AUTONOMOUS BRAKING SYSTEMS AND THEIR POTENTIAL EFFECT ON

WHIPLASH INJURY REDUCTION

Matthew Avery, Alix Weekes
Thatcham
United Kingdom

Paper Number 09-0328

ABSTRACT ACC and AEBS

The paper estimates the benefits of low speed ACC uses a radar unit mounted on the front grille
autonomous vehicle braking technologies (e.g. City of the car to sense the proximity and speed of
Safety from Volvo) on reducing whiplash injuries, vehicles ahead. This allows the functionality of a
and whether driver adaptation is likely. Potential standard cruise control system to be extended to
UK whiplash injury reduction and cost savings control braking as well as acceleration. The driver
associated with autonomous braking systems are can then let the ACC control acceleration and
calculated. Assuming standard fleet wide fitment, braking, and only has to provide steering input.
predictions show autonomous braking systems ACC is designed to work on motorways and dual
(City Safety) could annually prevent 263,250 carriageways and most systems are only
crashes, mitigate 87,750, and prevent 151,848 operational at over 30 km/h.
injuries, equalling nearly €2 billion savings in
repair costs and whiplash compensation. In driver ACC systems also have the facility to provide a
adaptation testing participants drove toward an warning to the driver if the car is at risk of a
inflatable target car at 15km/h without braking. collision. These warnings can take many forms
Responses were collected from 99 driver tests, including visual symbols or lights, audible beeps or
where the vehicle autonomously brakes preventing ‘bongs’, or a haptic tug on the seat belt.
impact. 11% of drivers braked instinctively when
approaching targets, and 95% of drivers stated they A further development of ACC is AEBS, which
would not rely on City Safety for normal driving, will automatically apply the vehicle brakes when
and understood that it was for emergency braking an imminent collision is identified. AEBS aims to
only. Feedback was also gathered from 11 drivers prevent the collision or to mitigate severity by
experiencing the system on thousands of kilometres reducing speed. AEBS functionality is known by
of normal UK roads. None reported either positive different names by individual manufacturers, such
interventions or false interventions. City Safety, an as Collision Mitigation Braking System (CMBS)
example of low speed autonomous braking by Honda, or Collision Mitigation by Braking
systems, shows huge potential for reducing crashes (CMbB) by Ford.
and whiplash injuries valued at nearly €2 billion in
insurance claim savings. Other current autonomous So both ACC and AEBS use radar sensors and
braking systems operating at higher speeds require show some potential for mitigating crashes, but
driver activation, and can only mitigate impact they are not designed to prevent crashes from
speeds. City Safety operates autonomously at low occurring completely. The potential effect for
speeds and can prevent collisions occurring reducing crashes and injuries may also be limited
completely, so no risk compensation issues are by certain HMI (Human Machine Interface) issues.
expected. The systems are only operational when activated by
the driver, and can be turned off easily if the driver
INTRODUCTION chooses. The systems also issue warnings to the
driver that they need to intervene to prevent a
Over the last few years vehicle manufacturers have collision. Since different systems issue different
been launching a wide range of primary safety types of warnings there is potential for confusion
technologies. These are technologies that are that might lead to either a lack of response from the
designed to prevent a collision from occurring by driver, or an inappropriate response, which limits
warning the driver to intervene, or to lessen the the effectiveness of the warning.
speed and severity of the collision by autonomous
vehicle braking. Some examples are Adaptive An example of a Low Speed Avoidance technology
Cruise Control (ACC), Automatic Emergency is City Safety, and that does not appear to have
Braking Systems (AEBS), and Low Speed these associated HMI issues. This uses LIDAR
Avoidance technologies.

