The development of driverless cars offers a wonderful opportunity to stem the number of road deaths and injuries, explains Professor Steve Furber.

The announcement by Google that it is to manufacture driverless cars could herald the greatest improvement in road safety since the car was invented.  After all, it takes out of the equation the biggest risk factor – the driver.

We kill more people on the roads every year than we do in wars – 1,750 deaths in the UK in 2012 and the UK has among the best road safety records. The global annual road casualty figure is over a million. The economic welfare costs of road accidents in the UK are estimated at around £16bn a year and road accidents are the leading killer of young adults aged 15 to 24. Why do we put up with this? It is now largely avoidable.

The key route to improved road safety is to get the human out of the loop; we are clearly the weak link in the system. At first this will come through enhanced technology supporting the driver – automatically enforced speed limits, automatic collision avoidance, and so on.

But the ultimate solution is driverless cars, such as those currently being developed by Google. Their use could lead to lower car ownership, fewer car parking issues, much lower environmental impact and other social benefits.

Mainstream car development is already bringing in driver safety support technologies, such as warnings when a driver crosses a white line on a motorway without signaling and collision avoidance radars. More controversial technologies, such as automated enforcement of speed limits, are already clearly possible and will be adopted over time. But these will face opposition from those who see their right to exceed the speed limit as more important than others’ right to life.

The technology to achieve fully driverless cars – where human control is restricted to an emergency override that stops the car when the rest of the system deems it safe – is complex.  But as the Google prototype shows, it is now within our reach.

The driverless car has to have an advanced vision system and powerful computers to understand the complex environment that every driver deals with. But computers have extremely good reaction times compared to humans, so streams of cars could assemble into road trains on motorways at spacings that would be far too dangerous if they relied on human reaction times. These cars will save fuel consumption as each runs in the slipstream of the vehicle in front.

Streams of traffic could cross rather like a motorcycle stunt team, missing each other by small margins, while eliminating the need for traffic lights and considerably increasing the traffic capacity of existing roads. Fully automated cars could report the local traffic speed to central servers and adjust routes rapidly to avoid congested areas.

The benefits of driverless cars are not just the elimination of road carnage. Many of us have had to counsel elderly parents to give up driving. A driverless car would remove the problem of isolation that comes with handing in a driving licence. The same situation would apply for those with visual and other impairments that affect their ability to drive.

Some interesting problems may emerge with cars that cannot collide with anything. Humans are good optimisers and pedestrians would soon realise that walking across the road whenever they want is perfectly safe, because cars will stop automatically without hitting them. This would be good for quiet side streets, where children could again play in total safety and pedestrian priority is appropriate.

On major thoroughfares, though, it could cause chaos. Some greater physical separation of cars and pedestrians may be necessary – which is desirable in any case and could be funded from the saving in health service costs.

Government could accelerate the development of driverless cars by encouraging relevant research and establishing a legal framework for autonomous vehicles – but there will be political consequences. The petrol heads will resent being banned from public roads and confined to race tracks, where their right to kill themselves and similarly-minded others might persist – but only after they have made huge insurance payments to cover the health risks.

Professional drivers will have to find new occupations. The number of cars required to keep us moving will drop to perhaps 15% of the current number ‘on the road’ – which typically means parked, but available for use any time the owner wants to go somewhere. This transition will take a few decades, so there is time to plan.

Personally, I look forward to this development. It might come at just the right time for me, as my own children will then worry about my inability to drive safely, just as I did for my parents not long ago. “Dad, please get a driverless car”, I can hear them pleading. That’s so much better than “Dad, we really think it’s time you got rid of the car”.

I will be easy to persuade as I already resent the lost time, as I see it, that I spend driving. I travel by train when it is practical, as I can get work done. This blog was written on the train to Manchester from London. If in 20 years’ time I can climb into the back of my car, tell it where I want to go and then sit down to write another blog, I’ll be very happy, and very much safer.