Road Safety News

Government and fleets should only buy cars with AEB: Thatcham

Thursday 14th May 2015

Thatcham Researchis calling on fleet operators and the new UK Government to make a “decisive shift” to buying only cars with Autonomous Emergency Braking (AEB), on the back of a paper released yesterday which shows AEB “reduces rear-end crashes by 38%”.

The paper*, from Euro NCAP and ANCAP, the independent safety bodies for Europe and Australasia, shows that low speed AEB technology leads to a 38% reduction in “real-world rear-end crashes”, with no significant difference between urban and rural crash benefits.

The publication, which pools real-world data from five European countries as well as Australia, concludes that low speed AEB technology needs widespread fitment for maximum benefits. The low speed option normally consists of an automatic brake function that operates effectively at speeds up to 50km/h.

Andrew Miller, chief technical officer of Thatcham Research and president of Euro NCAP, said: “Clearly, at this level of effectiveness, low speed AEB is potentially a hugely important active safety technology, and widespread fitment through the vehicle fleet should be encouraged in the interest of improved vehicle safety – a key recommendation since 2014.

“These findings strongly support Euro NCAP’s decision to make AEB technology a key discriminator in the safety rating of new vehicles – which is reflected in insurance group ratings. And from next year, Euro NCAP will include in its assessments AEB systems that recognise pedestrians.

“We would like to see the new Government take a lead in the UK – just as Swedish and Australian Governments have done – and mandate only AEB-equipped cars across its own fleet and champion the take-up of AEB across the all-important fleet market.”

*Footnote: the paper ‘Effectiveness of low speed autonomous emergency braking in real-world rear-end crashes’, has been published in the journal ‘Accident Analysis & Prevention’ and can be purchased from Science Direct.


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:

I would not expect that all improvements in safety tend to be eroded over time otherwise we would not have seen such a reduction in total casualties over the years. I do accept that there are some improvements, perhaps seat belts are one, where users take extra risks to balance out the reduced risk of injury gained by the presence of the improvement. Thankfully a majority of us appear to exhibit self preservation tendencies. Now where is my wing-suit.......
NIck, Lancashire

Agree (1) | Disagree (0)

I talked recently to the owner of a VW Golf fitted with AEB. He had been in the outside lane of a dual carriageway, negotiating a slight left-hand bend while overtaking another car on his left. There was a lane in the central reservation for the use of drivers who wished to turn right across the dual carriageway, and there was a stationary vehicle in that lane waiting to turn. As he approached that situation to his utter amazement his car performed an emergency stop, or so he claimed. Clearly the system had detected the stationary car and predicted that he was about to hit it. Had there been a vehicle following him, there would have been a real chance of his being rear-ended.

AEB and any other systems will have limitations, but I suspect that drivers will consider them foolproof and operate with even less attention being paid to their driving than at present. Even if they do not have their drawbacks, they simply cannot override the laws of physics. As has been stated, any potential improvements in safety tend to be eroded over time by their being used to increase performance in other areas.
David, Suffolk

Agree (0) | Disagree (1)

Nick, it would have been nice to see the Thatcham report to find out whether the results they quoted included those vehicles that had the Mobileye warning system as well as the AEB installed.
Duncan MacKillop. No surprise - No accident.

Agree (1) | Disagree (0)

If this technology removes up to 38% of rear end shunts then that is a good start. We need to identify ways of removing the 62% or more that remain. I think my earlier post should have allowed for some collisions caused by human errors not being able to be overcome by this technology - probably because it is low speed AEB technology?

This AEB technology actually stops the vehicles without input from the driver (preventative technology?) so is that different to the way the technology you described affected your driving by you changing the way you drove to avoid warnings? Or does the AEB emit warnings before it kicks in?
Nick, Lancashire

Agree (1) | Disagree (0)

According to the figures in the report the technology doesn't work in 62% of rear end crashes so clearly the 'human error' has not been removed entirely from the equation if at all.

I had an interesting exchange with Mr Yaakov Sandler of Mobileye technology (the people behind the AEB equipment) and although it will break Nick's post size limit I hope he will indulge me and replicate the exchange.

