Road Safety News

The ‘driverless cars’ debate: the verdict

Wednesday 29th November 2017

There was a clear shift in opinion when two experts went head-to-head at the 2017 National Road Safety Conference to debate the road safety implications of ‘driverless cars’.

The two participants, Dan Phillips, GATEway project manager at the Royal College of Art, and Christian Wolmar, writer and broadcaster, debated whether driverless cars will be good or bad for road safety.

Having won the toss and elected to bat first, in his opening remarks Christian Wolmar described driverless cars as “a future that will never happen”, adding that “all cars will have to be driverless” (to deliver safety benefits) and “the transition period to driverless cars will be endless”.

As an example of the “insuperable barriers” to the idea, he said because driverless cars “will be programmed not to kill people...they will have to stop”, adding that “pedestrians won’t let driverless cars through” and “‘bad people’ will go out of their way to stop them”.

In conclusion he said “forcing us into driverless pods is a political decision” and urged supporters of the technology to “be careful what you wish for”.

In response, Dan Phillips pointed out that “a significant majority (of the population) have a positive attitude to driverless vehicles” because they will be “cleaner, greener and safer”.

Dan Phillips said that “94% of crashes involve human choice or error” - adding that “fatality rates in manufacturing are falling as automation increases”.

Using specific examples, he explained how Tesla’s crash rate has “dropped by 40% after introducing anti-crash software”.

Again referencing that “almost all road deaths are down to driver error”, Dan Phillips said “driverless vehicles won’t get drunk, angry or be distracted”.

He concluded by pointing out the social benefits of driverless cars, asking delegates to “look at what people do (with their time) on public transport and as passengers in cars”.

What did the audience think?
The audience was asked, before and after the debate, whether they thought driverless cars would be good or bad for road safety.

While the majority voted ‘good’ in both polls, there was a significant shift in favour of ‘bad’ in the second vote, as shown in the graphic below.

Before the debate:
Good: 51%
Bad: 11%
Not sure: 38%

After the debate:
Good: 44%
Bad: 26%
Not sure: 30%

Category: Autonomous vehicles.




Comment on this story
Report a reader comment

What's your view - comment on this story:

I confirm that I have read and accept the moderation policy and house rules relating to comments posted on this website.
Your comment:
Your name and location:
Your email:

Some individual programmers may design/write software that's not following the best road safety practice. However the peer review, team thinking approach and design audit processes will flag up weaknesses.

I wonder when the new road safety accreditation for writers of software for autonomous vehicles will emerge - and who will be qualified to accredit anyway?
Pat , Wales

Agree (2) | Disagree (0)

That's assuming Peter, that those behind the programming of the autonomous vehicles are themselves better than the average driver, otherwise the status quo would be maintained, as all that would happen is that average human drivers will be replaced by vehicles programmed by average drivers, with the same propensity for collisions.
Hugh Jones, Cheshire

Agree (2) | Disagree (5)

Change cannot happen if we maintain the status quo. Pedestrians, bike riders and AV's will be required to work collaboratively. Big change. Difficult but current situation is not sustainable and not of our choosing. Result of the before/after poll reflects that CW tapped into the misinformed belief of most vehicle drivers that they are good drivers, indeed better than the rest. AVs will challenge that so don't expect objectivity.
Peter Treadgold

Agree (2) | Disagree (2)

Nick: I think David's concern (which I share) is that those programming the vehicles may not necessarily be that competent themselves behind the wheel, so their own shortcomings and bad habits may unwittingly come to be part of the autonomous vehicles' modus operandi. Human error behind the wheel may well be removed, but only to be counteracted by the introduction of human error further back up the 'chain of command' as it were.

For the record, I don't have a problem with autonomous vehicles being made available for those who want or need them - I just don't want them forced upon us. If I ever see one on the road, I shall watch its progress with interest...from a safe distance!
Hugh Jones, Cheshire

Agree (3) | Disagree (1)

David, why would a programmer need to be a skilled driver? They are moving an inert mass from one place to another whilst avoiding other moving (and stationary) objects. Removing the need for individual humans to be skilled (or indeed able) enough to do this by driving themselves is surely the point? Do all flight simulator designers/programmers hold the highest level of pilots licence?

There are obviously further considerations to account for but I think the point is valid?
Nick, Lancashire

Agree (4) | Disagree (1)

Hugh, it is probably true to say that "they" are regularly operating on the roads in the real world and have been doing for quite a while.

I do not think for one minute that they will be a like-for-like replacement of today's cars and vehicles. In cities, as population densities increase, the need and desire to own a car type vehicle will probably decrease as their efficiency in moving around decreases. Mass transit systems may be seen on the streets replacing queuing single occupant vehicles together with more space for pedestrians and cyclists to move around.

Many other people, including Christian Wolmar pointed out that autonomous vehicles may struggle to move in high density pedestrian areas. Let's not spend precious resources of time and money trying to replicate a system close to breaking point. Sometimes we have to accept change.
Nick, Lancashire

Agree (3) | Disagree (0)

Whilst these vehicles may well be programmed not to do anything 'wrong' - which I would imagine would not be difficult to do - that doesn't mean they will necessarily be doing everything 'right', which I would think would be much harder to program and an objective which may well have been under-estimated and unfortunately will only become apparent when as and when they begin to regularly operate in the real world.
Hugh Jones, Cheshire

Agree (3) | Disagree (0)

In a Channel 4 program that aired last Sunday, it was claimed that there are only 50-60 people in the world today with the skills to write the computer code for the Artificial Intelligence that is needed for autonomous vehicles to learn. What are the chances of any of those computer programmers being someone who understands driving and is a skilled driver?
David, Suffolk

Agree (3) | Disagree (2)

The opinions given on both sides of the driverless cars debate shows how little we actually understand about the driving process.
Duncan MacKillop. No surprise - No accident.

Agree (3) | Disagree (7)

I thought Mr Wolmar came across as a realist who had thought it through, whilst Mr Philips an idealist who perhaps hadn't.
Hugh Jones, Cheshire

Agree (4) | Disagree (5)