자율주행차(autonomous car)는 무인자동차(driverless car) 또는 로봇자동차(robo-car)로도 알려진, 인간의 조작이 줄어들거나 전혀 없이 운행이 가능한 자동차이다. 이들은 때때로 로보택시(robotaxi)라고 불리기도 하지만, 이 용어는 특히 차량 공유 회사가 운영하는 자율주행차를 지칭한다.

Autonomous car

{{Short description|Vehicle operated with reduced human input on public roads}} {{About|passenger cars on public roads|the general concept|Vehicular automation|wider applications|Unmanned ground vehicle}} {{Use American English|date=November 2022}} {{Use dmy dates|date=November 2022}} {{Self-driving car}}

A '''self-driving car''', also known as an '''autonomous car''', '''driverless car''', or '''robotic car''' ('''robo-car'''), is a [[car]] that is capable of operating with reduced or no [[User input|human input]]. They are sometimes called '''[[robotaxi]]s''', though this term refers specifically to self-driving cars operated for a [[ridesharing company]].

{{As of|2026}}, the term "self-driving" lacks an agreed standard definition and is also subject to commercial advertising and branding considerations. In 2020, [[Waymo]] was the first to offer rides in driverless taxis in the [[operational design domain]] (ODD) of limited geographic areas, but {{As of|2025|pre=as of late|bare=yes}}, no system has achieved full autonomy in all domains - sometimes referred to as "Level 5" on a scale of 0 to 5 levels of automation defined by the global standards organisation [[SAE International]], or simply "no driver" as given by the classification system proposed by [[Mobileye]] in the US.

Following a history of experimentation and development of [[Advanced driver assistance systems|advanced driver assistance]] systems (ADAS) after WWII, two main technologies are now primarily used: [[Lidar|LiDAR]] (Light Detection and Ranging), and visual sensors (cameras) which capture images and video like human eyes. These are combined with systems such as [[Global Positioning System|GPS]], [[Neural network (machine learning)|neural networks]], [[artificial intelligence]], and established ADAS engineering to deliver levels of driving autonomy.

With more self-driving cars on public roads, an increasing number of safety incidents, collisions and even deaths have been recorded around the world. The primary obstacle to self-driving is the advanced software and mapping required to make them work safely across the wide variety of conditions that drivers experience. Other issues include security of [[over-the-air update]]s, legal and [[Regulation of self-driving cars|regulatory]] issues, [[ethics]] and consumer confidence. Methods of testing and monitoring the reliability of cars have evolved in parallel with the deployment of cars with self-driving capabilities, with various standards for this being proposed. Should autonomous cars gain mass adoption, wider implications for urban infrastructure and the economy have also been discussed.

Public perception and acceptance of autonomous cars has been found to be mixed. A 2014 telephone poll in the US found 31.7% would not continue to drive once an automated car was available to them, while a survey in 2022 found only a quarter (27%) of the world's population would feel safe in one.

== History == {{Excerpt|History of self-driving cars|only=paragraphs}}

== Definitions == Organizations such as the global standards body [[SAE International]] (SAE) have proposed terminology to describe technical capabilities. However, most terms have no standard definition and are employed variously by vendors and others. Proposals to adopt aviation automation terminology for cars has also 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}}

The first consideration is the [[operational design domain]] (ODD). {{Excerpt|Operational design domain|only=paragraphs|paragraphs=2|hat=no|inline=yes}}

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|abbr=on}}.{{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 {{Convert|750,000|mi}}.{{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 {{Convert|130,000|mi}} 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=J.D. Power}}

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 [[Tesla Autopilot|Full Self-Driving]], [[Tesla, Inc.|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.{{Cite news |date=2018-06-11 |title=Car insurers warn on 'autonomous' vehicles |url=https://www.bbc.co.uk/news/technology-44439523 |access-date=2025-12-25 |work=BBC News |language=en-GB}}

=== Concepts === The following are useful in understanding the various definitions and criteria in use for self-driving cars.

==== Driving systems ==== [[Advanced driver-assistance system]]s (ADAS) automate 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 |bibcode=2024ETRR...16...39A |doi=10.1186/s12544-024-00654-0 |issn=1866-8887 |doi-access=free}} An ADAS requires a human driver to handle tasks that the ADAS does not support.

ADAS contrasts to an [[automated driving system]] (ADS), which would be classified by SAE J3016 as Level 3 or higher.{{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}}

==== Autonomy versus automation ==== 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 |bibcode=1991ICSys..11d...5A |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}}

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

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

==== Cooperative system ==== 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 |bibcode=2022IEEEA..1064920Y |doi=10.1109/ACCESS.2022.3183758 |doi-access=free}}

According to [[#SAE classification|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.}}

==== Self-driving ==== 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 |title=Self-Driving Cars Explained |url=https://www.ucsusa.org/clean-vehicles/how-self-driving-cars-work |website=Union of Concerned Scientists}}

The British [[Automated and Electric Vehicles Act 2018]] 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=Penningtons Law |date=24 July 2018 }}

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

=== British definitions === In [[British English]], the word "automated" 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 a [[UNECE Regulations|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 |title=Automated Vehicle Bill |url=https://bills.parliament.uk/publications/52908/documents/3984 |website=parliament.uk}}

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.

=== SAE classification{{anchor|automation-levels}} === [[File:Tesla Autopilot Engaged in Model X.jpg|thumb|[[Tesla Autopilot]] is classified as an SAE Level 2 system.{{cite web|url=https://www.tesla.com/support/autopilot |title=Support – Autopilot |work=[[Tesla, Inc.|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 |last=Baldwin |first=Roberto |date=9 March 2021 |title=Tesla Tells California DMV that FSD Is Not Capable of Autonomous Driving |url=https://www.caranddriver.com/news/a35785277/tesla-fsd-california-self-driving/ |website=[[Car and Driver]]}}]]

In 2014, the association for transport and other industries [[SAE International]] published a six-level classification system – ranging from fully manual to fully automated – as J3016, ''Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems''; the details are revised occasionally.{{cite web |publisher=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}}

The system was adopted by the [[United Nations Economic Commission for Europe]] (UNECE) in the form of [[Regulation of self-driving cars#UNECE WP.29 GRVA|UN Regulation No. 157]], which came into force in January 2021, and used in over 50 countries.{{Citation |title=World Forum for Harmonization of Vehicle Regulations (WP.29): how it works, how to join it |date=2019 |url=https://digitallibrary.un.org/record/3824138 |access-date=2025-12-29 |others=UN. ECE |publisher=UN}} After SAE updated its classification in 2016, (J3016_201609),{{Cite web |date=September 30, 2016 |title=Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles |url=https://www.sae.org/standards/j3016_201609-taxonomy-definitions-terms-related-driving-automation-systems-road-motor-vehicles |access-date=2025-12-29 |website=[[SAE International]]}} the US [[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=[[National Highway Traffic Safety Administration|NHTSA]] |location=US |access-date=1 December 2021}}{{Cite web |title=Automated Vehicles for Safety |url=https://www.nhtsa.gov/technology-innovation/automated-vehicles-safety |archive-url=https://web.archive.org/web/20240101060412/https://www.nhtsa.gov/technology-innovation/automated-vehicles-safety |archive-date=1 January 2024 |access-date=2025-12-29 |website=NHTSA |language=en |url-status=live }}

The classification is a topic of debate, having been criticized for its technological focus, 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 |url-access=subscription }}{{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 }} 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 |bibcode=2020ITSMg..39c..13S |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}}

==== Automation levels ====

The SAE classification of Levels is based on the role of the driver, rather than the vehicle's capabilities, although these are related in the form of a "driving mode" (aka driving [[Scenario (vehicular automation)|scenario]]). The mode is determined by both an [[operational design domain]] (ODD) and a "dyanmic driving requirement." The ODD is the circumstance in which the car is driving, and the driving requirement is what the system must do while remaining safe within the boundaries of that ODD. These two things define the SAE Level.{{cite web |year=2014 |title=Automated Driving – Levels of Driving Automation are Defined in New SAE International Standard J3016 |url=https://cdn.oemoffhighway.com/files/base/acbm/ooh/document/2016/03/automated_driving.pdf |url-status=dead |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 |work=[[SAE International]]}}

Cars may therefore switch levels according to the driving mode. A human analogy might be a mode in which you can stand unassisted on one leg (the requirement) if you are on the ground (the ODD), but would need support when standing on a [[tightrope]].

Above Level 1, level differences are related to how responsibility for safe movement is divided/shared between the [[advanced driver-assistance system]] (ADAS) and driver, rather than specific driving features (see [[#Concepts|Concepts]]). A car might therefore have Level 3 capability for lane management (the driving requirement) up to 100 km/h (the ODD), but Level 2 at speeds faster than that. Or it might have Level 4 for driverless navigation on designated highways, but Level 2 on urban side roads.{{Cite journal |last1=Taeihagh |first1=Araz |last2=Lim |first2=Hazel Si Min |date=2 January 2019 |title=Governing autonomous vehicles: emerging responses for safety, liability, privacy, cybersecurity, and people risk |journal=Transport Reviews |volume=39 |issue=1 |pages=103–128 |arxiv=1807.05720 |doi=10.1080/01441647.2018.1494640 |issn=0144-1647 |s2cid=49862783}}

ADAS that are considered Level 1 are adaptive [[cruise control]], emergency [[Emergency brake assist|brake assist]], automatic emergency brake assist, [[Lane departure warning system|lane-keeping]], and lane centering. ADAS that are considered Level 2 are: highway assist, autonomous [[obstacle avoidance]], and autonomous parking. At Level 3 and above, if a car is driving and the ODD changes - for instance, a clear highway becomes a construction zone with hand-signals from a worker - the system must recognize that it is exiting its ODD and fall back accordingly, including asking for driver intervention if necessary.{{Cite news |date=2022-05-24 |title=ADAS Level 0 to Level 5: Autonomous/self-driving grades explained |url=https://timesofindia.indiatimes.com/auto/cars/adas-level-0-to-level-5-autonomous-cars-self-driving-level-2-kia-ev6-features-explained/articleshow/91762964.cms |access-date=2023-07-24 |work=The Times of India |issn=0971-8257}}