Duncan (said):
Accidents are not 'caused' by anything, they are caused by everything! It's the overall variation in the system that does the damage and there are two ways of addressing the problem. The first is to reduce the variation in the system and the second is to increase the ability of the system users to sucessfully manage the variation that exists. The various electronic interventions discussed clearly don't reduce the variation in the system so they must work on the level of the system user's ability.

If there is a reduction in accidents with vehicles fitted with a technology then it is not the technology itself that has made the difference but the effect that it has had on the user. My own experiences with the development of an eco-driving app were that the rate the alarm sounded reduced rapidly with use which showed that the user must have learnt something external to the device itself. If the user had not learnt anything then the rate of alarm sound would have remained constant.

Knowing what the user has learnt then is key to understanding the effectiveness of an intervention such as the Mobileye. I suspect that with Mobileye as with my Eco-Driving app the user has learnt to be a better predictor of unfolding events which is actually what has caused the reduction in the number of collisions. The better the user is at prediction then the less likely it is for the alarm or warning to be triggered and I suspect that the data traces from the Mobileye units will confirm that. On initial installation there will be a significant number of warnings, but this number will reduce significantly over time until there are hardly any warnings at all.

As all accidents are the result of prediction failure then any reduction in the number of accidents must be due to an improvement in predictive ability. If the figures quoted are correct then the Mobileye must be an excellent predictive training device rather than being a preventative device in its own right.

Yaakov Sandler (replied):
Duncan you are 100% correct. Mobileye has been integrated with more than 40 different types of Fleet Management Systems. Once integrated with a Fleet Management System the fleet manager can see how many of the different type of warnings his drivers are getting over a period of time, and tests have shown that there is a significant reduction in the number of warnings over a 6 week period.

Duncan MacKillop (replied):
That seems to confirm our hypothesis that training people to be better predictors of future events will have a significant effect on the accident rate. Now all we have to do is to convince the road safety industry that it is prediction and not 'behaviour' that is key to accident reduction and by this time next year we'll all be millionaires!
Duncan MacKillop. No surprise - No accident.

Agree (2) | Disagree (1)

I take both of your points of view and wish to include within that framework several facts of life. That human beings are at times [a] complacent [b] lazy [c] apathetic. To that end they will accept the new innovations such as AEB and therefore consider that they are a safe distance behind another vehicle anticipating in the worse case scenario that the AEB will protect them at all times from a impending collision. That will not, I fear, be the case.
Bob Craven Lancs...Space is Safe Campaigner

Agree (1) | Disagree (0)

Leaving aside a potential discussion whether luck comes into play for a driver to avoid a collision or whether it is the result good driving, the point you raised was that you think the incidences of rear-end shunts would revert to former levels if all vehicles have AEB fitted.

I expect the levels would reduce due to the removal of the human error aspect of driving into the back of a vehicle. Do you disagree with that?
Nick, Lancashire

Agree (3) | Disagree (0)

Will the conventionally equipped car following the AEB equipped car be so lucky?
Duncan MacKillop. No surprise - No accident.

Agree (0) | Disagree (3)

Thanks for that Duncan.

However, I would expect that the difference now is the AEB reduces the "human error/inability to operate the brakes in time" aspect of the collision, therefore the following vehicle is more likely to avoid the rear-end of the car in front?
Nick, Lancashire

Agree (3) | Disagree (0)

To give it a historical reference Nick, I think that there was a broadly similar situation with the introduction of cars fitted with disc brakes. Now we've all got them the collision rate in shunts is probably much the same as it was before their introduction. Collisions are much more survivable now though, what with all those passive safety features that cars are fitted with nowadays.
Duncan MacKillop. No surprise - No accident.

Agree (0) | Disagree (6)


Please can you explain your common sense rationale on how the crash rate would revert to the previous rate once all vehicles are fitted with AEB?

