834 |journal=IEEE Transactions on Intelligent Vehicles |volume= 8 |issue= 4 |language=en |pages=3040–3072 |doi=10.1109/TIV.2022.3169762 |arxiv= 2207.02387 |s2cid= 250311612 }}

The US has standardized the use of turquoise lights to inform other drivers that a vehicle is driving autonomously. It will be used in the 2026 Mercedes-Benz EQS and S-Class sedans with Drive Pilot, an SAE Level 3 driving system.{{Citation needed|date=February 2024}}

As of 2023, the Turquoise light had not been standardized by the P.R.C or the UN-ECE.{{Cite web |last=Tucker |first=Sean |date=2023-12-19 |title=Thanks to Mercedes, Turquoise Lights Mean Self-Driving |url=https://www.kbb.com/car-news/thanks-to-mercedes-turquoise-lights-mean-self-driving/ |access-date=2024-02-03 |website=Kelley Blue Book |language=en-US}}

Artificial intelligence (AI) plays a pivotal role in the development and operation of autonomous vehicles (AVs), enabling them to perceive their surroundings, make decisions, and navigate safely without human intervention. AI algorithms empower AVs to interpret sensory data from various onboard sensors, such as cameras, LiDAR, radar, and GPS, to understand their environment and improve its technological ability and overall safety over time.{{Cite web |date=2022-03-07 |title=How AI Is Making Autonomous Vehicles Safer |url=https://hai.stanford.edu/news/how-ai-making-autonomous-vehicles-safer |access-date=2024-04-23 |website=hai.stanford.edu |language=en}}

File:Autonomous Delivery Vehicle Pileup.jpg

The primary obstacle to ACs is the advanced software and mapping required to make them work safely across the wide variety of conditions that drivers experience.{{cite news |last=Henn |first=Steve |date=31 July 2015 |title=Remembering When Driverless Elevators Drew Skepticism |newspaper=NPR.org |publisher=NPR |url=https://www.npr.org/2015/07/31/427990392/remembering-when-driverless-elevators-drew-skepticism |access-date=14 August 2016}} In addition to handling day/night driving in good and bad weather on roads of arbitrary quality, ACs must cope with other vehicles, road obstacles, poor/missing traffic controls, flawed maps, and handle endless edge cases, such as following the instructions of a police officer managing traffic at a crash site.

