Designing for Vehicle Security: towards a crime free car

Dean Southall Paul Ekblom

CRIME PREVENTION UNIT: PAPER 4 LONDON: HOME OFFICE Editor: John Webster Home Office Crime Prevention Unit 50 Queen Anne’s Gate London SW1H 9AT

© Crown Copyright 1985 First published 1985

Crime Prevention Unit Papers The Home Office Crime Prevention Unit was formed in 1983 with a view to promoting preventive action against crime. Included in its remit is a particular responsibility for the dissemination of information on crime prevention topics. The present series of occasional papers, while based largely upon research material, has been established primarily for practitioners whose work has a direct bearing on the reduction of crime.

ISBN 0 86252 222 6 (ii)

Foreword Although theft of cars and their contents probably causes less anxiety to the public than household burglary, it is a crime which occurs one and a half times more frequently. It is also expensive: in 1983 net losses to owners, even after any insurance claims were met, amounted to some £260 million. This report examines, for a technical readership, the security weaknesses of existing vehicle designs and considers the feasibility of reducing theft of cars and their contents through careful attention to security issues from the earliest stages of car design. It is based on research commissioned by the Home Office from the Institute for Consumer Ergonomics, Loughborough University, and written jointly by the Institute and the Home Office Crime Prevention Unit. The research clearly indicates that car security can be greatly improved with minimal interference to the design of cars, at relatively little cost and in a manner which imposes no inconvenience to the motorist. Building on the clear but limited gains from such measures as the steering column lock, the authors argue that significant and progressive reduction in car crime could be achieved within the next decade. The research naturally has implications for car manufacturers, but increasing sales of accessory products such as alarms suggests a growing interest among consumers for car security, and that opportunities for the responsive manufacturer will arise. It is encouraging that a number of British manufacturers are already beginning to respond by introducing for example high security locks on car doors and steering columns. What is missing is an approach to automotive design which considers the security of the car as a whole. I trust that this report will stimulate efforts by the industries concerned to give car security the priority it deserves and the consumer the security he needs.

GILES SHAW, MP Minister of State Home Office November 1985

(iii)

Acknowledgements

Many individuals and organisations contributed to the research on which this paper is based. We are especially grateful to the Working Group on Vehicle Security of the Home Office Standing Conference on Crime Prevention, chaired by John Curtis; Detective Superintendent Gerry Pope, Metropolitan Police C1(6) Branch; the consultants to the project at Loughborough University — Jim Cooling, Department of Electronic and Electrical Engineering, Gordon Lucas and Eric Cresswell, Department of Transport Technology, and David Parrish, Department of Electrical Engineering; and Margaret Galer of the Institute for Consumer Ergonomics. DEAN SOUTHALL PAUL EKBLOM

(iv)

Contents

Page Foreword

(iii)

Acknowledgements

(iv)

1 Introduction

1

2 Assessing the feasibility of security by design

3

3 Mechanical and bodywork security

6

preventing entry

6

vehicle immobilisation

9

preventing resale

10

meeting design criteria

10

4 Electrical and electronic security

12

detecting interference

12

preventing entry

14

detecting entry

17

vehicle immobilisation

19

preventing resale

20

the cost of electronic security devices: a final note

21

5 Conclusion

22

Bibliography

25

(v)

