Motorists doubtful over driverless car technology, survey finds

IAM research also shows drivers split right down the middle on merits of self-driven cars' adherence to speed limit

by Simon_MacMichael   November 20, 2012  

Car wheel at speed copyright Simon MacMichael.jpg

Two thirds of motorists say they’re unsure about the merits of driverless cars, while four in ten state that they would never even consider driving one, according to a survey from road safety charity IAM. Opinion regarding such vehicles being restricted to driving within the speed limit is split right down the middle – half of motorists polled say it is a positive feature, but half see it as a drawback.

The finding that most people who actually drive cars are dubious about that task being given to a machine is perhaps understandable - for those who are most attached to their vehicles, it's presumably a bit like asking turkeys to vote for Christmas.

The poll of 1,111 motorists comes at a time when development of the technology is gathering pace – as we reported last month, California’s Governor Jerry Brown signed a bill that paves the way for self-driven cars to take to the state’s roads, while a vehicle being developed by Google has passed 300,000 miles in testing without mishap.

Professor Garel Rhys, who heads up Cardiff University’s Centre for Automotive Industry Research has said because the majority of road traffic incidents are due to, self-driving cars would lead to roads becoming safer. 

Despite that, only one in three respondents to IAM’s survey believed that taking away human interaction would improve road safety, and three quarters said that instead of trying to improve automotive technology, it would be better to focus on ways to encourage safer driving habits among motorists.

Simon Best, chief executive of IAM, said: “The presence of driverless technology in every car is still many years away. In the meantime, more should be done immediately to improve driver standards and deal with the most common human errors through better training, as well as incentives by the government and insurance companies.

"Of course technology has a huge role to play in road safety, but as long as there are cars on the road people will want to drive them. What we need to aim for is first class drivers operating first class vehicles.”

Other findings of the research include:

Around half of respondents (500 of 1,088) feel that driverless cars are a good initiative for the future.

Half of motorists don’t think that driverless cars will become popular.

56 per cent do not think there is a possibility that driverless cars will be the norm
within the next ten years.

98 of 1,088 people think that driverless car technology is irresponsible.

22 per cent of respondents would use a driverless car.

Over half of motorists think that automated systems should take control to prevent a crash.

92 per cent of people like the fact that with driverless car technology, the car behind would not be allowed to drive too close to you.
 

21 user comments

Oldest firstNewest firstBest rated

The Google car (an automated self-driving Toyota Prius) has had well over 100,000 miles on the clock, driving the congested streets of the cities of California. The only accident it has ever had was when a human was driving it and not the computer...

I suspect people don't like the idea of not being in control and not having the option to speed...

posted by pk1234 [3 posts]
20th November 2012 - 11:43

3 Likes

A driverless car would also potentially remove the fun of telling other road users half-remembered and downright false bits of the Highway Code

posted by Sadly Biggins [266 posts]
20th November 2012 - 12:04

5 Likes

78% of respondents would prefer to drive themselves.
92% of respondents would prefer everyone else to hand control to the computer.

Once again this study has proved that 90% of the population rate their driving ability as "Above Average"...

Me? I can't wait - bring on the Robots!

posted by Mr Will [91 posts]
20th November 2012 - 12:14

7 Likes

I quite like driving, but anyone who would rather manually drive the daily commute/to the shops etc rather than let a computer take over while you relax and read a book/chat with the family must be friggin nuts.

Fact is though, journey times would only improve markedly once automated vehicles were mandatory - any less than that and the necessary margin for error (and litigation-happy society) would mean that the computers would be ever-cautious.

Last night I would have considered trading a very loud baby for a really nice bike.

posted by notfastenough [3476 posts]
20th November 2012 - 12:30

4 Likes

"92 per cent of people like the fact that with driverless car technology, the car behind would not be allowed to drive too close to you."

And what do most people do themselves...?

Basically +1 for Mr Will's post. This survey basically shows just why they should come sooner rather than later.

posted by Chuck [435 posts]
20th November 2012 - 13:02

2 Likes

I really, really can't wait until drivers are taken out of the loop. It will mark a step change in road safety. Once they start to appear on the road in any numbers I expect the accident/collision statistics show will soon show how safe they are (or how dangerous human drivers are depending on your view).

Once the stats show that you're 10, 20, 50, whatever times more likely to cause an accident if you drive yourself, then legislation and insurance will soon get rid of the human menace!

I only wish those who think that helmets/hi-viz are the answer would put their time and effort into driverless cars.

posted by qwerky [150 posts]
20th November 2012 - 13:31

5 Likes

Are you kidding? I most definitely have more faith in a computer than in some moron who dropped out of high school and is talking on their cell while trying to overtake me on a country road. How on earth would driverless cars not be safer?

new-to-cycling's picture

posted by new-to-cycling [47 posts]
20th November 2012 - 13:50

6 Likes

notfastenough wrote:
I quite like driving, but anyone who would rather manually drive the daily commute/to the shops etc rather than let a computer take over while you relax and read a book/chat with the family must be friggin nuts.