Other obstacles include cost, liability,{{Cite book|title=Being digital|last=Negroponte|first=Nicholas|date=1 January 2000|publisher=Vintage Books|isbn=978-0679762904|oclc=68020226}}{{cite web|url=http://www.technewsworld.com/story/83102.html|title=Feds Put AI in the Driver's Seat|last=Adhikari|first=Richard|date=11 February 2016|work=Technewsworld|access-date=12 February 2016}} consumer reluctance,{{cite press release|url=http://www.prnewswire.com/news-releases/new-allstate-survey-shows-americans-think-they-are-great-drivers---habits-tell-a-different-story-126563103.html|title=New Allstate Survey Shows Americans Think They Are Great Drivers – Habits Tell a Different Story|date=2 August 2011|agency=PR Newswire|access-date=7 September 2013}} ethical dilemmas,{{cite magazine |last=Lin |first=Patrick |date=8 October 2013 |title=The Ethics of Autonomous Cars |url=https://www.theatlantic.com/technology/archive/2013/10/the-ethics-of-autonomous-cars/280360/ |magazine=The Atlantic}}{{cite journal |last1=Skulmowski |first1=Alexander |last2=Bunge |first2=Andreas |last3=Kaspar |first3=Kai |last4=Pipa |first4=Gordon |date=16 December 2014 |title=Forced-choice decision-making in modified trolley dilemma situations: a virtual reality and eye tracking study |journal=Frontiers in Behavioral Neuroscience |volume=8 |page=426 |doi=10.3389/fnbeh.2014.00426 |pmc=4267265 |pmid=25565997 |doi-access=free}} security,{{Cite journal |last1=Alsulami |first1=Abdulaziz A. |last2=Abu Al-Haija |first2=Qasem |last3=Alqahtani |first3=Ali |last4=Alsini |first4=Raed |date=15 July 2022 |title=Symmetrical Simulation Scheme for Anomaly Detection in Autonomous Vehicles Based on LSTM Model |journal=Symmetry |language=en |volume=14 |issue=7 |pages=1450 |bibcode=2022Symm...14.1450A |doi=10.3390/sym14071450 |issn=2073-8994 |doi-access=free}}{{cite web |last=Moore-Colyer |first=Roland |date=12 February 2015 |title=Driverless cars face cyber security, skills and safety challenges |url=https://www.v3.co.uk/v3-uk/analysis/2394924/driverless-cars-face-cyber-security-skills-and-safety-challenges |access-date=24 April 2015 |website=v3.co.uk}}{{Cite journal |last1=Petit |first1=J. |last2=Shladover |first2=S.E. |date=1 April 2015 |title=Potential Cyberattacks on Automated Vehicles |journal=IEEE Transactions on Intelligent Transportation Systems |volume=16 |issue=2 |pages=546–556 |doi=10.1109/TITS.2014.2342271 |issn=1524-9050 |s2cid=15605711}}{{cite web |last=Tussy |first=Ron |date=29 April 2016 |title=Challenges facing Autonomous Vehicle Development |url=http://auto-sens.com/the-challenges-facing-autonomous-vehicles/ |access-date=5 May 2016 |publisher=AutoSens}} privacy,{{cite news |last=Gomes |first=Lee |date=28 August 2014 |title=Hidden Obstacles for Google's Self-Driving Cars |magazine=MIT Technology Review |url=http://www.technologyreview.com/news/530276/hidden-obstacles-for-googles-self-driving-cars/ |url-status=dead |access-date=22 January 2015 |archive-url=https://web.archive.org/web/20150316001705/http://www.technologyreview.com/news/530276/hidden-obstacles-for-googles-self-driving-cars/ |archive-date=16 March 2015}} and legal/regulatory framework.{{cite news|url=https://www.forbes.com/sites/quora/2013/09/24/will-regulators-allow-self-driving-cars-in-a-few-years/|title=Will Regulators Allow Self-Driving Cars in a Few Years?|date=24 September 2013|work=Forbes|access-date=5 January 2014}} Further, AVs could automate the work of professional drivers, eliminating many jobs, which could slow acceptance.{{cite web|url=https://www.theverge.com/2013/11/18/5120270/reliance-on-autopilot-is-now-the-biggest-threat-to-flight-safety|title=Reliance on autopilot is now the biggest threat to flight safety, study says|date=18 November 2013 |work=The Verge |first=Casey |last=Newton |access-date=19 November 2013}}

Tesla calls its Level 2 ADAS "Full Self-Driving (FSD) Beta".{{cite web |last=Stumpf |first=Rob |title=Tesla Admits Current "Full Self-Driving Beta" Will Always Be a Level 2 System: Emails

|url=https://www.thedrive.com/tech/39647/tesla-admits-current-full-self-driving-beta-will-always-be-a-level-2-system-emails |access-date=29 August 2021 |website=The Drive|date=8 March 2021 }} US Senators Richard Blumenthal and Edward Markey called on the Federal Trade Commission (FTC) to investigate this marketing in 2021.{{cite web |author=Keith Barry |title=Senators Call for Investigation of Tesla's Marketing Claims of Its Autopilot and "Full Self-Driving" Features

|url=https://www.consumerreports.org/advertising-claims/call-to-investigate-tesla-marketing-claims-autopilot-fsd-a1181594362/

|website=Consumer Reports |access-date=13 April 2020 }} In December 2021 in Japan, Mercedes-Benz was punished by the Consumer Affairs Agency for misleading product descriptions.{{cite news|date=10 December 2021 |url=https://www3.nhk.or.jp/news/html/20211210/k10013383761000.html| title=メルセデス・ベンツ日本に措置命令 事実と異なる記載 消費者庁 |trans-title=Administrative order to Mercedes-Benz Japan Co., Ltd. for the descriptions that are different from the fact – The Consumer Affairs Agency |language=ja |work=NHK, Japan |access-date=13 April 2022 }}