1 INTRODUCTION Motor vehicles are peculiarly vulnerable to theft — both of themselves and their contents. British Crime Survey estimates suggest that in 1983, for example, nearly 300,000 private vehicles were stolen and that there were some 1.3 million thefts from vehicles. These offences are thus costly in the cumulative distress and inconvenience which they cause. Moreover, the initial police response to theft from and of cars, together with handling recovered vehicles, were estimated at over £10 million in England and Wales in 1984: this is the equivalent of about 800,000 man-hours and represents resources unavailable for tackling more serious crime. The courts, prisons and probation service all add further to the expenditure on autocrime falling to the public purse. One of the several reasons for the attraction of the motor vehicle to the thief lies in its inherent insecurity. Although cars are not totally devoid of protection — they are provided with door and steering column locks, they have ignition keys and locked engine compartments — in many cases these devices are of poor quality, easily overcome by even the most inexperienced thief and often disregarded by the motoring public who forget to lock doors or who leave valuables on view. In attempting to tackle this latter problem successive governments have stressed the importance of locking cars and not leaving valuable items in unattended vehicles. But it is the design of the car itself which, ultimately, must form the cornerstone of our strategies against autocrime. There is little point in urging the public to engage a lock which can be overcome with ease by any passing delinquent. This report makes the case for designed-in car security. It will argue that the technology is available now which could lead to significant reductions in vulnerability, and that the prospects for the longer term, making full use of micro-electronic technology, a r e w o r t h y o f d e v e l o p m e n t b y m o t o r manufacturers. In making this case, the report draws on a study funded by the Home Office into the feasibility of increasing the security of mass-market cars. This work was carried out by the Institute for Consumer Ergonomics at Loughborough University, aided by consultants with motor industry experience from the University’s Departments of Transport Technology and Electronic and Electrical Engineering. The Institute also drew on the experience of a wide range of British, European and Japanese vehicle and component manufacturers, motoring organisations, police, government officials and the Working Group on Vehicle Security of the Home Office Standing Conference on Crime Prevention. The research also involved holding a number of discussion groups which explored motorists’ assumptions about the security of their cars and the extent to which they were prepared to pay for improved security. The results of these discussions are described in the full report (see Bibliography). 1

Following a brief consideration of the criteria against which car security might be judged, and a review of some general issues in relation to security by design, the main section of this report falls into two parts. In the first, mechanical security measures currently available or feasible are considered and assessed against the design criteria. The second part repeats this exercise in relation to electrical and electronic devices, although the latter are assessed against several additional criteria. Finally, in the concluding section, consideration is given to the security features which might be available as standard in five and ten years time if action by manufacturers is set in hand now.

2

2 ASSESSING THE FEASIBILITY OF SECURITY BY DESIGN There is little point in developing a highly secure lock which is too complex or costly to install or awkward for the driver to operate. As well as being effective against criminals, therefore, design solutions to autocrime must be such that they can be readily implemented and, if possible, are more likely to be used by the motorist than less. There are a number of factors to be considered in assessing security devices. Those listed below were developed in consultation with car and vehicle components manufacturers, the police and car users, as criteria against which to judge devices currently available or likely to be developed in the near future; they comprise cost, convenience and reliability, safety, design freedom and cost effectiveness.

Cost All manufacturing costs are strictly controlled to a fraction of a penny per component; and motor manufacturers are correct to assume that car buyers are currently unready to pay much for added security. However, the group discussions held with 46 randomly-selected drivers indicated that if properly informed about the risks of autocrime and the scope for reducing it through design, people will be prepared to pay more, perhaps as much as £50 more, for a car. As car manufacturers come to compete on security the cost constraints may relax, and mass-production will also lower prices. For maximum cost-effect, the level of security to be set for a given model should reflect its value to the owner, and the risk of theft — in particular how popular it is with thieves, and how far it is likely to be stolen for resale by determined and organised offenders rather than merely taken for a drive by casual thieves and abandoned — a far more likely occurrence in most cases.

Convenience and reliability Given drivers’ current attitudes, they are unlikely to accept or use security features that are inconvenient in normal use. In the event of failure, security systems should not prevent the car being used, and should indicate the need for attention: currently locks may wear to the point of insecurity with no sign to the driver.

Safety As a minimum security features must comply with statutory regulations. BSAU 178: 1980 and ECE (United Nations Economic Commission for Europe) regulation 43 set out the standards for road vehicle safety glass and ECE regulation 18 specifies that a device for immobilising the vehicle should, on safety grounds, operate mechanically. 3