A fully self-driving car would therefore stay at Level 5 for all dynamic driving requirements in all circumstances (speeds, roads, terrains, visibilities, obstacle types, jurisdictions, etc.), while a car with no ADAS controlling the driving requirements would rely on the human driver for that, and stay at Level 0 regardless of the domain.{{Cite journal |last=Galvani |first=Marco |date=2019-02-04 |title=History and future of driver assistance |journal=IEEE Instrumentation & Measurement Magazine |volume=22 |issue=1 |pages=11–16 |bibcode=2019IIMM...22a..11G |doi=10.1109/MIM.2019.8633345 |issn=1941-0123 |s2cid=59600916}}{{Cite web |title=J3016_202104: Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles - SAE International |url=https://www.sae.org/standards/content/j3016_202104 |website=www.sae.org}} {| class="wikitable mw-collapsible" |+J3016 automation levels ! rowspan="2" |Mode ! rowspan="2" |Level ! rowspan="2" |Summary ! colspan="2" rowspan="2" |Description ! colspan="3" |Responsibility for |- !Direction & speed !Monitoring environment !Fallback |- style="background:#dff;" |n/a ! style="font-size:150%;background:#dff;" |0 |No Automation | colspan="2" |Full-time performance by the driver of all aspects of driving, even when "enhanced by warning or intervention systems" |Driver | rowspan="3" |Driver | rowspan="4" |Driver |- style="background:#bff;" | rowspan="3" |Some ! style="font-size:150%;background:#bff;" |1 |Driver Assistance |Driving mode-specific control by an ADAS of either steering or speed | rowspan="2" |ADAS uses information about the driving environment; driver is expected to perform all other driving tasks. |Driver and system |- style="background:#9ff;" ! style="font-size:150%;background:#9ff;" |2 |Partial Automation |Driving mode-specific execution by one or more ADAS for both steering and speed | rowspan="4" style="background:#dfd;" |System |- style="background:#dfd;" ! style="font-size:150%;background:#dfd;" |3 |Conditional Automation | rowspan="3" |Driving mode-specific control by an ADAS of all aspects of driving |Driver must appropriately respond to a request to intervene. | rowspan="3" |System |- style="background:#bfb;" |Many ! style="font-size:150%;background:#bfb;" |4 |High Automation |If a driver does not respond appropriately to a request to intervene, the car can stop safely. | rowspan="2" |System |- style="background:#9f9;" |All ! style="font-size:150%;background:#9f9;" |5 |Full Automation |System controls the vehicle under all conditions and circumstances. |}

=== Mobileye terminology === [[File:Mobileye Taxonomy for Consumer Autonomous Vehicles.png|thumb|upright=2|Mobileye taxonomy that explains the definitions of autonomous driving technology using the terms hands-on/off, eyes-on/off and no driver.]]

[[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 |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=General Motors}}{{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}}

==== Eyes-on/hands-on ==== 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 ==== 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 ==== 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., {{Convert|80|mph|abbr=on}} 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).

==== No driver ==== The highest level does not require a human driver in the car: monitoring is done either remotely (telepresence) or not at all.

==== Safety MRM ==== 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.

== Technology == {{Main|Vehicular automation}}

=== Architecture === 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 |journal=IEEE Transactions on Robotics |volume=37 |issue=6 |pages=1936–1955 |doi=10.1109/TRO.2021.3071615 |bibcode=2021ITRob..37.1936H }}{{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=[[European Technology Platform on Smart Systems Integration|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}}

Vendors such as Tesla and [[Motional]] have opted for monolithic "end to end" (E2E) neural networks, moving away from modular systems that separate perception from control.{{Cite web |last= |first= |date=2026-01-17 |title=How Motional's Robotaxi is Making Driverless Vehicles a Safe, Reliable and Accessible Reality |url=https://cleantechnica.com/2026/01/16/how-motionals-robotaxi-is-making-driverless-vehicles-a-safe-reliable-and-accessible-reality/ |access-date=2026-01-19 |website=CleanTechnica |language=en-US}}

=== Navigation === {{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 system]]s. 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.

==== Perception ==== ACs need to be able to perceive the world around them. Supporting technologies include combinations of cameras, [[Lidar|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 unit|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|bibcode=2007IJRR...26..519C |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 [[Bayes' theorem|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 |title=A Brief Survey on SLAM Methods in Autonomous Vehicle |url=https://www.researchgate.net/publication/329873195 |website=Research Gate}} 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 ==== [[Motion planning|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 |last=Shukla |first=Deepshikha |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 ==== 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 MIT[[MIT Computer Science and Artificial Intelligence Laboratory|Computer 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 ==== Sensors are necessary for the vehicle to properly respond to the driving environment. Sensor types include cameras, [[LiDAR]], [[ultrasound]], and [[radar]]. Control systems typically [[sensor fusion|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}}

=== Drive by wire === {{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.

=== Driver monitoring === {{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 |last=Dunoyer |first=Alain |date=27 January 2022 |title=Why driver monitoring will be critical to next-generation autonomous vehicles |url=https://www.sbdautomotive.com/en/news-insight-driver-monitoring |access-date=13 May 2022 |website=SBD Automotive |archive-date=31 May 2022 |archive-url=https://web.archive.org/web/20220531134243/https://www.sbdautomotive.com/en/news-insight-driver-monitoring |url-status=dead }} 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}}

=== Vehicle communication === {{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 |last=Beevor |first=Mike |date=11 April 2019 |title=Driving autonomous vehicles forward with intelligent infrastructure |url=https://www.smartcitiesworld.net/opinions/opinions/driving-autonomous-vehicles-forward-with-intelligent-infrastructure |access-date=27 April 2022 |website=Smart Cities World}}{{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=[[US National Highway Traffic Safety Administration|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 }}

[[International Organization for Standardization|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=[[International Organization for Standardization|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=[[International Organization for Standardization|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 |archive-date=22 November 2021 |archive-url=https://web.archive.org/web/20211122234007/https://www.connectedautomateddriving.eu/standards/standards-collection/ |url-status=dead }}

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

=== Software update === {{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 [[United Nations Economic Commission for Europe|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=[[United Nations Economic Commission for Europe|UNECE]] |access-date=20 March 2022 }}

=== Safety model === 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 |archive-date=18 July 2022 |archive-url=https://web.archive.org/web/20220718061845/https://sagroups.ieee.org/2846/ |url-status=dead }} 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 |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 }}

=== Notification === In accordance with the proposed SAE J3134 "ADS Marker Lamps" specification,{{Cite web |title=J3134 - Automated Driving System (ADS) Marker Lamp |url=https://www.sae.org/standards/j3134-automated-driving-system-ads-marker-lamp |access-date=2025-12-31 |website=www.sae.org}} 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.{{Cite web |date=2023-12-19 |title=Mercedes-Benz receives approvals for turquoise-colored automated driving marker lights in California and Nevada |url=http://media.mbusa.com/releases/mercedes-benz-receives-approvals-for-turquoise-colored-automated-driving-marker-lights-in-california-and-nevada |access-date=2025-12-31 |website=MBUSA Newsroom |language=en}}

As of 2023, the turquoise light had not been standardized in China or by the [[United Nations Economic Commission for Europe|UNECE]].{{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 === [[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}}

== Challenges == [[File:Autonomous Delivery Vehicle Pileup.jpg|thumb|Autonomous delivery vehicles stuck in one place by attempting to avoid one another]]

=== Obstacles === 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 |bibcode=2015ITITr..16..546P |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}}

=== Concerns ===

==== Deceptive marketing ==== 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 |last=Barry |first=Keith |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/ |access-date=13 April 2020 |website=[[Consumer Reports]]}} 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|archive-url=https://web.archive.org/web/20211210100939/http://www3.nhk.or.jp/news/html/20211210/k10013383761000.html|url-status=dead|archive-date=10 December 2021| 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 [[Mercedes-Benz E-Class (W213)|E-Class models]].{{cite news |last=Willems |first=Steph |date=28 July 2016 |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 |access-date=15 April 2022 |work=The Truth About Cars}} At that time, Mercedes-Benz rejected the claims and stopped its "self-driving car" ad campaign that had been running.{{cite news |last=Brown |first=Aaron |date=29 July 2016 |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 |access-date=15 April 2022 |work=The Drive}}{{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}}

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

==== Espionage ==== In 2018 and 2019, former Apple engineers were charged with stealing information related to Apple's self-driving car project.{{cite news |last=Chiu |first=Allyson |date=11 July 2018 |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/ |access-date=18 April 2022 |newspaper=[[The Washington Post]]}}{{cite news |last=Leswing |first=Kif |date=22 August 2022 |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 |access-date=23 August 2022 |publisher=[[CNBC]]}}{{cite news |last=O'Kane |first=Sean |date=30 January 2019 |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 |access-date=18 April 2022 |work=[[The Verge]]}} 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=[[United States Department of Justice|DOJ, US]] |access-date=14 June 2022 }}{{cite news |last=Benner |first=Katie |date=19 July 2021 |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 |access-date=14 June 2022 |newspaper=[[The New York Times]]}} China has prepared "the Provisions on Management of Automotive Data Security (Trial) to protect its own data".{{cite web |last1=Schaub |first1=Mark |first2=Atticus |last2=Zhao |first3=Mark |last3=Fu |date=24 August 2021 |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 |access-date=23 April 2022 |website=[[King & Wood Mallesons]]}}{{cite magazine |date=1 July 2022 |first=Justin |last=Ling |author-link=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 (magazine)|Wired]] |access-date=3 July 2022 }}

[[Cellular V2X|Cellular Vehicle-to-Everything]] technologies are based on [[Concerns over Chinese involvement in 5G wireless networks|5G wireless networks]].{{cite web |last=McLellan |first=Charles |date=4 November 2019 |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/ |access-date=8 May 2022 |publisher=[[ZDNet]]}} {{As of|November 2022}}, the [[United States Congress|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 (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}}

==== Driver communications ==== 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}}

==== Behavior prediction ==== 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=Tech Xplore}}

==== Handover ==== 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 }}

==== Trust ==== 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 }}

==== Economics ==== 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 US 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=Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |location=US |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=Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |location=US |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=Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |location=US |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=Bureau of Labor Statistics |publisher=Office of Occupational Statistics and Employment Projections |location=US |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=Bureau of Labor Statistics |location=US |access-date=24 April 2024}}

=== Ethical issues === {{See also|Machine ethics}}

==== Equity and inclusion ==== 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 {{!}} National Science Foundation |url=https://ncses.nsf.gov/pubs/nsf23315/report |access-date=2024-04-23 |website=ncses.nsf.gov}}

==== Pedestrian detection ==== 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}}

==== Rationale for liability ==== 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 |last1=Hevelke |first1=Alexander |last2=Nida-Rümelin |first2=Julian |year=2015 |title=Responsibility for Crashes of Autonomous Vehicles: An Ethical Analysis |journal=Sci Eng Ethics |volume=21 |issue=3 |pages=619–630 |doi=10.1007/s11948-014-9565-5 |pmc=4430591 |pmid=25027859}} Additionally, the infusion of [[Artificial intelligence|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}}

==== Trolley problem ==== 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 [[Deontological ethics|deontology]] (formal rules) and [[utilitarianism]] (harm reduction).{{Cite book |last1=Meyer |first1=G. |url={{google books |plainurl=y |id=TxmfDAAAQBAJ}} |title=Road vehicle automation |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|bibcode=2020ITEM...67..252K |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 |last1=Bonnefon |first1=Jean-François |last2=Shariff |first2=Azim |last3=Rahwan |first3=Iyad |year=2016 |title=The Social Dilemma of Autonomous Vehicles |journal=Science |volume=352 |issue=6293 |pages=1573–1576 |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.