Mercedes-Benz was criticized for a misleading US commercial advertising E-Class models.{{cite news |date=28 July 2016 |author=Steph Willems |title=Mercedes-Benz Slammed Over Misleading Commercial |url=https://www-thetruthaboutcars-com.translate.goog/2016/07/mercedes-benz-slammed-misleading-ad-ad-seems-go-missing/?_x_tr_sl=en&_x_tr_tl=ja&_x_tr_hl=ja&_x_tr_pto=op,sc

|work=The Truth About Cars |access-date=15 April 2022}} At that time, Mercedes-Benz rejected the claims and stopped its "self-driving car" ad campaign that had been running.{{cite news |date=29 July 2016 |author=Aaron Brown |title=Mercedes-Benz to Stop Running "Self-Driving Car" Ads |url=https://www-thedrive-com.translate.goog/news/4632/mercedes-benz-to-stop-running-self-driving-car-ads?_x_tr_sl=en&_x_tr_tl=ja&_x_tr_hl=ja&_x_tr_pto=op,sc |work=The Drive |access-date=15 April 2022}}{{cite news |date=25 April 2016 |title=Mercedes rejects claims about "misleading" self-driving car ads |url=https://www-reuters-com.translate.goog/article/us-mercedes-marketing-idUSKCN1081VV?_x_tr_sl=en&_x_tr_tl=ja&_x_tr_hl=ja&_x_tr_pto=op,sc |agency=Reuters |access-date=15 April 2022 |archive-date=31 May 2022 |archive-url=https://web.archive.org/web/20220531204702/https://www-reuters-com.translate.goog/article/us-mercedes-marketing-idUSKCN1081VV?_x_tr_sl=en&_x_tr_tl=ja&_x_tr_hl=ja&_x_tr_pto=op,sc |url-status=dead }} In August 2022, the California Department of Motor Vehicles (DMV) accused Tesla of deceptive marketing practices.{{cite news|date=9 August 2022 |url=https://www.cbtnews.com/california-dmw-accuses-tesla-of-deceptive-marketing-for-its-self-driving-tech/ |title=California DMV accuses Tesla of deceptive marketing for its self-driving tech |work=CBT Automotive Network |access-date=22 November 2022 }}

With the Automated Vehicles Bill (AVB) self-driving car-makers could face prison for misleading adverts in the United-Kingdom.{{Cite web |first=Matthew |last=Sparkes |date=13 November 2023 |title=Self-driving car-makers could face prison for misleading adverts in UK |url=https://www.newscientist.com/article/2402075-self-driving-car-makers-could-face-prison-for-misleading-adverts-in-uk/ |access-date=2024-02-02 |website=New Scientist |language=en-US}}

In the 2020s, concerns over ACs' vulnerability to cyberattacks and data theft emerged.{{cite journal |date=28 June 2021 |author=James Andrew Lewis |title=National Security Implications of Leadership in Autonomous Vehicles |url=https://www.csis.org/analysis/national-security-implications-leadership-autonomous-vehicles |website=CSIS |access-date=12 April 2022 }}

In 2018 and 2019, former Apple engineers were charged with stealing information related to Apple's self-driving car project.{{cite news|date=11 July 2018 |author=Allyson Chiu |title=Ex-Apple engineer arrested on his way to China, charged with stealing company's autonomous car secrets |url=https://www.washingtonpost.com/news/morning-mix/wp/2018/07/11/ex-apple-engineer-arrested-on-his-way-to-china-charged-with-stealing-companys-autonomous-car-secrets/ |newspaper=The Washington Post |access-date=18 April 2022}}{{cite news

