{
  "id": "robotaxi",
  "title": "Robotaxi",
  "lang": "en",
  "source_lang": "en",
  "summary": "A robotaxi is a self-driving taxi that carries paying passengers without a human driver, operated as a ride-hailing service.",
  "category": "technology",
  "subcategory": "transport",
  "type": "concept",
  "aliases": [
    "Autonomous taxi",
    "driverless taxi",
    "self-driving taxi",
    "autonomous ride-hailing",
    "driverless ride-hailing"
  ],
  "generated_at": "2026-10-16",
  "model": "pro/claude-haiku-5-5",
  "url": "https://aiwiki.page/en/robotaxi/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/robotaxi/",
      "json": "https://aiwiki.page/zh/robotaxi.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A robotaxi is a passenger vehicle that drives itself and carries members of the public for a fare, usually through a smartphone app and without a human driver in the vehicle. It is a commercial application of artificial-intelligence and robotics to road transport, and it differs from driver-assistance features in private cars because the operator, not the passenger, is responsible for the driving task within a defined service area. Most current robotaxis are battery-electric cars or vans fitted with electric drivetrains and a dense set of sensors, and they are typically run by a company that owns or manages the fleet."
    },
    {
      "heading": "Definition and classification",
      "text": "Robotaxis are usually described using the driving-automation scale published by SAE International, in which Level 4 describes a system that performs the entire driving task within a limited operating domain without relying on a human to take over. A robotaxi that operates without a safety driver in a geofenced zone meets that description. Systems that still require a supervising person in the driver's seat, or a remote operator who can intervene, are often called \"supervised\" or \"unsupervised with monitoring\" and are placed in a grey area between commercial pilot and full driverless service. Industry sources therefore differ on whether a given service counts as a robotaxi, which complicates comparisons of fleet sizes and ride volumes.\n\nRobotaxis are distinct from related concepts. Autonomous shuttles and buses run on fixed routes, while robotaxis serve on-demand trips. Driver-assistance systems, such as lane keeping or highway assist, require constant human supervision, and so do not qualify. Robot delivery vehicles and driverless trucks share much of the same technology stack but serve freight rather than passengers."
    },
    {
      "heading": "Technology",
      "text": "A robotaxi combines several technical layers. Perception relies on cameras, lidar (laser ranging), and radar to build a real-time model of the surroundings, and deep neural networks based on computer-vision and deep-learning classify objects such as pedestrians, cyclists, and other vehicles. Prediction models estimate how these agents are likely to move, and planning modules choose a trajectory that satisfies traffic rules and passenger comfort. Many operators also rely on detailed high-definition maps of their service areas, although newer systems aim to reduce dependence on prior mapping so they can expand more quickly.\n\nRedundancy is a central engineering concern. Operators typically duplicate critical components, including braking, steering, power, and computing, so that a single fault does not leave the vehicle without control. Remote assistance centers allow human staff to advise a vehicle that is stuck or uncertain, though operators differ on how much remote intervention they permit and how it is described publicly.\n\nHardware cost has been a major factor in commercial viability. Sensor prices have fallen as lidar units have become cheaper and more compact, and newer purpose-built vehicles are designed for driverless operation from the outset rather than retrofitted, removing steering wheels and pedals in some cases."
    },
    {
      "heading": "History",
      "text": "Early robotaxi development grew out of academic and military autonomous-vehicle research, most notably the DARPA Grand Challenges of the 2000s, and later out of large corporate programs. Commercial passenger services began in small, controlled deployments. In China, Baidu's Apollo Go service began public operation in Beijing in 2020. In the United States, Waymo, which grew out of a self-driving car project within Google, began offering public rides in Phoenix before expanding to other cities.\n\nThrough the mid-2020s the field moved from pilot programs toward paid service at scale. Ride volumes grew quickly, and several companies began expanding internationally, particularly into the Gulf states, Europe, and East Asia."