Avery 1
(Light Detection and Ranging) sensors, which is an POTENTIAL CITY SAFETY
optical remote sensing technology that measures EFFECTIVENESS ESTIMATES
properties of scattered light (laser) to find range
information of a distant target (vehicle in front). Since City Safety is designed to prevent low speed
These LIDAR sensors are mounted behind the collisions, it shows potential for reducing not only
windscreen and scan the road ahead for these crashes and the associated repair costs, but
approximately 6m. In a situation with a likely also whiplash injuries and costs. The main focus of
collision, the system will pre-charge the brakes to whiplash injury reduction countermeasures has
give a faster response if the driver does brake. been with better seat design. Data indicates that
Should the driver still fail to brake in an imminent 75% of all crashes occur below 30 Km/h [1] with
collision situation, automatic braking power up to the front to rear end crash at intersections being the
5m/s2 is applied, and throttle control by the driver most prevalent. British insurers report a cost in
is disconnected. In tests at speeds up to 22 km/h excess of €3 billion annually in the United
undertaken by Thatcham a car fitted with the City Kingdom due to whiplash [2]. In Sweden 70% of
Safety system successfully prevented contact with all injuries leading to disability are due to whiplash
another car. At speeds of up to 30 km/h the system injuries [3]. According to Watanabe [4] et al.
is able to mitigate collisions by 50%. The system is 43.5% of all injuries from vehicle crashes were
active for speeds up to 30 km/h. To prevent drivers from rear impacts, and approximately 90% of these
from adapting their normal driving to the system injuries were to the neck. Whiplash is an AIS 1+
the design of the system is intended to give a injury and the vast majority of occupants who
harsh/unpleasant braking sensation. The system is suffer initial soft tissue neck injuries typically
not operational against on-coming traffic, and is recover fully, although around 10% of the
operational against stationery or moving traffic. occupants with initial neck injury symptoms
The system calculates that the driver is taking continued to have symptoms after one year [5,6,7].
evasive action if they give a large steering, throttle, However collision avoidance technology offers a
or brake input, and the system is therefore huge potential to avoid the sorts of collisions that
overridden by the driver. typically cause whiplash injuries.

By default the system is always turned on when the Based on dose-response models Kullgren [8] has
vehicle starts, so it is always on and able to activate made estimates of the effectiveness of City Safety,
to mitigate/prevent a collision. Once the system has which indicates a 60% reduction in injured
operated the driver is given a display notice, but occupants. It is only possible to make estimates of
there is no warning given prior to intervention of the effectiveness of the system for preventing
City Safety. It is not possible to give a driver crashes at present, with only a small number of
warning of a potential collision since there is not vehicles in the fleet fitted with the system. Once a
enough time available once a collision risk is greater number of vehicles can be found on the
identified. Because City Safety is always turned on, road in the real world it will be possible assess the
and because it has no warnings, the HMI issues effectiveness of the system in detail. However by
associated with ACC and AEBS are not problems identifying those typical crash scenarios where the
for City Safety. system can be expected to have benefit, it is
possible to make some estimates of the potential
The City Safety system was launched as standard at savings in crash reduction, both in terms of damage
the end of 2008 on the Volvo XC60, and it is and injury costs.
expected to be fitted on other models from Volvo
as well as other manufacturers. However it will still Potential Crash Reduction
be a number of years before enough evidence can
be gathered about the effectiveness of City Safety Although there are many well established crash
in the real world to form a conclusion as to its frequency databases, such as GIDAS or CCIS,
potential for crash and injury prevention. This most of the criteria for inclusion relate to serious
paper outlines estimates of crash reduction and cost injures and typically require Police involvement or
savings offered by City Safety. It also presents two tow-aways. When comparing these to insurance
preliminary studies that have aimed to investigate statistics it is clear that the total amount of all
whether drivers are likely to adapt their driving crashes is far higher than the established databases
habits to the City Safety system by relying on its report. The types of crash and direction of impact
crash prevention technology, with the risk that they also tend to differ considerably. Overall there is a
consequently negate any advantages offered by the huge amount of under reporting is present in most
system by paying less attention to the road. published crash data sets when considering
whiplash injuries or non-injury crashes handled by
insurers. For example the Department for Transport
reported 247,780 casualties on UK roads in 2007
[9], and yet there are around 2.7 million motor