Design freedom The detailed specification of security measures to be implemented by all manufacturers across all models would restrict design options for vehicles and dictate other aspects of their bodywork or components to an unacceptable extent. Any security system must be capable of incorporation within a wide range of vehicle designs and must be evaluated, therefore, on this basis. General issues In addition to these specific criteria there area number of general points of which manufacturers should be aware. First, effective security requires design of the security of the car as a whole. Little is gained for example if a highly secure lock is fitted to a car with an easily defeated latch. Second, in a sense, there is a kind of ‘arms race’ between designers and thieves, who seek weak points in every new model, develop methods of attack and disseminate these sometimes to less experienced offenders. The designers can keep further ahead of the thieves by adopting a wide range of alternative approaches to particular security functions, so that no universal method of overcoming them can exist. The more flexible electronic approaches to security offer wider possibilities for keeping ahead. Design freedom thus aids effectiveness. Additionally, the designers can maintain good contact with police officers familiar with the latest methods of attack. Figure 1 shows the scope for the prevention of autocrime by design as a series of obstacles which can be placed before the thief attempting to get into the car, steal from it or drive it away. It indicates the greater scope in the longer run for electrical and electronic approaches. Car security can be improved through a staged approach. For example, on low value cars, mechanical solutions may currently be the most feasible strategy; for models of higher value, electrical/electronic solutions may be feasible. Many mechanical solutions are available now with a minimal requirement for further development whereas the application of electronic security features and their eventual employment in lower value cars depends on further research and development.

4

Figure 1: The scope for prevention of autocrime by design

Note: page numbers are in brackets 5

3 MECHANICAL AND BODYWORK SECURITY This section considers mechanical means of reducing the vulnerability of vehicles under three of the headings illustrated in Figure 1 (page 4); preventing entry, immobilising the vehicle and preventing resale. It goes onto consider the extent to which the security measures proposed meet the design criteria. Preventing entry: perimeter security (i) Door locks The security provided by a lock depends on its design, the manufacturing tolerances and engineering standards to which the lock is made, and the security of the method of lock installation and assembly. Many locks are machined to less than ideal tolerances due, for example, to a poor specification from the customer, or the need to allow for electroplating or lubrication. A lock with freeplay or large machining tolerances is easy to jiggle or pick and more susceptible to operation with a close-cousin key. Lock vulnerability may significantly increase with wear within two years. A mild steel or die-cast zinc keyhole plate facilitates the insertion of a screw thread to extract the barrel. Figure 2: Component parts of a standard door lock

The disc tumbler locks now standard in car doors (Figure 2) provide an example of the vulnerability of the fittings on current vehicles. They may be overcome, for example, by any of the following methods: Key interpassing: a different key operates the lock 6

Jiggling: a key of the correct cross-sectional profile is manipulated within the lock to align the discs and thereby operate the lock Picking: a fine tool aligns the discs and operates the lock Cylinder rotation: inserting a screwdriver into the keyhole and twisting with the aid of a tommy bar or mole wrench to rotate the inner barrel of the lock or the whole lock assembly, and in effect to operate the lock Barrel extraction: a tool such as a slide hammer or even a corkscrew is screwed into the barrel and force applied. The means of overcoming the weaknesses of many current locking systems are already on the market. Examples of the kind of improvements that can be made in design, manufacture and assembly are listed below: Lock design: two major producers of vehicle locks have achieved fundamental redesign. The first of these newly designed locks, in which wear does not reduce security, are being fitted as standard on the latest range of cars produced by a major manufacturer. The other new lock — an ‘anti-pick’ lock — is also in service in steering column assemblies. These locks have been produced with minimal additional on-cost. The standard lock costs the manufacturer approximately 70p; the locks described cost about 6p extra and 2—2½p respectively. Neither incurs direct additional installation costs. Lock manufacture: one major supplier of locks has stated that for an additional penny per lock the cylinder could be machined to a higher tolerance which would give much greater security. Engineering and production constraints may limit the scope for the use of harder materials for discs. However, a hardened steel keyhole plate is already fitted to some locks. Lock installation and assembly: in many modern vehicles the lock is located directly in the door panel and held in place by a U-shaped spring clip, which may offer little resistance to the rotation of the whole lock within the door. A lock may be protected if housed within the handle casting provided this is sufficiently strong, which may exclude some plastic mouldings. However, depending on the vehicle model there may be severe space constraints within the door panel which not only limit the room for alternative designs but also make it hard to find a method of line assembly installation as easy as the current clip method. On assembly itself, locks assembled from the rear are often less prone to extraction of the barrel than those which are front-loaded.