    },
    {
      "heading": "Current deployment (as of late 2026)",
      "text": "Reporting from 2026 indicates that the sector has moved from testing into sustained commercial operation, although the scale varies widely between operators and figures are often company-reported.\n\nUnited States\n\nWaymo is the largest operator in the United States. In March 2026 the company said it was providing 500,000 paid robotaxi rides every week across 10 U.S. cities, having grown from roughly 50,000 weekly rides in May 2024. In September 2026 it announced paid public service in Denver, San Diego, and Tampa, which brings the number of US cities where its driverless ride-hail service is available to 14. Waymo has said it aims for more than one million paid weekly rides by the end of 2026 and for operations in more than 20 cities during the year, including international locations such as London and Tokyo. Its February 2026 financing round of about USD 16 billion was intended to support that expansion.\n\nTesla has taken a different path, combining a large consumer vehicle base with a purpose-built vehicle, the Cybercab. Its service began in Austin and was extended to the San Francisco Bay Area, and in April 2026 it expanded to Dallas and Houston with unsupervised vehicles. Tesla had earlier signalled a broader rollout to several more U.S. cities during 2026. Independent analysts have questioned how many vehicles actually operate without a safety monitor, and the gap between announced plans and verified fleet size remains a point of debate. Motional, a joint venture associated with Hyundai, has targeted fully driverless operations in Las Vegas by late 2026.\n\nChina and Asia\n\nChinese operators have the largest robotaxi fleets outside the United States. Baidu's Apollo Go reported in its first-quarter 2026 results that it delivered 3.2 million fully driverless operational rides, with weekly rides peaking above 350,000 in March. Pony.ai reported a fleet of 1,975 vehicles as of 30 June 2026 and said it was targeting more than 20 cities worldwide by the end of 2026. WeRide reported that its robotaxi fleet in China had grown to about 1,000 vehicles by April 2026. Both Pony.ai and WeRide have reported city-level operations in the four largest Chinese cities and additional regional hubs. Apollo Go has also begun testing in Hong Kong and in London, and has been operating in Switzerland with a public transport partner.\n\nMiddle East and Europe\n\nThe United Arab Emirates has become a notable early market because regulators have issued permits for fully driverless operation. Apollo Go launched fully driverless commercial service in Dubai in 2026 through its own app and, in August 2026, became available to riders on the Uber platform there. In Abu Dhabi, Apollo Go and its partner AutoGo began a fully driverless public service on Yas Island, with plans to expand across the emirate. Pony.ai has reported commercial operations in Doha, Qatar, and in Zagreb, Croatia, in partnership with Uber and a Croatian vehicle developer."
    },
    {
      "heading": "Business models",
      "text": "Three broad models have emerged. In the vertically integrated model, a company develops the technology, owns the fleet, and runs the ride-hailing app, as Waymo does in some markets and Tesla intends to do. In the platform model, an autonomous-driving developer supplies the software while a ride-hailing platform such as Uber or Lyft supplies demand, and a fleet operator manages the vehicles. In the manufacturing model, a carmaker or contract manufacturer builds purpose-designed robotaxis for several operators; Hyundai and a Chinese manufacturer, Geely's Farizon unit, have both announced such arrangements. These models affect who bears capital costs and who controls customer data.\n\nOperators present the economics of robotaxis as a path to lower per-kilometre costs, because the largest cost in a conventional taxi, the driver's wage, is removed, and vehicles can operate for longer periods. Pony.ai has stated that its robotaxi business reached city-wide unit-economics breakeven in Shenzhen in February 2026, a claim reported by the company and not independently audited. Vehicle depreciation, remote support, insurance, cleaning, charging, and maintenance remain significant costs, and profitability at scale has not been demonstrated across the industry."
    },
    {
      "heading": "Regulation and governance",
      "text": "Robotaxis are regulated at several levels. National governments set vehicle safety standards and, in some countries, rules for testing and commercial operation. Subnational authorities often control permits for specific cities or zones. In the United States, regulation is largely state-based, and in some cities the legal framework is unclear. For example, one 2026 report noted that Waymo received a permit to test its vehicles with a human operator in New York City, but no rules exist for fully autonomous service. In the UAE and China, permits have been issued by transport authorities in designated zones, often with conditions on speed, weather, and area of operation.\n\nRegulators typically require operators to report crashes, disengagements, and incidents, and to demonstrate safety through testing mileage. Liability is a key legal question: when a driverless vehicle causes harm, responsibility may fall on the manufacturer, the software developer, the fleet operator, or the owner, and national laws differ on how this is allocated. insurance markets are adapting to this uncertainty."
    },
    {
      "heading": "Safety and public acceptance",
      "text": "Operators argue that robotaxis may reduce crashes, since human error contributes to most road collisions. Companies have published safety statistics showing lower rates of injury crashes per mile than human drivers in their operating areas, but these figures are self-reported, depend on the environments and mileage chosen, and are difficult to compare with national crash statistics. Researchers also note that the most informative evidence will come from independent analysis over many years.\n\nPublic attitudes remain mixed. Surveys in various countries show concern about safety, data collection, and job losses among drivers. Vehicles have been involved in notable incidents, including collisions and blocked emergency access, which have prompted scrutiny from regulators and local officials. Questions of data-privacy, cybersecurity, and accessibility for disabled and elderly users are also active areas of policy discussion."
    },
    {
      "heading": "Economic and social effects",
      "text": "Robotaxis may affect the labor-market for professional drivers, and the scale of this effect depends on how quickly services expand and how they are priced. They may also change urban travel patterns. Some analysts expect that cheaper on-demand autonomous transport could reduce car ownership and parking demand, while others warn that it could increase total vehicle travel by drawing riders away from public-transport and walking. Planners therefore discuss how robotaxis should be integrated with transit, traffic management, and urban-planning. Their deployment also raises questions of ethics, including how algorithms should trade off risks to different road users."
    },
    {
      "heading": "Limitations and open problems",
      "text": "Robotaxis still face technical limits in unusual weather, complex construction zones, unmarked roads, and interactions with unpredictable human road users. Operating in dense urban areas at peak hours, or during heavy rain or snow, has been more difficult than in simpler conditions. Expansion to new cities requires new mapping, local validation, and regulatory approval, which slows growth. Many operators also depend on a small number of suppliers for lidar, chips, and vehicles, which creates supply-chain risk."