==== Privacy ==== 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}}

=== Road infrastructure === 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 |last=McEvoy |first=Steve |date=26 January 2023 |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/ |access-date=5 April 2023 |website=Automotive World}} 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, [[East Japan Railway Company|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 {{Convert|60|km/h|abbr=on}}.{{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=[[East Japan Railway Company|JR East]] |access-date=5 April 2023 }}

== Testing == === Approaches === 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}}

=== 2010s and disengagements === [[File:Waymo self-driving car front view.gk.jpg|thumb|A prototype of [[Waymo]]'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 Department of Motor Vehicles|California DMV]] |access-date=24 April 2022 }} This is one measure of system robustness (ideally, the system should never disengage).{{cite news |last=Templeton |first=Brad |author-link=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 |access-date=24 April 2022 |work=[[Forbes]]}}

=== 2020s ===

==== Disengagement definitions ==== Reporting companies use varying definitions of what qualifies as a disengagement, and such definitions can change over time.{{cite news |last=Bellan |first=Rebecca |date=10 February 2022 |title=Despite a drop in how many companies are testing autonomous driving on California roads, miles driven are way up |url=https://techcrunch.com/2022/02/10/fewer-autonomous-vehicle-companies-in-california-drive-millions-more-miles-in-testing/ |access-date=25 April 2022 |work=[[TechCrunch]]}} Executives of self-driving car companies have criticized disengagements as a deceptive metric, because it does not consider varying road conditions.{{cite news |last=Zipper |first=David |date=8 December 2022 |title=Self-Driving Taxis Are Causing All Kinds of Trouble in San Francisco |url=https://slate.com/technology/2022/12/san-francisco-waymo-cruise-self-driving-cars-robotaxis.html |access-date=9 December 2022 |work=[[Slate (magazine)|Slate]]}}

==== Standards ==== In April 2021, [[World Forum for Harmonization of Vehicle Regulations|WP.29]] GRVA proposed a "Test Method for Automated Driving (NATM)".{{cite web|date=13 April 2021 |title=(GRVA) New Assessment/Test Method for Automated Driving (NATM) – Master Document |url=https://unece.org/transport/documents/2021/04/working-documents/grva-new-assessmenttest-method-automated-driving-natm |website=[[United Nations Economic Commission for Europe|UNECE]] |access-date=23 April 2022 }}

In October 2021, Europe's pilot test, L3Pilot, demonstrated ADAS for cars in [[Hamburg]], Germany, in conjunction with [[World Congress on Intelligent Transport Systems|ITS World Congress 2021]]. SAE Level 3 and 4 functions were tested on ordinary roads.{{cite web|date=15 October 2021 |title=L3Pilot: Joint European effort boosts automated driving |url=https://www.connectedautomateddriving.eu/blog/l3pilot-joint-european-effort-boosts-automated-driving/ |website=Connected Automated Driving |access-date=9 November 2021}}{{cite web |date=13 October 2021 |title=From the Final Event Week: On Motorways |url=https://l3pilot.eu/detail/news/from-the-final-event-week-on-motorways |website=L3Pilot |access-date=27 April 2022 |archive-date=27 April 2022 |archive-url=https://web.archive.org/web/20220427221346/https://l3pilot.eu/detail/news/from-the-final-event-week-on-motorways |url-status=dead }}{{cite web |date=28 February 2022 |title=L3Pilot Final Project Results published |url=https://l3pilot.eu/detail/news/l3pilot-final-project-results-published |website=L3Pilot |access-date=27 April 2022 |archive-date=22 May 2022 |archive-url=https://web.archive.org/web/20220522015155/https://l3pilot.eu/detail/news/l3pilot-final-project-results-published |url-status=dead }}

In November 2022, an International Standard ISO 34502 on "[[Scenario (vehicular automation)|Scenario]] based safety evaluation framework" was published.{{cite web|date=November 2022 |title=ISO 34502:2022 Road vehicles — Test scenarios for automated driving systems — Scenario based safety evaluation framework |url=https://www.iso.org/standard/78951.html |website=[[International Organization for Standardization|ISO]] |access-date=17 November 2022 }}{{cite web|date=16 November 2022 |title=New International Standard Issued for the Scenario-Based Safety Evaluation Framework for Automated Driving Systems Formulated by Japan |url=https://www.meti.go.jp/english/press/2022/1116_003.html |website=[[Ministry of Economy, Trade and Industry|METI, Japan]] |access-date=14 December 2022 }}

==== Collision avoidance ==== In April 2022, collision avoidance testing was demonstrated by [[Nissan]].{{cite web|author= |title=New driver-assistance technology dramatically improves collision-avoidance performance |url=https://www.nissan-global.com/EN/INNOVATION/TECHNOLOGY/ARCHIVE/ADVANCED_DRIVER_ASSISTANCE_TECH/ |website=[[Nissan]] |access-date=15 December 2022 }}{{cite web |last=Hope |first=Graham |date=26 April 2022 |title=Nissan Tests Collision Avoidance Tech for Self-Driving Cars |url=https://www.iotworldtoday.com/2022/04/26/nissan-tests-collision-avoidance-tech-for-self-driving-cars/ |access-date=15 December 2022 |website=IoT World Today}} [[Waymo]] published a document about collision avoidance testing in December 2022.{{cite web|author= |date=14 December 2022 |title=Waymo's Collision Avoidance Testing: Evaluating our Driver's Ability to Avoid Crashes Compared to Humans |url=https://blog.waymo.com/2022/12/waymos-collision-avoidance-testing.html |website=[[Waymo]] |access-date=15 December 2022 }}

==== Simulation and validation ==== In September 2022, Biprogy released Driving Intelligence Validation Platform (DIVP) as part of Japanese national project "SIP-adus", which is interoperable with Open Simulation Interface (OSI) of [[Association for Standardisation of Automation and Measuring Systems|ASAM]].{{cite web |date=6 September 2022 |title=SIP自動運転の成果を活用した安全性評価用シミュレーションソフトの製品化～戦略的イノベーション創造プログラム（SIP）研究成果を社会実装へ～ |trans-title=Commercial product of the achievement of SIP-adus: Driving Intelligence Validation Platform |url=https://www8.cao.go.jp/cstp/stmain/20220906divp.html |access-date=10 September 2022 |website=[[Cabinet Office (Japan)|Cabinet Office, Japan]] |language=Japanese}}{{cite web|title=DIVP |url=https://divp.net/ |website=DVIP |access-date=10 September 2022 }}{{cite web |last=Kuzumaki |first=Seigo |title=Development of "Driving Intelligence Validation Platform" for ADS safety assurance |url=https://en.sip-adus.go.jp/wp/wp-content/uploads/presentation_material2021.pdf |access-date=12 September 2022 |website=SIP-adus}}

==== Toyota ==== In November 2022, [[Toyota]] demonstrated one of [[Toyota GR Yaris|its GR Yaris]] test cars, which had been trained using professional rally drivers.{{cite news|date=17 November 2021 |author= |title=Toyota pushes AI to drive like pros |url=https://japannews.yomiuri.co.jp/business/companies/20221117-71380/ |work=[[Yomiuri Shimbun]] |access-date=20 November 2022 }} Toyota used its collaboration with [[Microsoft]] in FIA [[World Rally Championship]] since the 2017 season.{{cite web|date=20 September 2016 |title=Microsoft and Toyota Join Forces in FIA World Rally Championship |url=https://global.toyota/en/detail/13770530 |website=[[Toyota]]l |access-date=20 November 2022 }}

==== Pedestrian reactions ==== In 2023 David R. Large, senior research fellow with the Human Factors Research Group at the [[University of Nottingham]], disguised himself as a car seat in a study to test people's reactions to driverless cars. He said, "We wanted to explore how pedestrians would interact with a driverless car and developed this unique methodology to explore their reactions." The study found that, in the absence of someone in the driving seat, pedestrians trust certain visual prompts more than others when deciding whether to cross the road.{{cite news | url=https://www.bbc.co.uk/news/av/uk-england-nottinghamshire-65857348 | title=Driver disguises himself as car seat for study | work=BBC News }}

== Safety == A [[meta-analysis]] published in [[Nature Communications]] in 2024 compared various sources of safety data for autonomous (AV) and human-driven vehicles (HDV). This collected 2,100 AV and 35,133 HDV incident records which accurately reflected the incident details.{{Cite journal |last1=Abdel-Aty |first1=Mohamed |last2=Ding |first2=Shengxuan |date=2024-06-18 |title=A matched case-control analysis of autonomous vs human-driven vehicle accidents |url=https://www.nature.com/articles/s41467-024-48526-4 |journal=Nature Communications |language=en |volume=15 |issue=1 |pages=4931 |doi=10.1038/s41467-024-48526-4 |pmid=38890354 |bibcode=2024NatCo..15.4931A |issn=2041-1723|pmc=11189485 }} Some AVs in the comparisons (such as [[robotaxi]]s) were effectively autonomous, while others were equipped with Advanced Driving Systems (ADS) or Advanced Driver Assistance Systems (ADAS).

The study concluded that AVs are safer in most circumstances,{{Cite journal |last1=Abdel-Aty |first1=Mohamed |last2=Ding |first2=Shengxuan |date=2024-06-18 |title=A matched case-control analysis of autonomous vs human-driven vehicle accidents - Introduction |url=https://www.nature.com/articles/s41467-024-48526-4 |journal=Nature Communications |language=en |volume=15 |issue=1 |pages=4931 |doi=10.1038/s41467-024-48526-4 |bibcode=2024NatCo..15.4931A |issn=2041-1723|pmc=11189485 }} and AVs had far fewer crashes involving pedestrians (3% against 15%) per mile travelled.{{Cite journal |last1=Abdel-Aty |first1=Mohamed |last2=Ding |first2=Shengxuan |date=2024-06-18 |title=A matched case-control analysis of autonomous vs human-driven vehicle accidents - Results |url=https://www.nature.com/articles/s41467-024-48526-4 |journal=Nature Communications |language=en |volume=15 |issue=1 |pages=4931 |doi=10.1038/s41467-024-48526-4 |pmid=38890354 |bibcode=2024NatCo..15.4931A |issn=2041-1723|pmc=11189485 }} The comparison also highlighed some other disaparities: while AVs were found to be significantly less likely to crash in heavy rain or fog than an unassisted human, they were more than five times more vulnerable to collisions at dawn and dusk.