|date=22 August 2022 |author=Kif Leswing |title=Former Apple engineer accused of stealing automotive trade secrets pleads guilty |url=https://www.cnbc.com/2022/08/22/former-apple-employee-xiaolang-zhang-pleads-guilty-.html |publisher=CNBC |access-date=23 August 2022}}{{cite news|date=30 January 2019 |author=Sean O'Kane |title=A second Apple employee was charged with stealing self-driving car project secrets |url=https://www.theverge.com/2019/1/30/18203718/apple-self-driving-trade-secrets-china-titan |work=The Verge |access-date=18 April 2022}} In 2021 the United States Department of Justice (DOJ) accused Chinese security officials of coordinating a hacking campaign to steal information from government entities, including research related to autonomous vehicles.{{cite web|date=19 July 2021 |title=Four Chinese Nationals Working with the Ministry of State Security Charged with Global Computer Intrusion Campaign Targeting Intellectual Property and Confidential Business Information, Including Infectious Disease Research |url=https://www.justice.gov/opa/pr/four-chinese-nationals-working-ministry-state-security-charged-global-computer-intrusion |website=DOJ, US |access-date=14 June 2022 }}{{cite news|date=19 July 2021 |author=Katie Benner |title=The Justice Dept. accuses Chinese security officials of a hacking attack seeking data on viruses like Ebola. |url=https://www.nytimes.com/2021/07/19/us/politics/chinese-hackers-justice-dept.html |newspaper=The New York Times |access-date=14 June 2022 }} China has prepared "the Provisions on Management of Automotive Data Security (Trial) to protect its own data".{{cite web|date=24 August 2021 |author1=Mark Schaub |author2=Atticus Zhao |author3=Mark Fu |title=China MIIT formulating new rules on data security |url=https://www.kwm.com/cn/en/insights/latest-thinking/china-issues-new-rules-on-data-security-in-auto-industry.html |website=King & Wood Mallesons |access-date=23 April 2022 }}{{cite magazine |date=1 July 2022 |author=Justin Ling |title=Is Your New Car a Threat to National Security? |url=https://www.wired.com/story/china-cars-surveillance-national-security/ |magazine=Wired |access-date=3 July 2022 }}

Cellular Vehicle-to-Everything technologies are based on 5G wireless networks.{{cite web|date=4 November 2019 |author= Charles McLellan |title=What is V2X communication? Creating connectivity for the autonomous car era |url=https://www.zdnet.com/article/guide-to-autonomous-vehicles-what-business-leaders-need-to-know/ |publisher=ZDNet |access-date=8 May 2022 }} {{As of|November 2022}}, the US Congress was considering the possibility that imported Chinese AC technology could facilitate espionage.{{cite news|date=21 November 2022 |url=https://www.wired.co.uk/article/autonomous-vehicles-china-us-national-security |title=Autonomous Vehicles Join the List of US National Security Threats |magazine=Wired |access-date=22 November 2022 }}

Testing of Chinese automated cars in the US has raised concern over which US data are collected by Chinese vehicles to be stored in China and any link with the Chinese communist party.{{cite news|title=US lawmakers raise concerns over Chinese self-driving testing data collection |first=David |last=Shepardson |date=16 November 2023 |publisher=[Reuters] |url=https://www.reuters.com/business/autos-transportation/lawmakers-voice-concern-chinese-autonomous-vehicle-firms-collecting-us-testing-2023-11-16/ |access-date=February 1, 2024}}

ACs complicate the need for drivers to communicate with each other, e.g., to decide which car enters an intersection first. In an AC without a driver, traditional means such as hand signals do not work (no driver, no hands).{{cite web|url=https://www.nbcnews.com/mach/science/what-s-big-orange-covered-leds-start-s-new-approach-ncna897151|title=What's big, orange and covered in LEDs? This start-up's new approach to self-driving cars|date=3 August 2018 |publisher=NBC News}}

ACs must be able to predict the behavior of possibly moving vehicles, pedestrians, etc, in real time in order to proceed safely. The task becomes more challenging the further into the future the prediction extends, requiring rapid revisions to the estimate to cope with unpredicted behavior. One approach is to wholly recompute the position and trajectory of each object many times per second. Another is to cache the results of an earlier prediction for use in the next one to reduce computational complexity.{{cite conference |last1=Crosato |first1=Luca |last2=Shum |first2=Hubert P. H. |last3=Ho |first3=Edmond S. L. |last4=Wei |first4=Chongfeng |last5=Sun |first5=Yuzhu |date=2024 |title=A Virtual Reality Framework for Human-Driver Interaction Research: Safe and Cost-Effective Data Collection |url=https://doi.org/10.1145/3610977.3634923 |conference=2024 ACM/IEEE International Conference on Human Robot Interaction |publisher=ACM/IEEE|doi=10.1145/3610977.3634923 }}{{Cite web |last=City University of Hong Kong |date=September 6, 2023 |title=Novel AI system enhances the predictive accuracy of autonomous driving |url=https://techxplore.com/news/2023-09-ai-accuracy-autonomous.html |website=techxplore.com}}