    },
    {
      "heading": "Data reliability",
      "text": "Figures for robotaxi fleets, rides, and cities are frequently company-reported and are not standardised. Sources also differ on how to count cities; for example, one count of Apollo Go's footprint listed 22 cities while a later company release cited 28. Counts of \"driverless\" operation sometimes include vehicles with safety monitors, and some third-party statistics pages are compiled from secondary sources. Readers should treat specific numbers as indicative of direction and scale rather than as precise, audited measurements."
    }
  ],
  "links": [
    "artificial-intelligence",
    "robotics",
    "automobile",
    "electric-vehicle",
    "computer-vision",
    "deep-learning",
    "internet",
    "insurance",
    "data-privacy",
    "cybersecurity",
    "labor-market",
    "public-transport",
    "urban-planning",
    "ethics"
  ],
  "sources": [
    {
      "title": "Waymo Expands Paid Robotaxi Rides to Denver, San Diego and Tampa - Bloomberg",
      "url": "https://www.bloomberg.com/news/articles/2026-09-01/waymo-expands-paid-robotaxi-rides-to-denver-san-diego-and-tampa"
    },
    {
      "title": "Waymo’s skyrocketing ridership in one chart",
      "url": "https://techcrunch.com/2026/03/27/waymo-skyrocketing-ridership-in-one-chart/"
    },
    {
      "title": "Waymo Co-CEO Outlines Path to 1 Million Weekly Trips in 2026",
      "url": "https://www.claimsjournal.com/news/national/2026/02/12/335668.htm"
    },
    {
      "title": "Waymo Targets 1 Million Robotaxi Rides A Week",
      "url": "https://www.forbes.com/sites/alanohnsman/2025/12/10/waymo-targets-1-million-robotaxi-rides-a-week/"
    },
    {
      "title": "Robotaxi Market Size, Share",
      "url": "https://www.fortunebusinessinsights.com/robo-taxi-market-103661"
    },
    {
      "title": "Pony AI Inc. - Form 6-K - FY2026",
      "url": "https://www.sec.gov/Archives/edgar/data/1969302/000110465926034888/tm269906d1_ex99-2.htm"
    },
    {
      "title": "Pony AI Inc. - Form 6-K - FY2026",
      "url": "https://www.sec.gov/Archives/edgar/data/0001969302/000110465926098113/tm2623382d1_ex99-2.htm"
    },
    {
      "title": "WeRide Inc. - Form 6-K - FY2026",
      "url": "https://www.sec.gov/Archives/edgar/data/0001867729/000110465926059704/tm2614059d1_ex99-1.htm"
    },
    {
      "title": "Baidu Announces First Quarter 2026 Results",
      "url": "https://ir.baidu.com/news-releases/news-release-details/baidu-announces-first-quarter-2026-results"
    },
    {
      "title": "Uber Technologies, Inc. - Uber Launches Baidu’s Fully Driverless Apollo Go in Dubai, Establishing the First Multi-Partner Autonomous Network Globally",
      "url": "https://investor.uber.com/news-events/news/press-release-details/2026/Uber-Launches-Baidus-Fully-Driverless-Apollo-Go-in-Dubai-Establishing-the-First-Multi-Partner-Autonomous-Network-Globally/default.aspx"
    },
    {
      "title": "Baidu's Apollo Go Goes Live on Uber in Dubai, Offering a New Way for Users to Hail Fully Driverless Rides",
      "url": "https://www.prnewswire.com/news-releases/baidus-apollo-go-goes-live-on-uber-in-dubai-offering-a-new-way-for-users-to-hail-fully-driverless-rides-302856106.html"
    },
    {
      "title": "Baidu's Apollo Go and K2's AutoGo Commence Fully Autonomous Ride-Hailing Service on Yas Island, Announce Phased Expansion Across Abu Dhabi",
      "url": "https://www.prnewswire.com/news-releases/baidus-apollo-go-and-k2s-autogo-commence-fully-autonomous-ride-hailing-service-on-yas-island-announce-phased-expansion-across-abu-dhabi-302663468.html"
    },
    {
      "title": "Tesla Robotaxi",
      "url": "https://en.wikipedia.org/wiki/Tesla_Robotaxi"
    },
    {
      "title": "Tesla expands Unsupervised Robotaxi service to two new cities",
      "url": "https://www.teslarati.com/tesla-expands-unsupervised-robotaxi-service-two-new-cities/"
    },
    {
      "title": "Apollo Go（autonomous driving ride-hailing service platform）_Baiduwiki",
      "url": "https://baike.baidu.com/en/item/Apollo%20Go/1462954"
    },
    {
      "title": "Robotaxi Statistics 2026: Trips, Fleets, Revenue and Safety Data",
      "url": "https://sqmagazine.co.uk/robotaxi-statistics/"
    }
  ],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}