=== Incidents ===

==== Tesla ==== {{See also|List of Tesla Autopilot crashes}}

As of 2023, Tesla's ADAS [[Tesla Autopilot|Autopilot]]/Full Self Driving (beta) was classified as Level 2 ADAS.{{Cite web |last=Mulac |first=Jordan h |date=2023-02-28 |title=Tesla admits its semi-autonomous driving tech is not the world's most advanced |url=https://www.drive.com.au/news/tesla-full-self-driving-level-2-autonomous/ |access-date=2024-02-02 |website=Drive |language=en-AU}}

On 20 January 2016, the first of five known fatal crashes of a Tesla with Autopilot occurred, in China's Hubei province.{{cite web|title=Tesla Fatalities Dataset|url=https://datasetsearch.research.google.com/search?query=tesla%20deaths&docid=QYBSmRGFh9iXZv8xAAAAAA==|access-date=17 October 2020}} Initially, Tesla stated that the vehicle was so badly damaged from the impact that their recorder was not able to determine whether the car had been on Autopilot at the time. However, the car failed to take evasive action.

Another fatal Autopilot crash occurred in May in Florida in a Tesla Model S{{cite news|url=https://qz.com/783009/the-scary-similarities-between-teslas-tsla-deadly-autopilot-crashes|title=There are some scary similarities between Tesla's deadly crashes linked to Autopilot|work=Quartz|first1=Josh|last1=Horwitz|first2=Heather|last2=Timmons|date=20 September 2016|access-date=19 March 2018}}{{cite web|url=http://c.m.163.com/news/a/C0UBU2I0002580S6.html|title=China's first accidental death due to Tesla's automatic driving: not hitting the front bumper|language=zh|work=[[Media of China|China State Media]]|date=14 September 2016|access-date=18 March 2018}} that crashed into a [[tractor-trailer]]. In a civil suit between the father of the driver killed and Tesla, Tesla documented that the car had been on Autopilot.{{cite web|url=https://jalopnik.com/two-years-on-a-father-is-still-fighting-tesla-over-aut-1823189786|title=Two Years On, A Father Is Still Fighting Tesla Over Autopilot And His Son's Fatal Crash|first=Ryan|last=Felton|website=Jalopnik |date=27 February 2018|access-date=18 March 2018}} According to Tesla, "neither Autopilot nor the driver noticed the white side of the tractor-trailer against a brightly lit sky, so the brake was not applied." Tesla claimed that this was Tesla's first known Autopilot death in over {{convert|130|e6mi|e6km|abbr=off|sp=us}} with Autopilot engaged. Tesla claimed that on average one fatality occurs every {{convert|94|e6mi|e6km|abbr=off|sp=us}} across all vehicle types in the US.{{cite news |last1=Yadron |first1=Danny |last2=Tynan |first2=Dan |date=1 July 2016 |title=Tesla driver dies in first fatal crash while using autopilot mode |url=https://www.theguardian.com/technology/2016/jun/30/tesla-autopilot-death-self-driving-car-elon-musk |access-date=1 July 2016 |work=[[The Guardian]] |location=San Francisco}}{{cite news |last1=Vlasic |first1=Bill |last2=Boudette |first2=Neal E. |date=30 June 2016 |title=Self-Driving Tesla Involved in Fatal Crash |url=https://www.nytimes.com/2016/07/01/business/self-driving-tesla-fatal-crash-investigation.html |access-date=1 July 2016 |work=[[The New York Times]]}}{{cite press release|title=A Tragic Loss|date=30 June 2016|publisher=[[Tesla Motors]]|url=https://www.teslamotors.com/blog/tragic-loss|quote=This is the first known fatality in just over 130 million miles where Autopilot was activated. Among all vehicles in the US, there is a fatality every 94 million miles. Worldwide, there is a fatality approximately every 60 million miles.|access-date=1 July 2016}} However, this number also includes motorcycle/pedestrian fatalities.{{cite web|url=https://www.forbes.com/sites/samabuelsamid/2016/07/05/adding-some-statistical-perspective-to-tesla-autopilot-safety-claims|title=Adding Some Statistical Perspective To Tesla Autopilot Safety Claims|last=Abuelsamid|first=Sam|website=[[Forbes]]}}{{cite web|url=https://www-fars.nhtsa.dot.gov/Main/index.aspx|title=FARS Encyclopedia|last=Administration|first=National Highway Traffic Safety}} The ultimate [[National Transportation Safety Board]] (NTSB) report concluded Tesla was not at fault; the investigation revealed that for Tesla cars, the crash rate dropped by 40 percent after Autopilot was installed.{{cite web|url=https://www.theverge.com/2017/1/19/14323990/tesla-autopilot-fatal-accident-nhtsa-investigation-ends|title=Fatal Tesla Autopilot accident investigation ends with no recall ordered|date=19 January 2016|access-date=19 January 2017|website=[[The Verge]]}}

In February 2025, a Tesla Cybertruck crashed while in Full-Self Driving mode, raising concerns about autonomous driving and prompting an investigation from Tesla who said the crash would be probed "“in line with standard protocol when any of our electric vehicles are involved in an accident while in FSD mode.”{{Cite web |last1=McAndrew |first1=Siobhan |last2=Cervantes Jr. |first2=Fernando |name-list-style=and |title=Tesla Cybertruck crash into a pole in Nevada was in self-driving mode: owner |url=https://www.usatoday.com/story/news/nation/2025/02/14/tesla-cybertruck-crash-in-reno-nevada-self-driving-mode/78643113007/ |access-date=2025-02-15 |website=USA TODAY |language=en-US}}{{Cite web |last=Shepardson |first=David |date=2025-02-15 |title=Tesla Launches Investigation Into Cybertruck FSD Crash Ahead of Robotaxi Launch |url=https://londoninsider.co.uk/tesla-launches-investigation-into-cybertruck-fsd-crash-ahead-of-robotaxi-launch/ |access-date=2025-02-15 |website=London Insider |language=en-US}}

==== Google Waymo ==== {{main|Waymo#Accidents}}

[[File:Google driverless car at intersection.gk.jpg|thumb|Google's in-house [[Waymo|automated car]]]]

In June 2015, Google confirmed that 12 vehicles had suffered collisions as of that date. Eight involved rear-end collisions at a stop sign or traffic light, in two of which the vehicle was side-swiped by another driver, one in which another driver rolled a stop sign, and one where a driver was controlling the car manually.{{cite news|url=https://www.latimes.com/business/la-fi-google-cars-20150603-story.html|title=Google founder defends accident records of self-driving cars|date=3 June 2015|agency=[[Associated Press]]|newspaper=[[Los Angeles Times]]|access-date=1 July 2016}} In July 2015, three employees suffered minor injuries when their vehicle was rear-ended by a car whose driver failed to brake. This was the first collision that resulted in injuries.{{cite web|url=http://www.govtech.com/transportation/Google-Autonomous-Car-Experiences-Another-Crash.html|title=Google Autonomous Car Experiences Another Crash|first=Vishal|last=Mathur|date=17 July 2015|access-date=18 July 2015|newspaper=[[Government Technology]]}}

According to Google Waymo's accident reports as of early 2016, their test cars had been involved in 14 collisions, of which other drivers were at fault 13 times, although in 2016 the car's software caused a crash.{{cite news |date=29 February 2016 |title=For the first time, Google's self-driving car takes some blame for a crash |url=https://www.washingtonpost.com/news/innovations/wp/2016/02/29/for-the-first-time-googles-self-driving-car-takes-some-blame-for-a-crash/ |newspaper=The Washington Post}} On 14 February 2016 a Google vehicle attempted to avoid sandbags blocking its path. During the maneuver it struck a bus. Google stated, "In this case, we clearly bear some responsibility, because if our car hadn't moved, there wouldn't have been a collision."{{cite magazine|url=https://www.wired.com/2016/02/googles-self-driving-car-may-caused-first-crash/|title=Google's Self-Driving Car Caused Its First Crash|magazine=Wired|date=February 2016}}{{cite news|url=https://www.latimes.com/local/lanow/la-me-ln-google-self-driving-car-bus-collision-20160229-story.html|title=Passenger bus teaches Google robot car a lesson|date=29 February 2016|newspaper=Los Angeles Times}} Google characterized the crash as a misunderstanding and a learning experience. No injuries were reported.

==== Uber's Advanced Technologies Group (ATG) ==== In March 2018, [[Death of Elaine Herzberg|Elaine Herzberg]] died after she was hit by an AC tested by [[Uber|Uber's]] Advanced Technologies Group (ATG) in Arizona. A safety driver was in the car. Herzberg was crossing the road about 400 feet from an intersection.{{cite news|last1=Bensinger|first1=Greg|last2=Higgins|first2=Tim|title=Video Shows Moments Before Uber Robot Car Rammed into Pedestrian|url=https://www.wsj.com/articles/video-shows-final-seconds-before-fatal-uber-self-driving-car-crash-1521673182|access-date=25 March 2018|work=[[The Wall Street Journal]]|date=22 March 2018}} Some experts said a human driver could have avoided the crash.{{cite news|url=https://www.bloomberg.com/news/articles/2018-03-22/video-said-to-show-failure-of-uber-s-tech-in-fatal-arizona-crash|title=Human Driver Could Have Avoided Fatal Uber Crash, Experts Say|newspaper=Bloomberg |date=22 March 2018}} Arizona governor [[Doug Ducey]] suspended the company's ability to test its ACs citing an "unquestionable failure" of Uber to protect public safety.{{cite news|agency=Associated Press|title=Governor Ducey suspends Uber from automated vehicle testing|url=https://www.abc15.com/news/region-phoenix-metro/central-phoenix/governor-ducey-suspends-uber-from-autonomous-vehicle-testing|access-date=27 March 2018|publisher=[[KNXV-TV]]|date=27 March 2018}} Uber also stopped testing in California until receiving a new permit in 2020.{{cite news | url=https://www.sfchronicle.com/business/article/Uber-pulls-out-of-all-self-driving-car-testing-in-12785490.php | title=Uber puts the brakes on testing robot cars in California after Arizona fatality | first=Carolyn | last=Said | date=27 March 2018 | newspaper=San Francisco Chronicle | access-date=8 April 2018 }}{{Cite news |date=5 February 2020 |title=Uber self-driving cars allowed back on California roads |language=en-GB |publisher=BBC News |url=https://www.bbc.com/news/technology-51393808 |access-date=24 October 2022}}