Fact is though, journey times would only improve markedly once automated vehicles were mandatory - any less than that and the necessary margin for error (and litigation-happy society) would mean that the computers would be ever-cautious.

I agree, I can't wait, long journeys where I'm doing the driving is tiring and it will also increase motorway capacity with automated cars being able to drive closely to one another at a uniform speed safely.

I'm a bit nervous about riding around on the roads with these machines on the road as they might struggle with rider behaviour just as much if not more that the planks we have behind the wheel at the moment, perhaps separate cycle lanes are required...

posted by s_smith [16 posts]
20th November 2012 - 14:01

5 Likes

In fact, my commute is too far to ride - I want a self-driving van with the back converted into a turbo-training suite for optimal use of the driving time! Big Grin

Last night I would have considered trading a very loud baby for a really nice bike.

posted by notfastenough [3476 posts]
20th November 2012 - 14:27

7 Likes

Two thirds of motorists say they’re unsure about the merits of driverless cars, while four in ten state that they would never even consider driving one

That's ok, they're, er, driverless.

Doctor Fegg's picture

posted by Doctor Fegg [141 posts]
20th November 2012 - 14:34

4 Likes

I've actually been behind the wheel of a driverless car that the TRL put together a couple of years back. If you ask me, it was dangerous.

OldRidgeback

posted by OldRidgeback [2308 posts]
20th November 2012 - 15:48

9 Likes

Quote:
Professor Garel Rhys, who heads up Cardiff University’s Centre for Automotive Industry Research has said because the majority of road traffic incidents are due to, self-driving cars would lead to roads becoming safer.

Majority of incidents are due to … what? Human error, I assume.

I would expect most people to be very sceptical that autonomous cars would be safe, but this scepticism will vanish once production standard cars are demonstrated.

two wheels good; four wheels bad

posted by cat1commuter [1380 posts]
20th November 2012 - 17:42

4 Likes

Drive to the shops? Why would you drive to the shops? You just ring up, tell them car is on the way and what you want, they pop your order in the boot and the car comes back ;- ))

posted by horizontal dropout [170 posts]
20th November 2012 - 19:19

6 Likes

"Drive to the shops? Why would you drive to the shops?"

exactly -and why not get junior a car to take him to school why mum gets taken in her car to the spa before work rather than doing the school run, oh and a car to take the older daughter to meet her friends at the shopping mall - no more dad's taxi

so in affluent areas possibly many more vehicles - so though cars "flow" (hurrah) congestion/energy consumption may not reduce. Of course taxes/road pricing could deal with this but this would punish the less well off getting to their inconviently located employment and not affect those that can pay with no impact on behaviour - suppose could insist that cars pick up extra passengers from "car share stops" since car knows where it is going and the technology exists to match destinations
...oh hang on that sounds like a private motorist funded public transport system - better start again

antigee's picture

posted by antigee [174 posts]
20th November 2012 - 21:53

4 Likes

I wouldn't trust any vehicle to drive itself quite honestly. Human intervention is paramount imo.

posted by Karbon Kev [682 posts]
21st November 2012 - 9:57

4 Likes

OldRidgeback wrote:
I've actually been behind the wheel of a driverless car that the TRL put together a couple of years back. If you ask me, it was dangerous.

Can you tell us more? Companies have thus far been fairly quiet on the detail of how this actually works. What was your experience?

Last night I would have considered trading a very loud baby for a really nice bike.

posted by notfastenough [3476 posts]
21st November 2012 - 11:07

7 Likes

notfastenough wrote:
OldRidgeback wrote:
I've actually been behind the wheel of a driverless car that the TRL put together a couple of years back. If you ask me, it was dangerous.

Can you tell us more? Companies have thus far been fairly quiet on the detail of how this actually works. What was your experience?

Ok, it was built in partnership with Ford and a few other firms and based on a Ford Explorer. It was intended as a comparatively low-cost demonstrator using available technologies.

The vehicle used GPS navigation so it knew where it was and you could prgram it to follow a route and get from A-B. Some of the other people were impressed, and thought it might be a boon for ageing drivers. I was horrified at this suggestion as it had some serious faults.

Firstly, there was no forward scanning radar (also available technology), so any driver would have to be alert to jump on the brakes if anyone got in the way, such as a cyclist or pedestrian. It steered and drove - all the peron behind the wheel had to do was apply the brakes should an unanticipated hazard appear.

My feeling was that the technology would allow the vehicle to drive itself and the person behind the wheel could very easily become complacent and let the vehicle do everything, with potentially disastrous results, should anything or anyone get in the way. The problem is forward scanning radar technology is as yet limited by range and the sophisticated systems that would allow safe operation at speed are prohibitively expensive. Some firms, like Volvo and Honda, do offer scanning radar but these are very limited with regard to the vehicle speeds. However, unless you have a scanning radar that operates at up to 70mph and allows safe braking at this speed, driverless cars aren't safe. You can buy such technology, but it is prohibitively expensive.