My initial thoughts are that if a rear-end shunt is less likely for one car fitted with AEB then if the car behind is similarly equipped then that would have a further positive effect? This might even stop Google Cars getting rear-ended?
nick, Lancashire

Agree (5) | Disagree (0)

AEB systems do not make the driver more reliant on the system, most drivers already follow too closely. I've spend a long weekend demonstrating these systems and can assure you that the driver will not follow more closely because AEB is fitted. You end up very, very close before the system kicks in. Let's not confuse opinion with practical experience and facts. "Only a fool brakes the 2 second rule."
Stephen Muirhead South Yorkshire

Agree (5) | Disagree (1)

Well said Duncan. An example is the way that government policy on safety was usurped by manufacturers. I remember in the 70s when the capacity of learner motorcycles was reduced from 250cc. to 125cc. One of the greatest mistakes ever made when it comes to motorcycle road safety as bikes up to and including the old 250cc were generally only just capable of maybe 70 mph. When the legislation changed to the lower capacity we found that there were more accidents as those lower capacity machines were just as fast as the previous ones but of a lesser weight and a higher centre of gravity when ridden thus making the bike and rider less stable and with tyres and brakes that were initially poorer.

The British motorcycle base was washed away and foreign power bikes took their place and maximum speeds for the new bikes was a competition by manufacturers and a definite consideration for prospective buyers. Since then 125c motorcycles manufacturers have continued to developed greater power from lower brake horse power and some can now reach speeds well in excess of 80 mph.
Bob Craven Lancs...Space is Safe Campaigner

Agree (5) | Disagree (1)

Pretty well spot-on in your last comment Bob.

The Law of Stretched Systems says that "Every system is continuously stretched so that it operates at its capacity".

Essentially what this means is that if you provide an enhancement to the system that has been designed to increase the safety margin then people will use that extra margin to improve performance instead. In this case if the driver knows that the AEB system will be much better at braking than they are then they are more likely to put themselves in situations that might require that higher level of braking.
Duncan MacKillop. No surprise - No accident.

Agree (4) | Disagree (5)

I for one would have liked flashing brake lights on my motorcycle for the last 5 or 6 years but always considered them, as did others, as not being legal. Am I now able to modify my motorcycle to legally have flashing brake lights or has there been no change in the laws to accommodate them? As regards rear enders as said it works up to 30 mph approx. but not after that.

I would much prefer the advice to leave at least the full stopping distance between myself and any other vehicle on the road. That means one lamp post or more apart if possible in 30 and 40 mph areas, two lamp posts apart on rural main roads and the distance between the small marker posts on the side of a motorway, they being 100 meters apart and about equal to the stopping distance at 70 mph. Quite a simple thing for drivers to remember I should think.

Back to the flashing lights. Does this new system take into account smaller objects such as dogs or children? I would like to know. Perhaps someone has bought the paper and can come back with an answer.

Knowing that one's car has this automatic system may to my mind lead drivers to be over confident in driving too close to the vehicle in front without any regard to the greater safety accommodated to all road users by being at least the correct distance behind and the increased visibilities that it would afford to not only the driver but to all other road users. This system encourages dangerous driving.
Bob Craven Lancs....Space is Safe Campaigner

Agree (3) | Disagree (4)

That's very interesting to know Lee. Now I wonder how many people out there in the real world actually know what a flashing or higher intensity brake light means?
Duncan MacKillop. No surprise - No accident.

Agree (3) | Disagree (1)

AEB cars are as likely to get rear ended if the driver behind is not paying attention to the road as much as car without AEB.

But most cars with AEB have a flashing brake light system so when you're doing a very heavy braking or the the AEB system has had to auto brake the car brake lights flash to warn the people behind.

Most modern cars even without AEB have flashing brake lights or make the Brake lights more then twice as bright when a manual heavy braking or ABS braking is in operation, but personally flashing brake lights is the best way as it's more noticeable.
Lee, Warrington

Agree (2) | Disagree (2)

Does the paper show what the crash rate is where both the vehicles involved are equipped with AEB? Common sense would suggest that once every vehicle was equipped with AEB then the crash rate would revert to what it was before they were so equipped. Also common sense suggests that as AEB use increases the number of crashes where the AEB equipped vehicle becomes the crashee rather than the crasher will also increase.

Of course all these questions may well be answered in the paper, but unlike those research papers published in the aviation safety industry this one is hidden behind a paywall.
Duncan MacKillop. No surprise - No accident.

Agree (7) | Disagree (4)