The [[National Transportation Safety Board|NTSB's]] final report determined that the immediate cause of the accident was that safety driver Rafaela Vasquez failed to monitor the road, because she was distracted by her phone, but that Uber's "inadequate safety culture" contributed. The report noted that the victim had "a very high level" of [[methamphetamine]] in her body.{{Cite news |date=19 November 2019 |title=Uber back-up driver faulted in fatal autonomous car crash |work=Financial Times |url=https://www.ft.com/content/6d0c5544-0afb-11ea-bb52-34c8d9dc6d84 |access-date=24 October 2022}} The board called on federal regulators to carry out a review before allowing automated test vehicles to operate on public roads.{{Cite web |title='Inadequate Safety Culture' Contributed to Uber Automated Test Vehicle Crash – NTSB Calls for Federal Review Process for Automated Vehicle Testing on Public Roads |url=https://www.ntsb.gov/news/press-releases/Pages/NR20191119c.aspx |access-date=24 October 2022 |website=ntsb.gov}}{{Cite magazine |last=Smiley |first=Lauren |title='I'm the Operator': The Aftermath of a Self-Driving Tragedy |language=en-US |magazine=Wired |url=https://www.wired.com/story/uber-self-driving-car-fatal-crash/ |access-date=24 October 2022 |issn=1059-1028}}

In September 2020, Vasquez pled guilty to endangerment and was sentenced to three years' probation.{{Cite web |last=Vanek |first=Corina |date=July 21, 2023 |title=Arizona driver in fatal autonomous Uber crash in 2018 pleads guilty, sentenced to probation |url=https://www.azcentral.com/story/news/local/tempe/2023/07/28/rafaela-vasquez-pleads-guilty-in-in-fatal-uber-self-driving-crash-killed-pedestrian-elaine-herzberg/70488361007/ |access-date=2024-02-02 |website=The Arizona Republic |language=en-US}}{{cite news |last1=Billeaud |first1=Jacques |last2=Snow |first2=Anita |date=2023-07-28 |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 |access-date=2024-09-01 |work=Associated Press}}

==== NIO Navigate on Pilot ==== On 12 August 2021, a 31-year-old Chinese man was killed after his [[NIO ES8]] crashed in a tunnel.{{Cite web |last=Bobylev |first=Denis |date=2025-09-08 |title=Third-gen Nio ES8 crashed in China ahead of the sales start |url=https://carnewschina.com/2025/09/08/third-gen-nio-es8-crashed-in-china-ahead-of-the-sales-start/ |access-date=2025-12-28 |website=Car News China |language=en-US}} NIO's self-driving feature was in beta and could not deal with static obstacles.{{cite web |last1=Rearick |first1=Brenden |date=16 August 2021 |title=NIO Stock: 10 Things to Know About the Fatal Crash Dragging Down Nio Today |url=https://investorplace.com/2021/08/nio-stock-10-things-to-know-about-the-fatal-crash-dragging-down-nio-today/ |access-date=17 February 2022 |website=InvestorPlace |language=en-US}} The vehicle's manual clearly stated that the driver must take over near construction sites. Lawyers of the deceased's family questioned NIO's private access to the vehicle, which they argued did not guarantee the integrity of the data.{{cite web |last=Ruffo |first=Gustavo Henrique |date=17 August 2021 |title=Nio's Autopilot, NOP, Faces Intense Scrutiny With First Fatal Crash in China |url=https://www.autoevolution.com/news/nio-s-autopilot-nop-faces-intense-scrutiny-with-first-fatal-crash-in-china-167486.html |access-date=17 February 2022 |website=autoevolution |language=en}}

==== Pony.ai ==== In November 2021, the [[California Department of Motor Vehicles]] (DMV) notified [[Pony.ai]] that it was suspending its testing permit following a reported collision in Fremont on 28 October.{{cite news |last=Liao |first=Rita |date=14 December 2021 |title=California suspends Pony.ai driverless test permit after crash |url=https://techcrunch.com/2021/12/14/pony-ai-suspension-driverless-pilot-california/ |access-date=23 April 2022 |work=[[TechCrunch]]}} In May 2022, DMV revoked Pony.ai's permit for failing to monitor the driving records of its safety drivers.{{cite news |last=Bellan |first=Rebecca |date=25 May 2022 |title=Pony.ai loses permit to test autonomous vehicles with driver in California |url=https://techcrunch.com/2022/05/24/pony-ai-loses-permit-to-test-autonomous-vehicles-with-driver-in-california/ |access-date=30 May 2022 |work=[[TechCrunch]]}}

==== Cruise ==== {{See also|Cruise (autonomous vehicle)#Incidents}} In April 2022, [[Cruise (autonomous vehicle)|Cruise]]'s testing vehicle was reported to have blocked a [[fire engine]] on emergency call, and sparked questions about its ability to handle unexpected circumstances.{{cite magazine |last=Marshall |first=Aarian |date=27 May 2022 |title=An Autonomous Car Blocked a Fire Truck Responding to an Emergency |url=https://www.wired.com/story/cruise-fire-truck-block-san-francisco-autonomous-vehicles/ |access-date=30 May 2022 |magazine=[[Wired (magazine)|Wired]]}}{{cite news |last=Hope |first=Graham |date=29 May 2022 |title=GM's Cruise Autonomous Car Blocks Fire Truck on Emergency Call |url=https://www.iotworldtoday.com/2022/05/29/gms-cruise-autonomous-car-blocks-fire-truck-on-emergency-call/ |access-date=30 May 2022 |work=IoT World Today}}

==== Ford ==== In February 2024, a driver using the Ford BlueCruise hands-free driving feature struck and killed the driver of a stationary car with no lights on in the middle lane of a freeway in Texas.{{Cite web |last=Moreno |first=Julie |date=2024-03-15 |title=NTSB investigating deadly crash in San Antonio involving a semi-autonomous vehicle |url=https://www.ksat.com/news/local/2024/03/15/ntsb-investigating-deadly-crash-in-san-antonio-involving-a-semi-autonomous-vehicle/ |access-date=2024-09-08 |website=KSAT |language=en}}

In March 2024, a drunk driver who was speeding, holding her cell phone, and using BlueCruise on a Pennsylvania freeway struck and killed two people who had been driving two cars.{{Cite web |last=Torrejón |first=Rodrigo |date=2024-09-03 |title=Woman who used hands-free driving system in fatal DUI crash on I-95 turns herself in to face charges, police say |url=https://www.inquirer.com/crime/dimple-patel-dui-crash-195-philadelphia-charges-20240903.html |access-date=2024-09-08 |website=Philadelphia Inquirer |language=en}} The first car had become disabled and was on the left shoulder with part of the car in the left driving lane. The second driver had parked his car behind the first car presumably to help the first driver. The [[National Transportation Safety Board|NTSB]] is investigating both incidents.{{Cite web |last=Bell |first=Sebastien |date=March 18, 2024 |title=Feds Investigate Ford BlueCruise's Role In Deadly Mustang Mach-E Crash |url=https://www.carscoops.com/2024/03/feds-investigate-bluecruises-involvement-in-fatal-ford-mustang-mach-e-crash/ |access-date=September 8, 2024 |website=Carscoops}}

==== Total incidents ==== The US [[National Highway Traffic Safety Administration]] began mandating incident reports from autonomous vehicle companies in June 2021. Some reports cite incidents from as early as August 2019, with current data available through June 17, 2024.{{Cite web |title=Standing General Order on Crash Reporting {{!}} NHTSA |url=https://www.nhtsa.gov/laws-regulations/standing-general-order-crash-reporting |access-date=2024-08-14 |website=www.nhtsa.gov |language=en}}

There have been a total of 3,979 autonomous vehicle incidents (both ADS and ADAS) reported during this timeframe. 2,146 of those incidents (53.9%) involved Tesla vehicles.{{Cite web |title=Data Analysis: Self-Driving Car Accidents [2019-2024] |url=https://www.craftlawfirm.com/autonomous-vehicle-accidents-2019-2024-crash-data/ |access-date=2024-08-14 |website=Craft Law Firm |language=en-US}}

== Public opinion surveys ==

=== 2010s === In a 2011 online survey of 2,006 US and UK consumers, 49% said they would be comfortable using a "driverless car".{{cite web|url=http://newsroom.accenture.com/article_display.cfm?article_id=5146|title=Consumers in US and UK Frustrated with Intelligent Devices That Frequently Crash or Freeze, New Accenture Survey Finds|date=10 October 2011|publisher=Accenture|access-date=30 June 2013}}

A 2012 survey of 17,400 vehicle owners found 37% who initially said they would be interested in purchasing a "fully autonomous car". However, that figure dropped to 20% if told the technology would cost US$3,000 more.{{cite web|url=http://reviews.cnet.com/8301-13746_7-57422698-48/many-car-buyers-show-interest-in-autonomous-car-tech/|title=Many car buyers show interest in autonomous car tech|last=Yvkoff|first=Liane|date=27 April 2012|publisher=CNET|access-date=30 June 2013}}

In a 2012 survey of about 1,000 German drivers, 22% had a positive attitude, 10% were undecided, 44% were skeptical and 24% were hostile.{{cite web|url=http://www.motorvision.de/service-ratgeber/selbstfahrende-autos-deutschland-22-prozent-deutschen-grosse-akzeptanz-roboterfahrzeuge-39281.html|archive-url=http://arquivo.pt/wayback/20160515125001/http://www.motorvision.de/service-ratgeber/selbstfahrende-autos-deutschland-22-prozent-deutschen-grosse-akzeptanz-roboterfahrzeuge-39281.html|url-status=dead|archive-date=15 May 2016|title=Große Akzeptanz für selbstfahrende Autos in Deutschland|date=9 October 2012|publisher=motorvision.de|access-date=6 September 2013}}

A 2013 survey of 1,500 consumers across 10 countries found 57% "stated they would be likely to ride in a car controlled entirely by technology that does not require a human driver", with Brazil, India and China the most willing to trust automated technology.{{cite web|url=http://www.autosphere.ca/collisionmanagement/2013/05/22/autonomous-cars-found-trustworthy-in-global-study/|title=Autonomous Cars Found Trustworthy in Global Study|date=22 May 2013|publisher=autosphere.ca|access-date=6 September 2013}}

In a 2014 US telephone survey, over three-quarters of licensed drivers said they would consider buying a self-driving car, rising to 86% if car insurance were cheaper. 31.7% said they would not continue to drive once an automated car was available.{{cite web|url=http://www.insurance.com/about-us/news-and-events/2014/07/autonomous-cars-bring-em-on-drivers-say-in-insurance.com-survey.html|title=Autonomous cars: Bring 'em on, drivers say in Insurance.com survey|date=28 July 2014|website=Insurance.com|access-date=29 July 2014}}

In 2015, a survey of 5,000 people from 109 countries reported that average respondents found manual driving the most enjoyable. 22% did not want to pay more money for autonomy. Respondents were found to be most concerned about hacking/misuse, and were also concerned about legal issues and safety. Finally, respondents from more developed countries were less comfortable with their vehicle sharing data.{{cite journal|last1=Kyriakidis|first1=M.|last2=Happee|first2=R.|last3=De Winter|first3=J. C. F.|year=2015|title=Public opinion on automated driving: Results of an international questionnaire among 5,000 respondents|journal=Transportation Research Part F: Traffic Psychology and Behaviour|volume=32|pages=127–140|doi=10.1016/j.trf.2015.04.014|bibcode=2015TRPF...32..127K |s2cid=2071964 |url=https://repository.tudelft.nl/islandora/object/uuid%3Aa779b2bc-dcca-4f8c-b8e0-8209791a3146/datastream/OBJ/download}} The survey reported consumer interest in purchasing an AC, stating that 37% of surveyed current owners were either "definitely" or "probably" interested.