Avery 2
crashes resulting in an insurance claim annually in
the UK [10].
Table 1.
Estimates of the effectiveness of City Safety for Summary of Estimated Crash Repair Cost
reducing all crashes (not just casualties) can be Savings from City Safety
generated from the insurance claims data. These
estimates assume a fleet wide fitment of City Crash Crash
Safety. According to analysis of motor insurance prevention mitigation
claims data, around 26% of claims are for rear-end % of crashes
75% under 25% over
impacts where one vehicle runs into the back of over/under
30km/h 30km/h
another [11]. This represents 702,000 crashes from 30km/h
the 2.7 million motor crashes that result in an Number of
insurance claim [10]. Many of these crashes occur crashes without 263,250 87,750
at intersections, junctions and traffic islands and driver braking
result from poor driver attention. Most of these Average crash
€3,736 €3,736
crashes occur at low speed (under 30km/h) in the repair cost
speed range where City Safety is active. City Sub-total repair
€983,502,000 €327,834,000
Safety is designed to specifically operate on rear- cost savings
end impacts, but it could also have a positive effect Total repair cost
€1.3 billion
in other crash types. Effectiveness estimates for saving
City Safety are therefore only focussed on the
front-into-rear impact scenario.
Potential Whiplash Injury Reduction
Research [12] has shown that in a front-into-rear
collision situation 50% of drivers will respond by Whiplash is a high cost burden to both the motor
applying braking. When City safety detects that the insurers, all those who purchase motor insurance
driver is braking it will disengage since the driver and the wider society in general. Costs in excess of
is in control. However for the other 351,000 £2 billion are reported annually by British insurers
crashes (50%) the driver will not brake and the due to whiplash [2]. Statistics from the Comité
system could therefore help to prevent or mitigate Européen des Assurances [16] show that four
the crash. Previous estimates were made by the countries have a very high rate of claims for
authors in [13]. These were more cautious whiplash injuries, including the United Kingdom
estimates based on only 30% of drivers no applying (76% of bodily injuries), Italy (66%), Norway
braking [14], rather than the 50% [12] used in this (53%), and Germany (47%). Average claims costs
paper to show the greater potential effectiveness. linked to cervical trauma can be very high, for
example Switzerland has the highest average cost
Over 75% of crashes are at speeds under 30 km/h per claim [16] with approximately €35,000 per
[1]. This data suggests that for the 263,250 crashes claim, followed by the Netherlands (€16,500), and
that are under 30 km/h City Safety could help to Norway (€6,050).
prevent the impact from occurring completely, and
for the other 87,750 crashes it could help to The annually UK has 432,000 whiplash injury
mitigate the severity (speed) of the impact. insurance claims [17]. Analysis by Thatcham of
whiplash injury claims cases [18] reveals that 70%
According to crash repair costs analysis [15] the of whiplash claims come from front impacts and
average repair cost per vehicle is €1,868 making a rear impacts, which equates to 303,696 whiplash
total repair cost of €3,736 per crash. So for the injuries.
263,250 crashes under 30km/h that City Safety
could help to prevent this equates to a saving of Until now there have been no technologies to
€983,502,000. For the 87,750 higher speed crashes prevent or mitigate whiplash injuries in frontal
it is assumed that City Safety lowers the crash collisions. City Safety is the first technology that
speed and consequently the repair costs are brought offers any potential to tackle the issue of frontal
down to the average level of €3,736 per crash, whiplash, and can prevent this injury from
equating to a saving of €327,834,000. This gives a occurring at low speeds which is an important
total saving of approximately €1.3 billion, as contribution to reducing the societal burden of
summarised in Table 1. whiplash.

In order to calculate the possible effect on whiplash

frequency the same method of estimates was used
to calculate the efficacy of the City Safety system.
With acceptance criteria of low speed rear end
crashes where the striking car does not brake (50%