Secondly, running a guidance system using GPS for location is troublesome in towns or hilly/mountainous regions where 'canyoning' occurs. This is when the GPS can't get enough poistional fixes on satellites due to obstructions from buildins, hills or mountains for example. Ever used a car GPS and it says "Unable to locate satellites?" Not such a problem if you've got a map, but if the entire directional capability of your vehicle relies on that tool, you're in trouble. Yes, you can have interial navigation as used in aircraft and industry, but it's expensive. When the Ford/TRL vehicle lost satellite, it would slow to a crawl and rely on stored data. I did point out to its developers that there are loops in the roadway it could use for example, but these do not use standard systems between the US and Europe so you'd require different technologies. And those are only any good anyway on major routes with the loops installed. I also pointed out to the engineering team responsible that there are autonomised vehicles operating successfully in the mining industry that they could have taken a look at - but the team hadn't heard of them and so had tried to design their own wheel after the fact (and cheaply and without key safety features) so to speak.

Driverless cars will rely on new and more accurate GPS signals. Yes, the European Galileo system will come on-line eventually (at staggering cost and questionable value for money benefit for us European taxpayers). But can we fully rely on its GPS signals, and that of the Russian Glonass or US GPS or the Chinese system (I forget its name)?

I could go on, but I'll spare you more technicalities.

I wasn't impressed and thought that if this was the best the automotive sector could do, it was by no means good enough. The team hadn't bothered to do research into other autonomised vehicles and could've spared themselves a lot of time and come up with something a lot better. I did say so, and I don't think the engineers were veary pleased when I pointed out the product had several serious flaws.

But without a lot of expensive technology, specifically a military grade forward scanning radar and an aircraft's intertial guidance system, I don't see how you can make a safe driverless vehclce.

OldRidgeback

posted by OldRidgeback [2308 posts]
21st November 2012 - 11:38

4 Likes

Hmm, lot of interesting stuff there, thanks for replying. The Darpa Grand Challenge in the USA (and by extension, the Google self-drive project) does seem to have overcome many of these. Clearly, the forward-scanning radar is an obvious requirement, since without it all you've got is a GPS connected to automated steering/pedals/gears. My thinking is more about the kind of object recognition software that needs to be inherent in the sensors. Traffic lights - (unless they broadcast a new data signal to indicate the light colour for each lane/direction), a cyclist, a cyclist with their arm out and moving across lanes, temporary roadworks with a man stood in the middle with a red/green stop/go lollipop. Roundabouts - object detection from the side, cars coming from the right but that change from a lane where they don't conflict with my use of the roundabout, to a lane where if I move out there will be a collision... ad infinitum.

Some cars have adaptive cruise control; that means that while maintaining a given speed, a forward-facing radar (with a display that shows a range of about 100m) will slow the vehicle if, for example, a car moves into my lane ahead of me. Obviously this cannot anticipate events, but this drawback is taken to the extreme. For example, if the car in the adjacent lane begins to indicate an intention to move into my lane, it doesn't take a genius to realise what is going to happen next. Yet the radar cannot see this, and waits until the car completes the move before slowing down. Similarly, it might be, to a human, patently obvious that someone is about to run out in front of a car, but a human driver can take pro-active measures, whereas the car is purely reactive.

That said, if they can crack this stuff, then in theory I think it would be great.

To take this a step further, one of my problems with the concept of flying cars is that drivers today have enough problems working in two dimensions, so introducing a third would seem barmy. However, 100 metres up, many of these hazards disappear, and automated flight planning and collision avoidance systems would take care of much of the rest - maybe that's a better idea...

Last night I would have considered trading a very loud baby for a really nice bike.

posted by notfastenough [3476 posts]
21st November 2012 - 14:13

4 Likes

Have you seen how much the vehicles in the US desert challenge cost? Let's just say, you don't build one for less than $2 million. And there's so much technology in there, the price isn't going to come down so much.

OldRidgeback

posted by OldRidgeback [2308 posts]
21st November 2012 - 21:35

7 Likes

Do you guys really believe that once the roads are full of automated cars you'll be allowed to cycle on them without special transponders, laser rangefinding systems and speed limiters? Careful what you wish for...

posted by AlanD [12 posts]
21st November 2012 - 22:36

7 Likes

OldRidgeback wrote:
Have you seen how much the vehicles in the US desert challenge cost? Let's just say, you don't build one for less than $2 million. And there's so much technology in there, the price isn't going to come down so much.

Very true, although the same was true of early computers.

Aland, you may have a point. It would also 'be the cyclists fault' if for some reason your gadgets stopped working and an automated vehicle hit you...

Last night I would have considered trading a very loud baby for a really nice bike.

posted by notfastenough [3476 posts]
22nd November 2012 - 9:39

3 Likes