(ii) Door latches and latching mechanisms The latch and associated components within the door are frequently vulnerable and enable the lock to be bypassed. Simple tools including a steel rule, a length of bent wire or a ‘slim-jim’ can be inserted between the door glass and rubber 7

weather strip and activate door lock release mechanisms such as operating rods or exposed latch operating arms. There are several feasible ways to protect latches. Shields: shields may be added to the interior of the door to deflect a tool from the operating mechanism. One major company, for instance, has installed them in one of its more expensive models and a major lock set supplier has designed a moulded plastic one suitable for other vehicles. This would add around £2 per door to the showroom price of a car. Bowden cables: the operating rod which connects the interior door handle or locking button and the latch can be replaced with a Bowden cable (a multi-strand steel cable in an outer sleeve such as is used to operate bicycle brakes). Suitably mounted, this will resist direct manipulation by such tools as the ‘slim-jim’. Body holes: access to latching mechanisms provided by designed body holes, for example, those exposed when exterior trim is removed, can be blocked by their relocation. (iii) The sill-mounted door locking button This can be raised by a plastic strip or a length of wire inserted via the door or window. The button is still fitted to certain current models because, it is argued, it is convenient for the driver both to use and to check that all doors are locked. Some manufacturers however have ensured the button is flush with the sill when locked, or the function is incorporated into the interior door handle out of reach. This may in fact improve convenience and safety as a single control both unlocks and unlatches the door from inside the car. (iv) Bonnet security A secure passenger compartment protects the internal bonnet release handle. The engine compartment can be further protected by installing and shielding the bonnet catch release mechanism to prevent opening by a screwdriver, for example, via the radiator grille. This may protect engine components against theft but will not always prevent ‘hot-wiring’ to start the motor, as this may be possible via the ignition circuitry within the intrument panel assembly. (v) Boot security Security of boot locks and latches raises similar issues to door security. For extra carrying space, some vehicles now allow for part or all of the rear seat to tip forward, and if this can be unlatched from within the passenger compartment the boot no longer possesses independent security. One manufacturer deliberately placed the latch release in the boot itself in a recent model, but has received complaints of inconvenience from customers unaware of the purpose 8

behind the move. A remote boot release lever, a convenience feature usually mounted near the driver’s seat, should incorporate a lock. Estate cars pose harder problems, although a lockable floor compartment may be possible; but hatchbacks could be provided with a rear parcel shelf so that at least items left in the car may be concealed from view.

(vi) Side window glass: composition and installation The toughened glass in vehicle side windows is easily broken allowing theft of contents and access to the interior door locks. An increase in the penetration resistance of side window glass is an obvious solution. Laminated glass has a much greater penetration resistance, but weight/strength considerations may necessitate redesign of the window winding assembly and perhaps the installation of the glass within a frame, which would reduce visibility, The most serious objection is the cost — laminated glass alone is 30-40% more expensive than toughened glass, and the entire assembly will double or triple the cost of side and rear windows. Possible alternatives are plastic, and glass/plastic composites but it is too early to say if they will be adopted widely. They are currently either unable to meet abrasion requirements or so costly as to preclude widespread implementation. They must also await changes in glass standards regulations before they may be installed in front or front-side windows. On glass installation, the recent move from gaskets towards glueing or bonding may give extra protection, depending on the quality of the process. Although quarter lights usually do not open in modern cars, the rubber seal and adhesive method of fixing enables them to be prised out, sometimes with little damage. Quarter light installation should be modified to prevent this.

Vehicle Immobilisation (i) Steering column locks Steering column locks are subject to the same vulnerabilities as door locks — picking, jiggling, and barrel extraction. Rear-loaded lock barrels are more secure than front-loaded ones inserted from the key end, and in general it will be more effective to depend on the inherent strength of the installation method rather than on locating the lock within a confined space to reduce the criminal’s scope for inserting tools. The space required around the lock to allow for everyday use, maintenance and repair is inevitably sufficient for slide hammers and other forms of attack. High security locks are in production and should be fitted as standard. 9

(ii) Transmission locks Transmission locks operate on the gearbox rather than the steering column and may, for example, lock the car in reverse gear. Where this method is used it is very difficult to overcome, although this may reflect the high manufacturing standards of the locking mechanism as well as the availability of a more secure mounting.