Avery 3
of crashes) 151,848 whiplash injuries would be system was activated correctly, driving toward the
saved. traffic cones did not alert the driver in a realistic
manner because it did not resemble a real crash
The average whiplash claim cost is €4,000 [2]. This situation. The realistic size and shape of the
equates to an estimated cost saving of inflatable car aids the drivers understanding of the
€607,392,000 for the 151,848 whiplash injuries that situation, and so gives a more realistic response.
could be saved by City Safety. Combined with the
repair cost savings of €1.3 billion, a City Safety The collision assessment tests were carried out on a
equipped fleet could potentially reduce Insurance test track. The driver was asked to drive normally
Claims by nearly €2 billion annually. toward the stationery inflatable car at the required
speed, but not brake (see Figure 1). The test
Driver adaptation conditions and timings varied, for example some
tests were in the rain with the windscreen wiper
The potential effectiveness of any automatic system in operation, some in normal dry daylight
braking system like the City Safety system depends conditions, and some in partial darkness.
upon whether a driver will adapt and rely on it,
negating any crash reduction potential. There are There were 99 participant drivers. Participants were
progressively more and more automotive primary aged from 20 to 70 years, and all of them were
safety technologies coming onto the market from qualified to drive in the UK. Not all participants
increasing numbers of manufacturers, including were from the UK, with 10% from other countries
technologies offering similar automatic braking internationally. Most drivers were asked to
systems to City Safety. However there is little complete the survey immediately after completion
commonality between the different systems in of the test, and some were given chance to reflect
terms of functionality and system operation. The upon their experience.
introduction of these new systems raises a number
of important questions. Will drivers understand the
meaning of a warning when it is given, what the
warning is referring to, and its criticality? More
importantly will they react appropriately? Will
drivers adapt to these technologies reducing any
safety benefits that may have been available? In a
worse situation, if one vehicle usually indicates a
non-critical occurrence such as low fuel, in another
vehicle a similar warning may indicate an
imminent collision. Such misunderstandings could
be potentially fatal.

Two test types were undertaken on the City Safety

system. The first test involved creating a collision
scenario that is prevented by the system. The
second test type was normal driving on public
roads with the system operational, followed by
questionnaires used to investigate drivers’ reactions
and opinions of the system.

DRIVER COLLISION ASSESSMENT TEST

Method

The participants drove the test car toward an

inflatable target car at 16km/h (10m.p.h.) without Figure 1. Driver Collision Assessment Test.
braking. The City Safety system autonomously
braked the vehicle so preventing the impact. To Results
avoid risk of damage to vehicles or injury to
participants an inflatable target was used. The Only 4% of drivers believed prior to the test that
inflatable target was a life sized representation of a the car would not actually brake. 37% of drivers
car to elicit the appropriate response from the had seen the system operate for another driver so
driver – many people were frightened by its believed that the car would brake automatically.
realistic dimensions. Prior demonstrations of the 59% of drivers believed prior to the test that the car
system using traffic cones revealed that whilst the

Avery 4
would brake without having seen it operate braking to the vehicle that they would not
previously. participate in the collision assessment test. This
also confirms the trend that drivers are unlikely to
67% of drivers felt the urge to apply the brakes as adapt their driving style to rely on the system to
they approached the target balloon car and did not brake for them in normal driving. These 5 drivers’
act upon it. 11% of drivers felt the urge to brake responses are not counted in the analysis of the 99
and actually applied braking by pressing the brake drivers who did participate in the tests.
pedal. Some of these drivers actually had to repeat
the test several times in order to overcome their 2 drivers commented on their perceived increased
instinctive fear of a collision and their consequent risk of a rear-end impact in additional comments on
urge to apply braking. 22% of drivers did not feel the survey. Their concern was that the car behind
any urge to brake as they approached the target. would be more likely to run into the rear of their
car when City Safety braked suddenly. These
drivers were informed that City Safety cannot
Did you feel the urge to brake as you apply more braking force than the driver so cars
approached the target inflatable car?
autonomous braking is merely replacing that of the
driver. If the car does not have City Safety fitted
80% and the driver does not brake there would
Percentage of drivers