In 2016, a survey of 1,603 people in Germany that controlled for age, gender, and education reported that men felt less anxiety and more enthusiasm, whereas women showed the opposite. The difference was pronounced between young men and women and decreased with age.{{cite journal|last1=Hohenberger|first1=C.|last2=Spörrle|first2=M.|last3=Welpe|first3=I. M.|year=2016|title=How and why do men and women differ in their willingness to use automated cars? The influence of emotions across different age groups|journal=Transportation Research Part A: Policy and Practice|volume=94|pages=374–385|doi=10.1016/j.tra.2016.09.022|bibcode=2016TRPA...94..374H |url=http://mediatum.ub.tum.de/node?id=1340486 }}

In a 2016 US survey of 1,584 people, "66 percent of respondents said they think autonomous cars are probably smarter than the average human driver". People were worried about safety and hacking risk. Nevertheless, only 13% of the interviewees saw no advantages in this new kind of cars.{{cite news|url=https://techcrunch.com/2016/12/22/autonomous-cars-seen-as-smarter-than-human-drivers/|title=Autonomous cars seen as smarter than human drivers|first=Kristen|last=Hall-Geisler|work=[[TechCrunch]]|date=22 December 2016|access-date=26 December 2016}}

In a 2017 survey of 4,135 US adults found that many Americans anticipated significant impacts from various automation technologies including the widespread adoption of automated vehicles.{{cite web|url=http://www.pewinternet.org/2017/10/04/automation-in-everyday-life/|title=Automation in Everyday Life|last1=Smith|first1=Aaron|last2=Anderson|first2=Monica |date=4 October 2017}}

In 2019, a test of a method of surveying opinions about self-driving cars, termed the Autonomous Vehicle Acceptance Model (AVAM), found that users were less accepting of high autonomy levels and displayed significantly lower intention to use highly autonomous vehicles. Additionally, partial autonomy (regardless of level) was perceived as requiring uniformly higher driver engagement (usage of hands, feet and eyes) than full autonomy.{{cite book |last1=Hewitt |first1=Charlie |title=Proceedings of the 24th International Conference on Intelligent User Interfaces |last2=Politis |first2=Ioannis |last3=Amanatidis |first3=Theocharis |last4=Sarkar |first4=Advait |publisher=ACM Press |year=2019 |isbn=9781450362726 |pages=518–527 |chapter=Assessing public perception of self-driving cars: The autonomous vehicle acceptance model |doi=10.1145/3301275.3302268 |s2cid=67773581}}

=== In the 2020s === In 2022, a survey reported that only a quarter (27%) of the world's population would feel safe in self-driving cars.{{cite web|date=25 November 2022 |title=Majority of world's population feel self-driving cars are unsafe |url=https://www.lrfoundation.org.uk/en/news/majority-of-worlds-population-feel-self-driving-cars-are-unsafe/ |website=[[Lloyd's Register|Lloyd's Register Foundation]] |access-date=4 December 2022 }}

In 2024, a study by Saravanos et al.{{Cite journal |last1=Saravanos |first1=Antonios |last2=Pissadaki |first2=Eleftheria K. |last3=Singh |first3=Wayne S. |last4=Delfino |first4=Donatella |date=April 2024 |title=Gauging Public Acceptance of Conditionally Automated Vehicles in the United States |journal=Smart Cities |language=en |volume=7 |issue=2 |pages=913–931 |doi=10.3390/smartcities7020038 |doi-access=free |issn=2624-6511|arxiv=2402.11444 }} at New York University reported that 87% of their respondents (from a sample of 358) believed that conditionally automated cars (at Level 3) would be easy to use.

In July 2024, a biannual [[YouGov]] survey asking British adults how safe or unsafe people would feel in a self-driving car found that 4% would feel "Very safe" while 37% would feel "Very unsafe".{{Cite web |title=How safe Brits say they would feel in a driverless car |url=https://yougov.co.uk/topics/travel/trackers/how-safe-brits-say-they-would-feel-in-a-driverless-car |access-date=2025-12-31 |website=YouGov |language=en-gb}}

== Regulation == {{Main|Regulation of self-driving cars}}

{{See also|Regulation of algorithms}}{{Excerpt|Regulation of self-driving cars|only=paragraphs|hat=no}}

== Commercialization == {{See also|History of self-driving cars#2020s}}

While various models of car may be described by their manufacturers as being at a certain Level, the SAE technical specification means that a car can move between levels depending on the driving task and the circumstances it is operating in at a given time. This means that, for example, when a manufacturer says they have a "Level 4 car," they mean the vehicle has one or more specific ''features'' (like [[Automatic parking#Automated Valet Parking|automated valet parking]]) that are capable of operating at Level 4, but on the open highway they might operate at a lower Level.

{{As of|2023}} most commercially available ADAS vehicles are SAE Level 2. A couple of companies reached higher levels, but only in restricted ([[geofence]]d) locations.{{Cite web |date=2023-03-03 |title=Self-Driving Cars: Everything You Need To Know |url=https://www.kbb.com/car-advice/self-driving-cars/ |access-date=2023-04-09 |website=Kelley Blue Book |language=en-US}} Vehicles operating below Level 5 still offer many advantages.{{Cite journal |last1=Hancock |first1=P. A. |last2=Nourbakhsh |first2=Illah |last3=Stewart |first3=Jack |date=16 April 2019 |title=On the future of transportation in an era of automated and autonomous vehicles |journal=Proceedings of the National Academy of Sciences of the United States of America |volume=116 |issue=16 |pages=7684–7691 |bibcode=2019PNAS..116.7684H |doi=10.1073/pnas.1805770115 |issn=0027-8424 |pmc=6475395 |pmid=30642956 |doi-access=free}}

=== Level 2 – Partial automation === {{See also|Lane centering#Sample of level 2 automated cars|List of self-driving system suppliers#Date of first public road driverless operation}}

SAE Level 2 features are available as part of the ADAS systems in many vehicles. In the US, 50% of new cars provide driver assistance for both steering and speed.Jeff S. Bartlett. How Much Automation Does Your Car Really Have? November 4, 2021, https://www.consumerreports.org/cars/automotive-technology/how-much-automation-does-your-car-really-have-level-2-a3543419955/.

[[Ford Motor Company|Ford]] started offering BlueCruise service on certain vehicles in 2022; the system is named ActiveGlide in [[Lincoln Motor Company|Lincoln]] vehicles. The system provided features such as lane centering, street sign recognition, and hands-free highway driving on more than 130,000 miles of divided highways. The 2022 1.2 version added features including hands-free lane changing, in-lane repositioning, and predictive speed assist.{{Cite web |date=15 March 2023 |title=Ford BlueCruise Version 1.2 Hands-Off Review: More Automation, Improved Operation |url=https://www.motortrend.com/reviews/ford-bluecruise-version-1-2-first-drive-review/ |access-date=9 April 2023 |website=MotorTrend |language=en}}{{Cite web |title=Ford updates its BlueCruise driver assist with hands-free lane changing and more |url=https://www.engadget.com/ford-updates-its-blue-cruise-driver-assist-with-hands-free-lane-changing-and-more-204022978.html |access-date=9 April 2023 |website=Engadget |date=9 September 2022 |language=en-US}} In April 2023 BlueCruise was approved in the UK for use on certain motorways, starting with 2023 models of Ford's electric [[Ford Mustang Mach-E|Mustang Mach-E]] SUV.{{cite web|date=14 April 2023 |url=https://www.bbc.co.uk/news/business-65272929 |title=Ford launches hands-free driving on UK motorways |website=[[BBC]] |access-date=18 April 2023 }}

[[Tesla Autopilot|Tesla's Autopilot]] and its Full Self-Driving (FSD) ADAS suites are available on all Tesla cars since 2016. FSD offers highway and street driving (without geofencing), navigation/turn management, steering, and dynamic cruise control, collision avoidance, lane-keeping/switching, emergency braking, obstacle avoidance, but still requires the driver to remain ready to control the vehicle at any moment. Its driver management system combines eye tracking with monitoring pressure on the steering wheel to ensure that drives are both eyes on and hands on.{{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 }}{{Cite web |last=Lambert |first=Fred |date=2024-01-22 |title=Tesla finally releases FSD v12, its last hope for self-driving |url=https://electrek.co/2024/01/22/tesla-releases-fsd-v12-last-hope-self-driving/ |access-date=2024-02-03 |website=Electrek |language=en-US}}

Tesla's FSD rewrite V12 (released in March 2024) uses a single deep learning transformer model for all aspects of perception, monitoring, and control.{{Cite web |last=Templeton |first=Brad |date=April 18, 2024 |title=Tesla, Waymo, Nuro, Zoox And Many Others Embrace New AI To Drive |url=https://www.forbes.com/sites/bradtempleton/2024/04/18/tesla-waymo-nuro-zoox-and-many-others-embrace-new-ai-to-drive/ |access-date=2024-05-04 |website=Forbes |language=en}}{{Cite web |last=Mengdan |first=Shen |date=April 3, 2024 |title=Tesla's FSD enters a new phase as self-driving competition intensifies |url=https://www.shine.cn/biz/auto/2404035023/ |access-date=2024-05-04 |website=SHINE |language=en}} It relies on its eight cameras for its vision-only perception system, without use of LiDAR, radar, or ultrasound. As of January 2024, Tesla has not initiated requests for Level 3 status for its systems and has not disclosed its reason for not doing so.