The ADAS has to be able to safely accept control from and return control to the driver.{{cite web|date=25 April 2018 |title=Human Factors behind Autonomous Vehicles |url=https://www-robsonforensic-com.translate.goog/articles/autonomous-vehicle-human-factors-expert?_x_tr_sl=en&_x_tr_tl=ja&_x_tr_hl=ja&_x_tr_pto=op,sc |website=Robson Forensic

|access-date=17 April 2022 }}

Consumers will avoid ACs unless they trust them as safe.{{Cite journal|date=1 January 2015|title=Trust in Automation – Before and After the Experience of Take-over Scenarios in a Highly Automated Vehicle|journal=Procedia Manufacturing|volume=3|pages=3025–3032|doi=10.1016/j.promfg.2015.07.847|issn=2351-9789|last1=Gold|first1=Christian|last2=Körber|first2=Moritz|last3=Hohenberger|first3=Christoph|last4=Lechner|first4=David|last5=Bengler|first5=Klaus|doi-access=free}}{{Cite news|url=http://gmauthority.com/blog/2017/08/survey-data-suggests-self-driving-cars-could-be-slow-to-gain-consumer-trust/|title=Survey Data Suggests Self-Driving Cars Could Be Slow To Gain Consumer Trust|work=GM Authority|access-date=3 September 2018}} Robotaxis operating in San Francisco received pushback over perceived safety risks.{{Cite web |date=2023-08-11 |title=California agency approves San Francisco robotaxi expansion amid heavy opposition |url=https://www.cnbc.com/2023/08/11/san-francisco-robotaxi-expansion-approval-faces-heavy-opposition.html |access-date=2024-02-02 |website=CNBC |language=en}} Automatic elevators were invented in 1900, but did not become common until operator strikes and trust was built with advertising and features such as an emergency stop button.{{cite news|url=https://www.npr.org/2015/07/31/427990392/remembering-when-driverless-elevators-drew-skepticism|title=Remembering When Driverless Elevators Drew Skepticism|website=NPR.org}}{{cite news|url=https://www.npr.org/sections/money/2015/07/29/427467598/episode-642-the-big-red-button|title=Episode 642: The Big Red Button|website=NPR.org}} However, with repeated use of autonomous driving functions, drivers' behavior and trust in autonomous vehicles gradually improved and both entered a more stable state. At the same time, this also improved the performance and reliability of the vehicle in complex conditions, thereby increasing public trust.{{Cite journal |last1=Metz |first1=Barbara |last2=Wörle |first2=Johanna |last3=Hanig |first3=Michael |last4=Schmitt |first4=Marcus |last5=Lutz |first5=Aaron |last6=Neukum |first6=Alexandra |date=2021-08-01 |title=Repeated usage of a motorway automated driving function: Automation level and behavioural adaption |journal=Transportation Research Part F: Traffic Psychology and Behaviour |volume=81 |pages=82–100 |doi=10.1016/j.trf.2021.05.017 |issn=1369-8478|doi-access=free |bibcode=2021TRPF...81...82M }}