Preventing resale — vehicle identification (i) Vehicle identification numbers Some manufacturers currently rivet on a plate containing vehicle identification details, which may easily be removed and replaced without trace of tampering. Stamping the number directly into the metal of the component is obviously less vulnerable; alternatives have been proposed however, including attaching a thin plastic label which will be clearly damaged by attempted removal. Such labels may also be damaged during legitimate use of the vehicle, and while this is no proof of changed identity, it should raise the suspicion of a prospective purchaser or the police. Organised criminals may find other means of obtaining new vehicle identities, so the production and supply of new labels should have a high degree of security.

(ii) Window etching with registration number This is currently growing in popularity as an after-market (accessory) service and some manufacturers are contemplating offering it as standard through their dealership net work. Window etching enables an easy check against a vehicle’s registration plate, acting both as a visible deterrent and to reduce the resale value of the stolen car. At present it offers advantages to the individual motorist as the thief moves on to look for an un-etched car, but it is questionable whether etching would be useful to the community as a whole in reducing overall theft levels. If it became standard on all vehicles, a criminal re-selling stolen cars may simply adjust to the cost and inconvenience of having to replace their windows. To replace those in a medium priced small family car with a secondhand value of over £3,000 would currently cost around £150, and according to one expert requires no specialist tools and only a little practice.

Mechanical and body security: meeting design criteria Cost Many of the proposed solutions incur minimal on-costs. An improved door/ boot lock at an extra 2½p each will cost the manufacturer 7½p per car, which may rise to some 25p on the showroom price. Shields to protect against the 10

‘slim-jim’ would add less than £10 to showroom price. The most expensive item considered is laminated glass for side windows. Safety The proposed solutions will not reduce the current level of safety; some confer additional safety to offset against their cost. The Institute for Consumer Ergonomics Accident Research Unit suggests Bowden cables around the perimeter of the door as an alternative to plain rods on latch release mechanisms. They are less likely to cause the door to open through distortion during a side impact collision, possibly ejecting the occupant, who may then sustain more serious injury than otherwise. Claims have been made for the safety advantages of laminated side window glass, but these have yet to be verified in real accidents. Design freedom The proposed solutions are not design restrictive. They are, in the main, minor details which if considered at the appropriate point in product development can be readily implemented within the various design and styling criteria in use at the time. User convenience The solutions proposed do not require extra actions by the driver in using the vehicle. Inconvenience could still arise should the driver lock the keys in the car or lose them while the car is locked. The former can be avoided using a technique found on some current cars which requires the driver’s door to be locked with the key rather than the sill button. The problem of lost keys is harder to solve. Some manufacturers provide a partial solution by supplying a spare flat key to be kept in a wallet or purse. Alternatively, it is possible in principle to design the vehicle such that entry (and driving) can be done without the proper keys, but using skill, special tools and time, requiring a specialist call-out service. One security company currently provides a 24 hour service for owners of a particular make of car. Reliability No particular problems of reliability should arise with any of the mechanical security options described. Many involve only simple changes in passive components such as removal or relocation of body-holes or the addition of a fixed shield; others, such as improved locks, require construction with improved materials and machining to tighter tolerances, so wear is less likely to affect performance and reliability may even be increased. 11

4 ELECTRICAL AND ELECTRONIC SECURITY Electrical and electronic security systems already available as car accessories are bewildering in their range (see full report, Appendix 2). Figure 3 summarises the scope of security features which can be designed into cars. Electrical and electronic approaches require additional design criteria to those listed on page 5. These are shown in Table 1. Table 1: Design criteria for electrical and electronic security systems. User acceptance — cost and convenience Reliability Maintenance requirements Repairability Safety Failure modes (fail to safe requirements) Power supply requirements Manufacturing effort and integration into other vehicle systems Electromagnetic compatibility (EMC) requirements The electric and electronic security systems now available are designed with various aspects of security in mind: detecting interference, preventing entry, detecting entry, vehicle immobilisation and preventing resale. Under these headings, some of the more important security systems now available are discussed below, together with an assessment against the appropriate design criteria. Detecting interference The most likely candidate for this task is currently a motion detection system, typically using trembler or tilt sensors to set off an audible alarm. In practice the tilt sensor is preferable; when combined with appropriate electronics it should be much more flexible. Tremblers and other alarms that work on vibration have the disadvantage that they are easily set off maliciously or accidentally. In all cases the alarm can be sounded by the car horn if this is protected and the alarm mode is clearly distinguishable from the ordinary noise of a horn, for example by oscillation.

