Self-driving vehicles

Integrated sensors, including cameras, radars and light detection technologies, collect vast amounts of data that train the automated system running the SDV and allow it to perceive its surroundings, anticipate its route and alert about any potential obstacles. 

Photographs and videos (visual data) are labelled manually by humans in a way that enables the AI system to learn to recognise and react appropriately to other cars, bicycles and pedestrians. One advantage of this approach is that the information learned by one vehicle can be shared across a whole fleet of vehicles operated by the same AI system. 

SDVs are already in use in Phoenix, Arizona and San Francisco, California in the USA – where they are also deployed by Waymo (owned by Google’s parent company Alphabet) and General Motors subsidiary Cruise – to provide driverless taxi services, referred to as ​‘robotaxis’. This service is currently available only in San Francisco, but is due to be extended to other US cities. Driverless taxis operate in a way similar to Uber, the transport-on-demand company that enables users to hail a taxi using a smartphone app. 

SDVs that operate as taxis or offer ride-sharing options are thought to become a cheaper, safer and more sustainable alternative to individual car ownership. However, at present, they are much more expensive to run than regular taxis.

Additionally, SDVs promise safety benefits by reducing the risk of accidents caused by human drivers.

While presented as potentially safer than human-driven vehicles, road safety remains one of the main risks posed by the use of SDVs.

There have been crashes and fatalities involving SDVs, including a fatal crash involving an Uber SDV. Cruise has been suspended after an accident in San Francisco, where a pedestrian was injured by one of its SDVs. Also, emergency responders in San Francisco have voiced frustrations with driverless taxis blocking the roads, delaying rescue vehicles.

Taxi drivers in the Bay Area, where driverless taxis have been in circulation since 2022, have also raised concerns over the risk of losing their jobs to robotaxis.

Another risk is that, as this technology becomes more commonplace, people will become more dependent on cars, and less likely to use public or shared transport, which could have negative implications on road traffic and emissions.

Finally, SDVs are likely to work better in some environments – such as those with clear, well-marked roads with few pedestrians – than in others, leading to an unequal and uneven access to the technology.