Autonomy also presents various political and economic implications. The transportation sector holds significant sway in many political and economic landscapes. For instance, many US states generate much annual revenue from transportation fees and taxes.{{Cite web |last=Talbott |first=Selika Josiah |title=The Political Economy Of Autonomous Vehicles |url=https://www.forbes.com/sites/selikajosiahtalbott/2020/06/23/the-political-economy-of-autonomous-vehicles/ |access-date=2024-04-23 |website=Forbes |language=en}} The advent of self-driving cars could profoundly affect the economy by potentially altering state tax revenue streams. Furthermore, the transition to autonomous vehicles might disrupt employment patterns and labor markets, particularly in industries heavily reliant on driving professions. Data from the U.S. Bureau of Labor Statistics indicates that in 2019, the sector employed over two million individuals as tractor-trailer truck drivers.{{cite web |title=Occupational Outlook Handbook: Heavy and Tractor trailer Truck Drivers |url=https://www.bls.gov/ooh/transportation-and-material-moving/heavy-and-tractor-trailer-truck-drivers.htm |website=U.S. Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |access-date=24 April 2024}} Additionally, taxi and delivery drivers represented approximately 370,400 positions, and bus drivers constituted a workforce of over 680,000.{{cite web |title=Occupational Outlook Handbook: Delivery Truck Drivers and Driver/Sales Workers |url=https://www.bls.gov/ooh/transportation-and-material-moving/delivery-truck-drivers-and-driver-sales-workers.htm |website=U.S. Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |access-date=24 April 2024}}{{cite web |title=Occupational Outlook Handbook: Taxi Drivers, Shuttle Drivers, and Chauffeurs |url=https://www.bls.gov/ooh/transportation-and-material-moving/taxi-drivers-and-chauffeurs.htm |website=U.S. Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |access-date=24 April 2024}}{{cite web |title=Occupational Outlook Handbook: Bus Drivers |url=https://www.bls.gov/ooh/transportation-and-material-moving/bus-drivers.htm |website=U.S. Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |access-date=24 April 2024}} Collectively, this amounts to a conceivable displacement of nearly 2.9 million jobs, surpassing the job losses experienced in the 2008 Great Recession.{{cite web |last1=Goodman, Mance |first1=Christopher, Steven |title=Employment Loss and the 2007–09 Recession: An Overview |url=https://www.bls.gov/opub/mlr/2011/04/art1full.pdf |website=U.S. Bureau of Labor Statistics |access-date=24 April 2024}}

The prominence of certain demographic groups within the tech industry inevitably shapes the trajectory of autonomous vehicle (AV) development, potentially perpetuating existing inequalities.{{Cite web |title=Diversity and STEM: Women, Minorities, and Persons with Disabilities 2023 {{!}} NSF - National Science Foundation |url=https://ncses.nsf.gov/pubs/nsf23315/report |access-date=2024-04-23 |website=ncses.nsf.gov}}

{{See also|Machine ethics}}

Research from Georgia Tech revealed that autonomous vehicle detection systems were generally five percent less effective at recognizing darker-skinned individuals. This accuracy gap persisted despite adjustments for environmental variables like lighting and visual obstructions.{{cite web |last1=Samuel |first1=Sigal |title=A new study finds a potential risk with self-driving cars: failure to detect dark-skinned pedestrians |url=https://www.vox.com/future-perfect/2019/3/5/18251924/self-driving-car-racial-bias-study-autonomous-vehicle-dark-skin |website=Vox |date=5 March 2019 |publisher=VoxMedia |access-date=22 April 2024}}

Standards for liability have yet to be adopted to address crashes and other incidents. Liability could rest with the vehicle occupant, its owner, the vehicle manufacturer, or even the ADAS technology supplier, possibly depending on the circumstances of the crash.{{Cite journal

|author1=Alexander Hevelke |author2=Julian Nida-Rümelin |title=Responsibility for Crashes of Autonomous Vehicles: An Ethical Analysis |journal=Sci Eng Ethics |year=2015 |volume=21 |issue=3 |pages=619–630 |doi=10.1007/s11948-014-9565-5 |pmid=25027859 |pmc=4430591 }} Additionally, the infusion of ArtificiaI Intelligence technology in autonomous vehicles adds layers of complexity to ownership and ethical dynamics. Given that AI systems are inherently self-learning, a question arises of whether accountability should rest with the vehicle owner, the manufacturer, or the AI developer.{{Cite web |date=2022-05-18 |title=The Ethical Considerations of Self-Driving Cars |url=https://montrealethics.ai/the-ethical-considerations-of-self-driving-cars/ |access-date=2024-04-23 |website=Montreal AI Ethics Institute |language=en-US}}

