Technology

Self-driving vehicles may not be the road toll saviour


Earlier this decade the AA believed that New Zealand’s road toll could be less than 200 by 2020.

Sadly, in the last five years we have seen the long-term reduction in road deaths reverse and last year 380 people lost their lives. This tragic turn-around has reignited debate about what can be done to prevent the often senseless deaths and injuries the come from crashes? More and more often self-driving cars are being promoted as “the answer”.

Car makers and technology firms are putting huge work into producing vehicles that can drive without human control. The vision is that this will remove human error from driving and therefore greatly reduce deaths and injuries on our roads.
But if we pin all our hopes on self-driving cars stopping the nearly 400 deaths and 30,000 injuries from crashes currently happening then we are going to be waiting for a long time to make a real difference.
Amongst all the hype and bold predictions around self-driving cars there are also a lot of difficult obstacles still to be overcome. A few of the biggest are:

1. Will there be limitations to when and where vehicles are capable of ‘self-driving’?

Driving in the central city, or in a suburban cul-de-sac, or on a gravel road are all very different situations. It will be an enormous challenge to create a self-driving vehicle that can seamlessly negotiate every environment and the variety of signage, road marking and traffic complexity it may encounter. Computers, unlike humans, are not good at dealing with ambiguity. An American motor writer illustrated this when they were in a self-driving car that encountered roadworks and a worker with a stop-go sign. Without human intervention the car would have sat there for hours because there was a person on the road in front of it and its programming forbids it from crossing over into an oncoming lane.

2. Will people want to hand the driving over?

Only 12% of AA Members say they would currently choose a self-driving vehicle over one they drive themselves. That figure will grow over time but among car manufacturers there seems to be an assumption that if you build them, people will want them. This is actually a crucial issue because for self-driving vehicles to deliver maximum safety benefits the entire fleet needs to be made up of them. A situation where half the vehicles are self-driving and half are being driven by people would create its own safety challenges. What if a substantial number of people decide they do not want to use a self-driving car?

3. How quickly will true self-driving technology develop?

Self-driving vehicles currently being trialled on the roads tend to operate in tightly controlled conditions and require human oversight and responses. We are a long way from mass production of vehicles that people could jump into on any street and be taken wherever they want. Progress in self-driving cars will likely be messy and full of trial and error. Studies are also showing that it takes a long time for someone who isn’t driving to be able to react appropriately if suddenly required to. This has led some experts to now think that until autonomous vehicles can be made that never require human intervention they should not be on the roads.

4. The time lag

The average age of cars on New Zealand roads today is 14 years old. If that situation continues then you would expect that the cars coming off the production lines this year will be the average vehicle on our roads in 2032. And none of them are self-driving.

Along with the technical challenges yet to be solved, there are also serious legal, ethical and financial issues that are going to need to be dealt with around self-driving vehicles like who has responsibility if a self-driving car is involved in a crash. Another key question will be whether people need a driver’s licence to travel in a self-driving vehicle? If the car does everything itself that seems unnecessary but if people are sometimes needed to take control they will still need to be able to drive.

We all want to see our road toll coming down and evolutions in vehicle technology are going to be a key part of that but while the technological advancements develop the AA wants more action now.

The new Government has signalled a commitment to improving road safety and announced a package of roading upgrades but in the years ahead we need to be looking at taking action on a mass scale and investing much more.
Targeting the highest risk roads for safety improvements, installing many more crash barriers on our highways and getting more people into more modern vehicles with side-curtain airbags and ESC won’t grab headlines like a car with no steering wheel. But these improvements work. They mean that even when drivers get it wrong or do something dumb it is much less likely to cost lives. The AA wants more of these measures for road safety in the present rather than betting on technology being a silver bullet in the future.

If by 2050 we have all become passengers in self-driving vehicles and have a road toll of zero we will still have saved hundreds of lives and prevented thousands of injuries in the years it takes to get there. And if the future turns out to not quite be as rosy as some are predicting then the safety improvements we can make right now will be even more valuable.

Previous post
Next post
Backyard diagnostics
Read more
February 2018 petrol and diesel prices
Read more