Design assessment: detecting interference There should be no real problems with user acceptance, especially as this device is already widely used in security accessories. The need to provide a sensor does introduce a real cost increase; however the total increase would be minimised when the electronics are an integral part of the vehicle. Operation is simple, 12

13

whilst repair and maintenance are likely to involve replacement of defective units. The device should be reliable in service given proper design (and appropriate tuning of the sensors should prevent false alarms), and it should be relatively easy to integrate into existing and future vehicle design. There are no safety implications and fail to safe is not required; nor are there any E M C (electromagnetic compatibility) requirements above the general standard for vehicles. Power supply must be maintained when the device is armed.

Preventing entry Electrically operated locks promise a significant improvement in vehicle security. They enable the separation of the lock’s two functions — ‘coded’ control of access to the lock actuating mechanism (i.e. responding only to the valid key or its equivalent), and actuation itself (i.e. allowing the latch to be released when the door handle is operated) — with the result that many existing lock-picking techniques will not work.

(i) Lock access The three most likely candidates for access methods are keypads, hand-held remote controllers and keys. Keypads: these are often suggested; in fact one has already been fitted to an advanced experimental car and they are available on some models in the USA. This solution is feasible today and has the advantages (potentially attractive to car rental companies) that there are no keys to be lost and that the code can be reprogrammable, providing a ready means, in effect, of changing the locks. Hand-held controllers: this method is currently in successful use with accessory alarm systems, employing ultrasonic, infra-red or radio signalling. Typically the off-vehicle control unit is battery-powered and compact enough to fit on a key-ring. It eliminates the need to remember codes. Moreover, environmental aspects are not relevant. It is easy to use and making an unauthorised copy is difficult, but against this are the consequences of: —failure of transmitter or transmitter power —receiver/control unit failure —loss of vehicle power —spurious activation (likely in radio systems only); and —loss of controller. 14

The remote controller approach has too many weaknesses at present for use as the sole locking device; those systems currently available supplement conventional locks. Keys: this term is used in its widest sense to include plastic cards perhaps with bar coding. From a practical perspective the conventional type of key is most likely to satisfy the needs of the car system; an assessment of the qualities of the key approach is given later. (ii) Lock actuation Two devices well-suited for electrically-operated locks are solenoids and motors. The choice between them is not clear-cut. Whichever is adopted, the new locks must be relatively simple and self-contained replacements for existing ones. It is also essential that single-point failures should not disable all locks simultaneously. Trade-offs must be made between mounting, mechanical and electrical considerations: Mounting: the lock and keyhole have traditionally been located in the door and mechanically coupled to the release mechanism. With electricallyoperated locks this need no longer be the case, giving the vehicle designer greater freedom to locate components. There appears to be no significant size problem with either solenoids or motors. Mechanical considerations: by decoupling the lock from the latch, simpler mechanical systems can be used for the interior release mechanism. By using body mounted locks, only the latch is fitted to the door; the resulting mechanism within the door (where space is at a premium) is relatively simple. The complete locking/latching mechanism will be simpler than present ones, minimising the effects of corrosion and jamming. Electrical aspects: solenoids operate relatively quickly, taking a short burst of current. Motors can have a longer run-out time to release the latch, significantly reducing the current needed. Solenoids have been in service in security-vehicle door locks for some years and according to one company have a good reliability record. Either can be readily integrated into a central locking system. (iii) Overcoming the driver’s failure to lock-up About one in five cars on the street are left insecure, and even intensive publicity aimed at persuading drivers to lock up can fail to make an impact. There are three design approaches to this problem: automatic locking, increasing the convenience of locking and locking reminders. Automatic door locking: the electrical locking system described below can be extended to lock all doors automatically once the car is vacated; but although technically feasible, it is questionable whether any such system can take account of all the ways in which the vehicle is used. When frequent access is required, as in loading, repeatedly having to unlock the door may 15

