The safe view of autonomous cars

A former GM car factory in Detroit
Credit: Thomas Hawk on Flickr

There are three key factors which affect car insurance – accidents (claims), theft, and the policy holder.

Here’s a thought about autonomous cars and accidents: “A total of 25,160 people were killed or seriously injured in the year ending September 2016, up by 6 per cent from the previous year, and 182,560 casualties of all severities.” Of that 25,160, around 5% are caused by drink-driving. The current estimate is that 1,380 people were killed or seriously injured when at least one driver was over the limit.

Whilst these figures only overlap by a few months (the drink driving numbers are for the year 2015), let’s combine them as there’s nothing to suggest there was any statistical anomalies around that time.

According to the Institute of Advanced Motorists, “in 2014 driver/rider error or reaction were cited as contributory factors in 74% of accidents”. I doubt driver or rider skills changed much between 2014 and 2016, so some simple maths shows that around 18,618 people were killed or seriously injured by a car driver or rider who was at fault for the accident.

Of the fatalities, over 70% those people were under 60 years’ old.

It’s surprising with these numbers that there aren’t more mandatory safety features in cars. Saying that, I remember the negative public reaction to mandatory seat belts when I was 10 years old. During that time, car accidents were more than three times higher than today, with about 5,500 fatalities per year. So I’d be surprised for a government to fit speed limiters to cars, even though it does question why you need a car capable of 70+, 100+ or 150+ mph on the road.

Barely a day goes by without a new press release about the latest self-driving car technology. It’s likely the first public-scale auto-driving car will be from Tesla. There have already been fatal accidents in self-driving Teslas, but the statistics are incredible. The first fatality was after 130 million miles. Their share price dropped 1% when the accident occurred, and since then its increased 50%.

Why are self-driving cars being tested in the US? Most of the US has a culture of crossing a road at a junction, traffic light or other recognised crossing. Not so in the UK or many other European cities though. In the UK we cross the road pretty much wherever it’s convenient for the pedestrian. And that could have a profound issue on self-driving cars.

If the public learns a car is programmed to always stop for a pedestrian in danger, they’ll cross the road even if an automated car is approaching at speed because they’ll trust the car to stop or avoid them. It will result in a rough journey for car passengers.

Nissan announced it began testing an autonomous driving car in London this year. It only engages full autonomy for multi-lane highways. My bet is that “multi-lane” means motorways only because I often seen people crossing the road across a dual carriageway near my home.

Test subjects for autonomous cars in London

Can you imagine what that Nissan will be like on motorways in the near future? An autonomous car which can think quicker than a human will be able to drive closer to the autonomous car in front, which could save fuel (a bit like a car peloton) and speed up journeys. Why would you need a speed limit?

The other two areas which affect car insurance are security and the policy holder, or ownership.

Why do we still own cars?

I don’t know. In the UK, the typical car is parked 96.5% of the time. We only use our depreciating cars 3.5% of the time. Imagine an alternative, where you order a self-driving car to come to your location for a specific time. You could sync the schedule with your calendar, or ask the car to follow you around by tracking your smartphone.

Once car ownership has ceased, we’ll be less concerned about car theft. In fact, why would you steal a car that is programmed by others to drive certain routes? It might not even have a steering wheel.

So if the three main car insurance risks of car ownership, accidents and theft are all likely to decrease, where does this leave car insurers? It’s likely that insurance policies will move from a consumer model to a group policy. Volvo have stated that they intend to accept full liability of accidents involving their cars, and we’re likely to see other manufacturers follow this model. Or we might receive insurance through another membership such as your car breakdown company or another association. You might even receive car insurance in your Microsoft Office 365 subscription.

I’ve only outlined some thoughts about insurance for automated cars. If you’re interested in other side effects of automated cars, I encourage you to read Benedict Evans’ excellent post.

Sources for this article

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/588773/quarterly-estimates-july-to-september-2016.pdf

https://www.iamroadsmart.com/media-and-policy/road-safety/road-safety-statistics

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/588591/illegal-alcohol-levels-second-provisional-2015.pdf

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/533294/rrcgb-main-results-2015-infographic.pdf

http://researchbriefings.files.parliament.uk/documents/SN02198/SN02198.pdf

https://www.theguardian.com/technology/2016/jul/01/tesla-driver-killed-autopilot-self-driving-car-harry-potter

https://www.google.co.uk/finance?q=tesla&ei=hKQQWYHZDMePUKTouqgL

https://techcrunch.com/2017/01/13/nissans-first-european-self-driving-car-trials-begin-on-london-roads-next-month/

http://www.reinventingparking.org/2013/02/cars-are-parked-95-of-time-lets-check.html

https://www.theregister.co.uk/2015/10/13/volvo_to_accept_full_liability_for_crashes_involving_driverless_cars/

Leave a Reply

Your email address will not be published.