70%
60% inevitably be a crash, consequently leaving the
50% person travelling behind little time to respond
40% either since no brake lights would show. The
30%
20% autonomous braking of City Safety activating the
10% brake lights could indeed help to warn any
0%
following drivers earlier, hence adding to the
Felt urge to Felt urge to Did not feel
potential benefit of the system.
brake brake and urge to brake
applied braking
ROAD DRIVING TEST
Figure 2. Drivers urge to brake in response to
collision situation.
Method
Assuming that they could afford it, 93% of drivers
Participants were loaned the test vehicle shown in
said that they would choose to have City Safety
Figure 3 for a period of up to one week to allow
fitted on a car that they were purchasing.
familiarisation with the controls. The test car was
an S80 loaned by Volvo that was retro-fitted with
Drivers were asked if they would rely on the City
the City Safety system for purposes of the research.
Safety system to brake for them during normal
The system is only fitted to new cars, and was
driving conditions i.e. that they would adapt their
launched on the XC60 in November 2008.
driving style to incorporate the functionality of the
system. Only 5% of drivers stated that they would
rely on City Safety during normal driving. 95% of
drivers stated that they would not rely on City
Safety during normal driving, and that it was only
for automatic braking in emergency situations if the
driver was distracted.

Discussion

The collision assessment survey results from

drivers reveal a strong trend indicating that they are LIDAR
unlikely to adapt their driving style to a City Safety sensor
type system, and allow the system to brake for unit
them. 78% of drivers felt the urge to brake when
approaching the target and 95% of drivers stated
that they would not rely on the system during Figure 3. Road driving test vehicle fitted with
normal driving. Driver adaptation to the City City Safety.
Safety system therefore seems highly unlikely.
The drivers used the car for normal road driving
There was also a group of non-participants i.e. within the UK on varying urban and inter-urban
those drivers who refused to participate. They were journeys. Feedback was gathered from 11 drivers
so afraid of relinquishing control of the vehicle who regularly travel high mileages. The mileage

Avery 5
travelled included an equal split between of false interventions in this study is an important
motorways as well as urban and rural roads, all of finding.
which were normal UK roads, for a combined
distance of over 20,000 kilometres. Participants The majority of the participant drivers reported that
were aged between 25 and 55 years old, and all they felt safer, or no different to normal, driving
held full driving licences. when using the system. This indicates that most
drivers were content with City Safety on their car.
Results The 2 road drivers who were aware of the system
during normal driving noticed it because of the
During the road driving trials all the 11 drivers had prototypical nature of the equipment fitted onto the
the City Safety system operational, since it could loan vehicle’s windscreen with visible components
not be de-activated on the test vehicle. For all and wiring. Production vehicles have the system
drivers, no positive interventions of the City Safety sensors cosmetically encased and will consequently
system were reported, and no false interventions be less noticeable.
either.
CONCLUSIONS
50% of drivers reported that they felt safer than
usual knowing that they were driving the car fitted In order to identify an impending low speed impact
with City Safety that had the capability of the City Safety system uses a LIDAR sensor
preventing a low speed collision. 30% felt no mounted in the front windscreen. The car brakes
different driving the test vehicle compared to their are automatically applied when an imminent
usual driving. 10% of drivers felt more confident collision is identified. The automatic braking can
driving the car fitted with City Safety, and the prevent an impact under 15 km/h and can mitigate
remaining 10% felt more nervous. an impact between 15 and 30 km/h. The City
Safety system prevents common low speed crashes
where whiplash typically occurs. It shows potential
How did you feel whilst driving the car fitted
with City Safety?
for reducing the burden on the wider society as
well as insurers. The UK estimates presented
Percentage of drivers

60% indicate the system could affect 351,000 crashes

50% annually by preventing or mitigating the crash. The
40% estimates show that City Safety could also save
30% over 150,000 crashes involving whiplash injuries.
20%
This equates to an estimated cost saving of nearly
10%
€2 billion.
0%
Confident Nervous Safer No
Studies of driver responses in normal road driving
different
showed no interventions of the system, including
Figure 4. How drivers felt whilst driving car no false activations. Collision prevention testing
fitted with City Safety on normal UK roads. involved drivers driving toward an inflatable target
car resulting in automatic application of the brakes
Only 2 of the drivers were aware of the system to prevent an impact. In these collision assessment
whilst driving. These drivers were conscious that tests the majority of drivers felt the instinctive urge
they could see the components of the system or to brake in response to the collision situation that
they were monitoring whether the system was was created. Drivers also stated that they
operating. The majority of drivers (82%) were not understood that the system is designed for
aware of the system during normal driving, so it emergency situations only and they would not rely
was in the background and did not distract them. upon the system in normal driving. This driver
study indicates that driver adaptation to the City
Discussion Safety system seems unlikely.