The trolley problem is a thought experiment in ethics. Adapted for ACs, it considers an AC carrying one passenger confronting a pedestrian who steps in its way. The ADAS notionally has to choose between killing the pedestrian or swerving into a wall, killing the passenger.{{Cite journal|last=Himmelreich|first=Johannes|date=17 May 2018|title=Never Mind the Trolley: The Ethics of Autonomous Vehicles in Mundane Situations|journal=Ethical Theory and Moral Practice|volume=21|issue=3|pages=669–684|doi=10.1007/s10677-018-9896-4|s2cid=150184601|issn=1386-2820}} Possible frameworks include deontology (formal rules) and utilitarianism (harm reduction).{{Cite book|title=Road vehicle automation|url={{google books |plainurl=y |id=TxmfDAAAQBAJ}}|last1=Meyer|first1=G.|last2=Beiker|first2=S|publisher=Springer International Publishing|year=2014|pages=93–102}}{{Cite journal|last=Karnouskos|first=Stamatis|year=2020|title=Self-Driving Car Acceptance and the Role of Ethics|journal=IEEE Transactions on Engineering Management|volume=67|issue=2|pages=252–265|doi=10.1109/TEM.2018.2877307|s2cid=115447875|issn=0018-9391}}

One public opinion survey reported that harm reduction was preferred, except that passengers wanted the vehicle to prefer them, while pedestrians took the opposite view. Utilitarian regulations were unpopular.{{cite journal |author1=Jean-François Bonnefon |author2=Azim Shariff |author3=Iyad Rahwan |year=2016 |title=The Social Dilemma of Autonomous Vehicles |journal=Science |volume=352 |issue=6293 |pages=1573–6 |arxiv=1510.03346 |bibcode=2016Sci...352.1573B |doi=10.1126/science.aaf2654 |pmid=27339987 |s2cid=35400794}} Additionally, cultural viewpoints exert substantial influence on shaping responses to these ethical quandaries. Another study found that cultural biases impact preferences in prioritizing the rescue of certain individuals over others in car accident scenarios.

Some ACs require an internet connection to function, opening the possibility that a hacker might gain access to private information such as destinations, routes, camera recordings, media preferences, and/or behavioral patterns, although this is true of an internet-connected device.{{Cite journal|last1=Lim |first1=Hazel Si Min |last2=Taeihagh |first2=Araz |year=2018 |title=Autonomous Vehicles for Smart and Sustainable Cities: An In-Depth Exploration of Privacy and Cybersecurity Implications |journal=Energies |volume=11 |issue=5 |page=1062 |doi=10.3390/en11051062|bibcode=2018arXiv180410367L|arxiv=1804.10367|s2cid=13749987|doi-access=free}}{{Cite news|url=https://www.theatlantic.com/technology/archive/2016/03/self-driving-cars-and-the-looming-privacy-apocalypse/474600/|title=How Self-Driving Cars Will Threaten Privacy|last=Lafrance|first=Adrienne|date=21 March 2016|access-date=4 November 2016}}{{Cite journal|last=Jack|first=Boeglin|date=1 January 2015|title=The Costs of Self-Driving Cars: Reconciling Freedom and Privacy with Tort Liability in Autonomous Vehicle Regulation|url=http://digitalcommons.law.yale.edu/yjolt/vol17/iss1/4|journal=Yale Journal of Law and Technology|volume=17|issue=1}}

ACs make use of road infrastructure (e.g., traffic signs, turn lanes) and may require modifications to that infrastructure to fully achieve their safety and other goals.{{cite web|date=26 January 2023 |author=Steve McEvoy |title=What are the next steps to reaching Level 4 autonomy? |url=https://www.automotiveworld.com/articles/what-are-the-next-steps-to-reaching-level-4-autonomy/ |website=Automotive World |access-date=5 April 2023 }} In March 2023, the Japanese government unveiled a plan to set up a dedicated highway lane for ACs.{{cite web|date=1 April 2023 |title=Japan Planning 100-kilometer Lane for Self-Driving Vehicles |url=https://japannews.yomiuri.co.jp/society/general-news/20230401-100928/ |website=Yomiuri Shimbun |access-date=11 April 2023}} In April 2023, JR East announced their challenge to raise their self-driving level of Kesennuma Line bus rapid transit (BRT) in rural area from the current Level 2 to Level 4 at 60 km/h.{{cite web|date=4 April 2023 |title=気仙沼線 BRT における自動運転レベル4認証取得を目指します |trans-title=Challenging self-driving Level 4 approval of Kesennuma Line BRT |url=https://www.jreast.co.jp/press/2023/20230404_ho01.pdf |website=JR East |access-date=5 April 2023 }}