==== Development ==== [[General Motors]] is developing the "Ultra Cruise" ADAS system, that will be a dramatic improvement over their current "Super Cruise" system. Ultra Cruise will cover "95 percent" of driving scenarios on 2 million miles of roads in the US, according to the company. The system hardware in and around the car includes multiple cameras, short- and long-range radar, and a LiDAR sensor, and will be powered by the [[Qualcomm Snapdragon]] Ride Platform. The luxury [[Cadillac Celestiq]] electric vehicle will be one of the first vehicles to feature Ultra Cruise.{{Cite web |last=Hawkins |first=Andrew |date=7 March 2023 |title=GM's Ultra Cruise will use radar, camera, and lidar to enable hands-free driving |url=https://www.theverge.com/2023/3/7/23627656/gm-ultra-cruise-sensor-radar-lidar-hands-free |access-date=9 April 2023 |website=[[The Verge]] |language=en-US}}

=== Level 3 – Conditional automation === {{As of|2024|April}}, two car manufacturers have sold or leased Level 3 cars: Honda in Japan, and Mercedes in Germany, Nevada and California.{{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}}

Mercedes Drive Pilot has been available on the EQS and S-class sedan in Germany since 2022, and in California and Nevada since 2023. A subscription costs between €5,000 and €7,000 for three years in Germany and $2,500 for one year in the United States. Drive Pilot can only be used when the vehicle is traveling under {{Convert|40|mph|kph}}, there is a vehicle in front, readable line markings, during the day, clear weather, and on freeways mapped by Mercedes down to the centimeter (100,000 miles in California).{{Cite web |last=Jones |first=Rachyl |date=April 18, 2024 |title=Exclusive: Mercedes becomes the first automaker to sell autonomous cars in the U.S. that don't come with a requirement that drivers watch the road |url=https://fortune.com/2024/04/18/mercedes-self-driving-autonomous-cars-california-nevada-level-3-drive-pilot/ |access-date=2024-04-20 |website=Fortune |language=en}} As of April 2024, one Mercedes vehicle with this capability has been sold in California.

As of 2026, the only level 3 cars in China are [[Deepal SL03]] and [[Arcfox|Arcfox αS]].{{Cite web |last=X科技实验室 |title=无人送货车售价不到2万，买一台坐里面上下班，有没有搞头？_哔哩哔哩_bilibili |url=https://www.bilibili.com/video/BV1YtFPzSEx1/ |access-date=2026-02-18 |website=bilibili |language=zh-Hans}}

==== Development ==== [[Honda]] continued to enhance its Level 3 technology.{{Cite web |date=1 December 2022 |title=Honda Unveils Next-generation Technologies of Honda SENSING 360 and Honda SENSING Elite |url=https://global.honda/newsroom/news/2022/c221201eng.html?from=RSS&from=newsrelease_area |website=[[Honda]] |access-date=1 December 2022}}{{Cite news|date=1 December 2022 |url=https://jp.reuters.com/article/honda-autonomous/honda-to-develop-advanced-level-3-self-driving-technology-by-2029-idUSL4N32Q0T3 |title=Honda to develop advanced level 3 self-driving technology by 2029 |work=[[Reuters]] |access-date=1 December 2022}} As of 2023, 80 vehicles with Level 3 support had been sold.{{Cite web |first=Christopher |last=Smith |date=28 January 2022 |title=Level 3 Automated Driving Tech Has Major Limitations: Report |url=https://www.motor1.com/news/563945/level3-autonomous-tech-real-world/ |access-date=2024-02-02 |website=Motor1.com |language=en}}

[[Mercedes-Benz]] received authorization in early 2023 to pilot its Level 3 software in Las Vegas.{{cite web | url=https://www.autonomousvehicleinternational.com/news/adas/mercedes-benz-certifies-drive-pilot-for-use-in-nevada-becoming-the-first-l3-system-approved-for-us-highways.html |title=Mercedes-Benz Drive Pilot certified for use in Nevada – first L3 system approved for US highways | date= 27 January 2023 }} California also authorized Drive Pilot in 2023.{{Cite web |title=Mercedes Drive Pilot Level 3 ADAS Approved For Use In California |url=https://insideevs.com/news/671349/mercedes-drive-pilot-level-3-adas-approved-use-california/ |access-date=2024-02-02 |website=InsideEVs |language=en|first= Dan |last=Mihalascu |date=9 June 2023}}

[[BMW]] commercialized its AC in 2021.{{cite web |last=Sergeev |first=Angel |date=31 March 2017 |title=BMW Details Plan For Fully Automated Driving By 2021 |url=https://www.motor1.com/news/147074/bmw-fully-autonomous-car-2021/ |website=Motor1.com}} In 2023 BMW stated that its Level-3 technology was nearing release. It would be the second manufacturer to deliver Level-3 technology, but the only one with a Level 3 technology which works [[Automotive night vision|in the dark]].{{press release |first=Christophe |last=Koenig |title=Level 3 highly automated driving available in the new BMW 7 Series from next spring |url=https://www.press.bmwgroup.com/global/article/detail/T0438214EN/level-3-highly-automated-driving-available-in-the-new-bmw-7-series-from-next-spring |publisher=BMW |date=November 2023|access-date=2026-02-21}}

In 2023, in China, [[IM Motors]], [[Mercedes-Benz|Mercedes]], and [[BMW]] obtained authorization to test vehicles with Level 3 systems on motorways.{{Cite web |last=Kang |first=Lei |date=18 December 2023 |title=IM Motors gets permit to test L3 self-driving vehicles in Shanghai |url=https://cnevpost.com/2023/12/18/im-motors-permit-test-l3-self-driving-shanghai/ |website=CnEVPost}}{{Cite web|url=https://breakingthenews.net/Article/Mercedes-to-test-self-driving-cars-in-Beijing/61161747|title=Breaking The News|website=breakingthenews.net}}

In September 2021, [[Stellantis]] presented its findings from its Level 3 pilot testing on Italian highways. Stellantis's Highway Chauffeur claimed Level 3 capabilities, as tested on the [[Maserati Ghibli (M157)|Maserati Ghibli]] and [[Fiat 500X]] prototypes.{{cite news |last=Myles |first=Paul |date=17 September 2021 |title=Stellantis Shows Off its Level 3 Technology |url=https://www.tu-auto.com/stellantis-shows-off-its-level-3-technology/ |access-date=29 November 2021 |work=[[Informa]]}}

[[Polestar]], a [[Volvo Cars]]' brand, announced in January 2022 its plan to offer Level 3 autonomous driving system in the Polestar 3 SUV, a [[Volvo XC90]] successor, with technologies from [[Luminar Technologies]], [[Nvidia]], and Zenseact.{{cite news |last=Ramey |first=Jay |date=11 January 2022 |title=Polestar 3 with Level 3 Autonomous Tech on the Way |url=https://www.autoweek.com/news/green-cars/a38737805/polestar-3-level-3-autonomous-ride-pilot/ |access-date=31 May 2022 |website=Autoweek}}

In January 2022, [[Bosch (company)|Bosch]] and the [[Volkswagen Group]] subsidiary [[CARIAD]] released a collaboration for autonomous driving up to Level 3. This joint development targets Level 4 capabilities.{{cite news|date=26 January 2022 |publisher=Hannovermesse |title=Bosch and CARIAD advance automated driving |url=https://www.hannovermesse.de/en/news/news-articles/bosch-and-cariad-advance-automated-driving |website=hannovermesse |access-date=26 Jan 2022}}

[[Hyundai Motor Company]] is enhancing cybersecurity of [[connected car]]s to offer a Level 3 self-driving [[Genesis G90]].{{cite news |first1=Jin-woo |last1=Seo |first2=You-jung |last2=Jung |first3=Ha-yeon |last3=Lee |date=16 February 2022 |title=Korean firms enhance car cybersecurity before Level 3 autonomous car releases |url=https://pulsenews.co.kr/view.php?year=2022&no=147214 |access-date=22 April 2022 |work=Pulse by [[Maeil Business Newspaper]]}} Kia and Hyundai Korean car makers delayed their Level 3 plans, and will not deliver Level 3 vehicles in 2023.{{Cite web |last=Herh |first=Michael |date=2023-12-01 |title=Hyundai Motor Puts Level 3 Autonomous Driving Technology on Back Burner |url=https://www.businesskorea.co.kr/news/articleView.html?idxno=206898 |access-date=2024-02-02 |website=Businesskorea |language=ko}}

=== Level 4 – High automation === In 2024, companies such as [[Waymo]] began offering [[robotaxi]] services in parts of the US with fully autonomous vehicles without safety drivers.{{Cite news |last=Ludlow |first=Edward |date=2024-04-15 |title=Waymo, Cruise and Zoox Inch Forward Ahead of Tesla Joining Robotaxi Race |url=https://www.bloomberg.com/news/newsletters/2024-04-15/waymo-cruise-and-zoox-inch-forward-ahead-of-tesla-joining-robotaxi-race |access-date=2024-04-30 |work=Bloomberg |language=en}} These services all operate at a loss as of 2025, with operating costs of about {{convert|7-9|$/mile|abbr=in|order=flip|round=0.5}}, compared to {{convert|1|$/mile|abbr=in|order=flip|1}} for personal cars. The consultancy [[McKinsey & Company|McKinsey]] estimated that bringing costs down to less than {{convert|2|$/mile|abbr=in|order=flip|1}} would take until 2035.{{Cite news |date=2025-11-24 |title=The self-driving taxi revolution is here |url=https://www.economist.com/who-will-win-the-trillion-dollar-robotaxi-race |access-date=2025-11-25 |newspaper=The Economist |language=en}}

In April 2023 in Japan, a Level 4 protocol became part of the amended Road Traffic Act.{{cite news|date=1 April 2023 |url=https://japannews.yomiuri.co.jp/society/general-news/20230401-100936/ |title=Level 4 Autonomous Driving Allowed in Japan |work=[[Yomiuri Shimbun]] |access-date=3 April 2023 }} ZEN drive Pilot Level 4 made by [[National Institute of Advanced Industrial Science and Technology|AIST]] operates there.{{cite web |date=31 March 2023 |title=国内初！自動運転車に対するレベル4の認可を取得しました |trans-title=Domestically the first! Approved as Level 4 self-driving car |url=https://www.meti.go.jp/press/2022/03/20230331002/20230331002.html |access-date=3 April 2023 |website=[[Ministry of Economy, Trade and Industry|METI, Japan]] |language=Japanese}}

