Thursday 28 December 2023

A New Year wish – autonomous cars.

Driverless vehicles are in the news again. (Driverless cars: Tech possible for UK motorways by 2026, transport secretary says). 

The usual justification given for needing driverless cars is safety because most car accidents are caused by human error. I disagree because it seems unlikely driverless cars will ever be demonstrably and significantly more safe than cars driven by humans. However, that does not mean we do not need autonomous vehicles. I think the main reasons are improving equality of access and reducing GHG emissions. That may surprise you as it is often predicted autonomous vehicles will increase car use and emissions. However, driverless cars enable cheaper and more convenient public transport, and also enable policies that can discourage car use. 

First consider equity. 

  • People with low incomes struggle to buy and run a car – or even to learn to drive. 
  • Some people, for example the elderly with slower reactions, are more prone to accidents and should not be driving. However, giving up their licence is such a loss of liberty (because of the lack of alternatives) many keep driving longer than they should. 
  • Living in an area with poor public transport, such as a rural area makes life without a car even harder. This means low income or elderly people in rural areas have a particularly tough time.

Driverless cars would allow most people to use a car at least occasionally, as it would be more affordable than a conventional taxi with a driver. 

40% of the lowest income households do not have access to a car or van compared to 12% of the highest income households. Data from National Travel Survey Statistics NTS 0703

Accident rates where someone is killed or seriously injured are higher for drivers older than 70 or younger than 25. Chart from

The number of licence holders 70 years old or over increased by 50% in the 10 years to 2022, while younger age groups held fairly steady. Data from National Travel Survey Statistics NTS 0701

Consider now GHG emissions. 

As of now, car use causes unhealthy air pollution and congestion as well as GHG emissions. We need to reduce the need to travel (e.g. 15 minute cities) and also shift modes from car to public transport. So far this problem has proven intractable because public transport can never be as convenient as a private car. However, using driverless cars we can implement intermediate solutions. For example, driverless taxis will be cheaper because there is no need to pay a driver. Also driverless buses could be more frequent because you don’t have to pay a driver for each bus. Even if they do require staff on board to check tickets, this would speed up boarding and hence reduce travel times. Driverless buses are under test in Edinburgh, as of May this year [1].

Suppose you have your own car now. Perhaps you could accept these scenarios:

  1. You live in an area where private cars are not allowed to park on the road. This is great for kids, and you have better air quality too. You need a car, though, so you rent a space in a car park half a mile away and your driverless EV comes to your door when you need it. You don't bother with the car for shopping, though, unless the weather is bad. The bike with a trailer is about as convenient.
  2. There is a bus service of sorts but it is infrequent. Also, quite often you have a lot of stuff to carry so you need a car for a door to door service. Conventional taxis are too expensive so you have your own car. This is expensive to buy and insure, but since you have it anyway you use it a lot. More frequent driverless buses and affordable driverless taxis change the economics. You find you don’t need a car of your own any more. 
  3. You need a car to get to work because the buses are slow and unreliable. You would like to car-share to save money – the daytime parking fees are very high. However, sometimes you have to work late and that can leave you stranded – or your co-sharers stranded if you are the driver that day. Now you can call up an affordable driverless taxi to get you home – or even call your own car to come and fetch you. Car sharing has become practical. 

The car-less residential estates in the first scenario are almost unthinkable now in the UK. However, consider that in Tokyo on-street parking is illegal. Residents often have to rent a space in a private car park – and they have to walk to get their car when they need it [2].

Also, in Amsterdam you may not park in an area without a parking permit, which you have to apply for, and there is limited availability, and you have to pay a fee [3].

Acceptability of driverless cars is much higher among 24-55 year olds than older people [4]. The younger group are in the majority. Perhaps they will become more averse to safety critical technology as they grow older - or perhaps not.

The availability of driverless cars makes policies to reduce car use such as parking restrictions or high car parking fees more acceptable. They don’t solve all the problems in one go, but they provide significant nudges along the way. 


[1] UK's first driverless bus begins passenger service in Edinburgh (BBC) May 2023




  1. You say: "I disagree because it seems unlikely driverless cars will ever be demonstrably and significantly more safe than cars driven by humans." but what you mean is "... I think it seems unlikely driverless cars ..." unless you have some good evidence! Also it's only ICE cars that cause GHG emissions, and EVs resolve that problem. So I think you need to rethink.

    1. I do not say that EVs will not be safer, just that they are unlikely to be demonstrably and significantly safer. I should think they will be demonstrably safer than *some* drivers such as those under the influence of drink or drugs, drivers with fading eyesight, drivers making phone calls ... Also, EV cars have potential to eliminate GHG emissions but not the equity issues.


Comments on this blog are moderated. Your comment will not appear until it has been reviewed.