ACs can be tested via digital simulations,{{cite web |date=18 June 2018 |title=Automobile simulation example |url=https://www.cyberbotics.com/doc/automobile/index |access-date=18 June 2018 |work=Cyberbotics}}{{cite journal |last1=Hallerbach |first1=S. |last2=Xia |first2=Y. |last3=Eberle |first3=U. |last4=Koester |first4=F. |year=2018 |title=Simulation-Based Identification of Critical Scenarios for Cooperative and Automated Vehicles |url=https://www.researchgate.net/publication/324194968 |journal=SAE International Journal of Connected and Automated Vehicles |publisher=SAE International |volume=1 |issue=2 |pages=93–106 |doi=10.4271/2018-01-1066}} in a controlled test environment,{{cite web|url=http://www.mtc.umich.edu/|title=Mcity testing center|date=8 December 2016|work=University of Michigan|access-date=13 February 2017|archive-date=16 February 2017|archive-url=https://web.archive.org/web/20170216092805/http://www.mtc.umich.edu/|url-status=dead}} and/or on public roads. Road testing typically requires some form of permit{{cite web|url=https://www.dmv.ca.gov/portal/dmv/detail/vr/autonomous/testing|title=Adopted Regulations for Testing of Autonomous Vehicles by Manufacturers|date=18 June 2016|work=DMV|access-date=13 February 2017}} or a commitment to adhere to acceptable operating principles.{{cite web|url=https://www.gov.uk/government/publications/automated-vehicle-technologies-testing-code-of-practice|title=The Pathway to Driverless Cars: A Code of Practice for testing|date=19 July 2015|access-date=8 April 2017}} For example, New York requires a test driver to be in the vehicle, prepared to override the ADAS as necessary.{{cite web | url=https://dmv.ny.gov/dmv/apply-autonomous-vehicle-technology-demonstration-testing-permit | title=Apply for an Autonomous Vehicle Technology Demonstration / Testing Permit| date=9 May 2017}}

File:Waymo self-driving car front view.gk.jpg's self-driving car, navigating public streets in Mountain View, California in 2017]]

In California, self-driving car manufacturers are required to submit annual reports describing how often their vehicles autonomously disengaged from autonomous mode.{{cite web|title=Disengagement Reports |url=https://www.dmv.ca.gov/portal/vehicle-industry-services/autonomous-vehicles/disengagement-reports/ |website=California DMV |access-date=24 April 2022 }} This is one measure of system robustness (ideally, the system should never disengage).{{cite news|author=Brad Templeton |date=9 February 2021 |title=California Robocar Disengagement Reports Reveal Tidbits About Tesla, AutoX, Apple, Others |url=https://www.forbes.com/sites/bradtempleton/2021/02/09/california-robocar-disengagement-reports-reveal-about-tesla-autox-apple-others/?sh=170c8ff27fab |work=Forbes |access-date=24 April 2022 }}

In 2017, Waymo reported 63 disengagements over {{convert|352545|mi|abbr=on}} of testing, an average distance of {{convert|5596|mi|abbr=on}} between disengagements, the highest (best) among companies reporting such figures. Waymo also logged more autonomous miles than other companies. Their 2017 rate of 0.18 disengagements per {{convert|1000|mi|abbr=on}} was an improvement over the 0.2 disengagements per {{convert|1000|mi|abbr=on}} in 2016, and 0.8 in 2015. In March 2017, Uber reported an average of {{convert|0.67|mi|abbr=on}} per disengagement. In the final three months of 2017, Cruise (owned by GM) averaged {{convert|5224|mi|abbr=on}} per disengagement over {{convert|62689|mi|abbr=on}}.{{Cite news|url=https://www.nextbigfuture.com/2018/03/uber-self-driving-system-was-still-400-times-worse-waymo-in-2018-on-key-distance-intervention-metric.html|title=Uber' self-driving system was still 400 times worse [than] Waymo in 2018 on key distance intervention metric|last=Wang|first=Brian|date=25 March 2018|work=NextBigFuture.com|access-date=25 March 2018}}