==== Development ==== {{See also|Robotaxi#History}}

In July 2020, [[Toyota]] started public demonstration rides on [[Lexus LS]] (fifth generation) based TRI-P4 with Level 4 capability.{{Cite press release |url=https://global.toyota/en/newsroom/corporate/30344967.html |date=24 October 2019 |title=Toyota to Offer Rides in SAE Level-4 Automated Vehicles on Public Roads in Japan Next Summer |publisher=[[Toyota]] |access-date=17 March 2022}} In August 2021, Toyota operated a potentially Level 4 service using [[Toyota concept vehicles (2010–2019)#e-Palette|e-Palette]] around the [[2020 Summer Olympics|Tokyo 2020 Olympic]] Village.{{cite news |last=Davis |first=River |date=2 August 2021 |title=Hyperdrive Daily: The Driverless Shuttle Helping Toyota Win Gold |url=https://www.bloomberg.com/news/newsletters/2021-08-02/toyota-seizes-olympic-glory-by-shuttling-athletes-autonomously |access-date=7 November 2021 |work=[[Bloomberg News]]}}

In September 2020, [[Mercedes-Benz]] introduced world's first commercial Level 4 [[Automated Valet Parking]] (AVP) system named ''Intelligent Park Pilot'' for its new [[Mercedes-Benz S-Class (W223)|S-Class]].{{cite web |date=2 September 2020 |title=Automotive luxury experienced in a completely new way – The main points of the new Mercedes-Benz S-Class at a glance |url=https://media.mercedes-benz.com/article/92eb97a3-3477-4ca1-acf1-32ee0e07d646 |access-date=21 May 2022 |website=Mercedes me media |language=en}}{{cite web |title=Bosch – Stuttgart Airport Set to Welcome Fully Automated and Driverless Parking |url=https://iot-automotive.news/bosch-stuttgart-airport-set-to-welcome-fully-automated-and-driverless-parking/ |website=IoT Automotive News |access-date=21 May 2022 }} In November 2022, Germany’s Federal Motor Transport Authority (KBA) approved the system for use at [[Stuttgart Airport]].{{Cite web |last= |first= |date=2022-11-30 |title=Mercedes-Benz and Bosch driverless parking system: Approved for commercial use |url=https://group.mercedes-benz.com/innovation/product-innovation/autonomous-driving/intelligent-park-pilot.html |access-date=2024-02-03 |website=Mercedes-Benz Group |language=en}}

In September 2021, Cruise, General Motors, and Honda started a joint testing programme, using Cruise AV.{{Cite press release |url=https://global.honda/newsroom/news/2021/c210908eng.html |date=8 September 2021 |title=Honda to Start Testing Program in September Toward Launch of Autonomous Vehicle Mobility Service Business in Japan |publisher=[[Honda]] |access-date=16 March 2022}} In 2023, the Origin was put on indefinite hold following Cruise's loss of its operating permit.{{Cite news |last=MILLER |first=CALEB |date=November 29, 2023 |title=GM's Self-Driving Cruise Origin Indefinitely Delayed Amid Major Setbacks |url=https://www.caranddriver.com/news/a45989501/gm-self-driving-cruise-origin-delayed/ |work=[[Car and Driver]]}}

In January 2023, Holon announced an autonomous shuttle during the 2023 [[Consumer Electronics Show]] (CES). The company claimed that the vehicle is the world's first Level 4 shuttle built to automotive standard.{{cite news |last=James |first=Anthony |date=5 January 2022 |title=New Benteler brand Holon presents world's first autonomous mover built to automotive standards |url=https://www.autonomousvehicleinternational.com/news/mobility-solutions/new-benteler-brand-holon-presents-worlds-first-autonomous-mover-built-to-automotive-standards.html |access-date=21 January 2023 |work=ADAS & Autonomous Vehicle International}}

==See also==

[[History of self-driving cars]]
[[Regulation of self-driving cars]]
[[Vehicular automation]]
[[Autopilot]]
[[Driving]]
[[Car]]

== References == {{reflist}}

== Further reading == {{Commons category-inline|Autonomous automobiles}} {{Prone to spam|date=December 2014}} {{Refbegin|2}}

{{cite book|url={{google books |plainurl=y |id=DW2QAAAAQBAJ}}|title=Gridlock: Why We're Stuck in Traffic and What To Do About It|last=O'Toole|first=Randal|date=18 January 2010|publisher=Cato Institute|isbn=978-1-935308-24-9}}
{{cite book|url=http://www.inf.ed.ac.uk/publications/thesis/online/IM110982.pdf|title=A Simulated Autonomous Car|last=Macdonald|first=Iain David Graham|publisher=The University of Edinburgh|year=2011|type=thesis|access-date=17 April 2013}}
{{cite web|url=https://www.technologyreview.com/s/520431/driverless-cars-are-further-away-than-you-think/|title=The Future of Self-driving Cars|last=Knight|first=Will|date=22 October 2013|website=MIT Technology Review|access-date=22 July 2016}}
{{Cite journal|last1=Taiebat|first1=Morteza|last2=Brown|first2=Austin|last3=Safford|first3=Hannah|last4=Qu|first4=Shen|last5=Xu|first5=Ming|year=2019|title=A Review on Energy, Environmental, and Sustainability Implications of Connected and Automated Vehicles.|journal=Environmental Science & Technology|volume=52|issue=20|pages=11449–11465|doi=10.1021/acs.est.8b00127|pmid=30192527|arxiv=1901.10581|bibcode=2019arXiv190110581T|s2cid=52174043}}
{{cite report|url=http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_lrd_069.pdf|title=A Look at the Legal Environment for Driverless Vehicles|last=Glancy|first=Dorothy|publisher=Transportation Research Board|location=Washington, DC|volume=69|isbn=978-0-309-37501-6|access-date=22 July 2016|series=National Cooperative Highway Research Program Legal Research Digest|year=2016}}
{{cite web|url=http://theloadstar.co.uk/the-driving-forces-behind-what-would-be-the-next-revolution-in-the-haulage-sector/|title=The driving forces behind what would be the next revolution in the haulage sector|last=Newbold|first=Richard|date=17 June 2015|website=The Loadstar|access-date=22 July 2016}}
{{cite web|url=http://recode.net/2015/10/27/meet-the-companies-building-self-driving-cars-for-google-and-tesla-and-maybe-apple/|title=Meet the Companies Building Self-Driving Cars for Google and Tesla (And Maybe Apple)|last1=Bergen|first1=Mark|date=27 October 2015|website=re/code}}
{{cite web|url=http://ntl.bts.gov/lib/57000/57000/57076/Review_FMVSS_AV_Scan.pdf|title=Review of Federal Motor Vehicle Safety Standards (FMVSS) for Automated Vehicles: Identifying potential barriers and challenges for the certification of automated vehicles using existing FMVSS|author=John A. Volpe National Transportation Systems Center|date=March 2016|website=National Transportation Library|publisher=[[US Department of Transportation]]|author-link=John A. Volpe National Transportation Systems Center|access-date=6 April 2016|archive-date=16 June 2017|archive-url=https://web.archive.org/web/20170616204001/https://ntl.bts.gov/lib/57000/57000/57076/Review_FMVSS_AV_Scan.pdf|url-status=dead}}
{{cite web|url=http://knowledgecenter.csg.org/kc/system/files/CR_automomous.pdf|title=State Laws on Autonomous Vehicles|last=Slone|first=Sean|date=August 2016|website=Capitol Research – Transportation Policy|publisher=[[Council of State Governments]]|access-date=28 September 2016|archive-date=28 February 2021|archive-url=https://web.archive.org/web/20210228043344/http://knowledgecenter.csg.org/kc/system/files/CR_automomous.pdf|url-status=dead}}
{{cite news|url=https://www.npr.org/2015/07/31/427990392/remembering-when-driverless-elevators-drew-skepticism|title=Remembering When Driverless Elevators Drew Skepticism|first=Steve|last=Henn|date=31 July 2015}}
{{cite web|url=http://www.rand.org/content/dam/rand/pubs/research_reports/RR400/RR443-2/RAND_RR443-2.pdf|title=Autonomous Vehicle Technology: A Guide for Policymakers|first=James M.|last=Anderson|display-authors=etal|year=2016|publisher=[[RAND Corporation]]}} *{{Cite journal |date=2014 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation |url=https://link.springer.com/book/10.1007/978-3-319-05990-7 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-319-05990-7 |isbn=978-3-319-05989-1 |issn=2196-5544|url-access=subscription }} **{{Cite journal |date=2015 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation 2 |url=https://link.springer.com/book/10.1007/978-3-319-19078-5 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-319-19078-5 |isbn=978-3-319-19077-8 |issn=2196-5544|url-access=subscription }} ** {{Cite journal |date=2016 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation 3 |url=https://link.springer.com/book/10.1007/978-3-319-40503-2 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-319-40503-2 |isbn=978-3-319-40502-5 |issn=2196-5544|url-access=subscription }} ** {{Cite journal |date=2018 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation 4 |url=https://link.springer.com/book/10.1007/978-3-319-60934-8 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-319-60934-8 |isbn=978-3-319-60933-1 |issn=2196-5544|url-access=subscription }} ** {{Cite journal |date=2019 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation 5 |url=https://link.springer.com/book/10.1007/978-3-319-94896-6 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-319-94896-6 |isbn=978-3-319-94895-9 |s2cid=168659939 |issn=2196-5544|url-access=subscription }} ** {{Cite journal |date=2019 |editor-last=Meyer |editor-first=Gereon |editor2-last=Beiker |editor2-first=Sven |title=Road Vehicle Automation 6 |url=https://link.springer.com/book/10.1007/978-3-030-22933-7 |journal=Lecture Notes in Mobility |language=en |doi=10.1007/978-3-030-22933-7 |isbn=978-3-030-22932-0 |issn=2196-5544|url-access=subscription }} These books are based on presentations and discussions at the Automated Vehicles Symposium organized annually by [[Transportation Research Board|TRB]] and [[Association for Unmanned Vehicle Systems International|AUVSI]].
{{cite web|url=https://journals.sas.ac.uk/deeslr/issue/view/528|title=Autonomous vehicles – who will be liable for accidents?|first=Roger|last=Kemp|year=2018|publisher=[15 Digital Evidence and Electronic Signature Law Review (2018) 33 – 47]}} {{Refend}}

{{Self-driving cars and enabling technologies}} {{Navboxes|list= {{Emerging technologies|transport=yes}} {{Artificial intelligence navbox}} {{Automobile configuration}} {{Automated trains and fixed-guideway transit}} {{Mobile robots}} {{Computer vision footer}} {{SAE International}} }}

[[Category:Self-driving cars| ]] [[Category:Automotive technologies]] [[Category:Automotive safety]] [[Category:Driving]] [[Category:Transport culture]]

Autonomous car

From MOAI Insights