None of the drivers encountered an emergency The City Safety system appears to offer significant
situation where the system would activate, so the benefits to all drivers in preventing the most
City Safety system did not actually intervene for common sort of impacts. The system is low cost
any drivers during their road trials. None of the and can be readily made available across a new car
drivers encountered a situation where City Safety fleet. Estimates presented in this paper indicate that
was required to prevent a collision. Furthermore significant reductions in injuries and repair costs
there were no false interventions. False are possible. Due to the late activation of the
interventions could annoy drivers and lead them to system in the collision process and the harsh and
mistrust such technologies preventing their unpleasant emergency braking applied, an
widespread adoption in the vehicle fleet, so the lack

Avery 6
activation of City Safety is expected to discourage [12] Langweider, K., German Insurance
drivers from adapting to the technology. Association / Gesamtverband der Deutschen
Versicherungswirtschaft e.V. .
ACKNOWLEDGEMENTS
[13] Avery, M. and Weekes, A.M. (2008). Volvo
The authors would like to acknowledge the drivers City Safety - Collision avoidance technology and
who have tested the City Safety system and its potential to reduce whiplash injuries. Neck Pain
completed questionnaires. The authors would like in Car Crashes, Munich.
to thank Volvo Cars for the loan of the test vehicle.
[14] Breuer, J. (2008). Active Safety: Heritage,
REFERENCES
Current & Future Technology. Stuttgart, Mercedes-
[1] Volvo Car Corporation (2007). Estimates from Benz.
NASS and STO data. Gothenburg, Volvo Car
Corporation. [15] Thatcham; the Motor Insurance Repair
Research Centre (2007). Real World
[2] Association of British Insurers (2008). Motor Damageability: Project Report. Thatcham, UK.
Insurance Claims Data. London, UK, Association August 2007.
of British Insurers. www.abi.org.uk.
[16] Comite Europeen Des Assurances (2004).
[3] Folksam (2005). How safe is your car? Minor Cervical Trauma Claims: Comparitive
Stockholm, Folksam Research 10660 Stockholm Study. Brussels, Comite Europeen des Assurances.
Sweden. www.folksam.se.
[17] Compensation Recovery Unit (2007).
[4] Watanabe, Y., Ichikawa, H., Kayama, O., Ono, Compensation Recovery Unit data.
K., Kaneoka, K. and Inami, S. (2000). "Influence of
seat characteristics on occupant motion in low- [18] Thatcham; the Motor Insurance Repair
velocity rear-end impacts." Accident Analysis & Research Centre and Medico-Legal Consultancy
Prevention 32 (2):243-250. (2008). Real World Injury Claims Database.
Thatcham, UK, Motor Insurance Repair Research
[5] Nygren, A. (1984). "Injuries to car occupants - Centre.
some aspects of interior safety of cars." Acta Oto-
Laryngologica 395:1-164.

[6] Nygren, A., Magnusson, S. and Grant, G.

(2000). Nackskador efter bilolyckor Whiplash
associated disorders. Lund, Sweden,
Studentlitteratur.

[7] Squires, B., Gargan, M.F. and Bannister, G.C.

(1996). "Soft-tissue injuries of the cervical spine:
15 year follow-up." Journal of Bone and Joint
Surgery 78B (6):955-957.

[8] Kullgren, A. (2008). Dose-Response Models

and EDR Data for Assessment of Injury Risk and
Effectiveness of Safety Systems. 2008 International
IRCOBI Conference, Bern.

[9] Department for Transport (2008). Road

Casualties Great Britain: 2007. London,
Department for Transport. 25th September.

[10] Claims and Underwriting Exchange Data

(2007).

[11] Thatcham; the Motor Insurance Repair

Research Centre (2008). Analysis of Motor
Insurance Claims (FNOL) data. Thatcham, UK.

Avery 7