greatly inconvenience the user. An override could be provided for such circumstances, but might be misused. Increasing convenience: central door locking: central locking has been in service in more expensive cars for some years and conversion kits are available for the DIY motorist. They are now beginning to prove their reliability and are viewed very favourably by drivers. Currently central locking is a relatively expensive optional extra: whereas a mechanical latching system may cost the manufacturer £2 per door, solenoid operated central locking costs around £5 per door and the difference in showroom prices will be even greater. Component manufacturers are now developing electrical latching and locking mechanisms providing central locking much more cheaply than before. One prototype arrangement (described in Appendix 4 of the full report), while not yet meeting all the requirements specified in the previous section, shows the potential of this approach. Its security advantages include the impossibility of picking or jiggling the lock and the absence of mechanical linkages to be manipulated, which also saves space within the door and simplifies installation. The latching mechanism with its associated motors is no larger than a conventional latch and occupies much less space than current central locking systems. As it is the objective of the component manufacturer to supply these devices to manufacturers at only 20% above the price of mechanical systems then given the savings on installation the on-cost of this feature is minimised. Advanced forms of central locking are, therefore, a highly marketable convenience feature which has realistic potential for application to less expensive models. Reminding the driver to secure the vehicle: a reminder device has for some time been a standard feature on vehicles sold in the USA. A buzzer sounds when the ignition is turned off until the driver removes the key. As well as enhancing security, the buzzer would reduce the likelihood of a driver locking the keys in the car. British manufacturers have fitted these devices on exported cars. The switch is mounted in a special casting in the steering column lock assembly, and a major supplier of these fits the castings as standard whether or not the relevant car is for export. To provide the warning facility would only require the addition of the switch and buzzer which the component supplier estimated would cost the car manufacturer about 80p. If in future models door locks incorporate a switch, an audible tone could sound when all doors are closed (but not locked) and the key removed from the ignition. The tone would enhance the apparent vulnerability of the car and encourage as well as remind the driver to lock up. Design assessment: preventing entry There should be no problems of user acceptance as from the user’s perspective, access to electrical locks is virtually the same as to conventional locks. 16

Component manufacturers suggest that electric locks should cost little if any more than existing mechanical ones. Moreover, by having a totally integrated unit, no extra manufacturing effort is needed. Electrical locks could be integrated within existing vehicles but are likely to be used for new designs only. Reliability should be high, and highest where access is mainly mechanical and the actuating device a solenoid. In this case, use of the correct key would merely operate a ‘contact’ to energise the solenoid. Lowest reliability (relatively speaking only) will probably be found in motorised locks which use electronic/ optical/magnetic means of decoding. Maintenance should be minimal, possibly less than for conventional locks; repair will probably be by replacement, in line with garage servicing methods. The design features recommended eliminate the single point failure problem; a simple hand-held battery powered plug-in unit for use on individual locks will take care of power failure (and the adoption of a universal system would be convenient to motorists). As a ‘flat battery’ will still retain sufficient energy to actuate the lock, the plug-in unit would rarely be used. EMC requirements must be respected in the design of the electronics, as with any other electronic system. Extra precautions are needed to ensure the lock cannot be damaged or released by injecting high voltages into the power supply.

Detecting entry The car can be entered either directly by the door or by first breaking the glass. These routes into the car must be protected in different ways. (i) Breaking and entering Disturbing the vehicle can be detected by a tilt sensor as described earlier. Alternatively, breaking and entering will modify the interior space of the car, which can be detected by ultrasonic and passive infra-red devices (both assessed in detail in the full report, Appendix 3). An alarm can be sounded and the engine perhaps immobilised. Several modes of operation are available. For example, a high security door locking system may serve to arm and disarm the alarm. For a second best solution, entry by normal means may deactivate the system for a limited period. Unless a further correct set of actions are carried out, such as insertion of the key in the ignition, it would then re-arm itself, and the alarm sound. (ii) Detection of entry via the door Assuming any tamper alarm has been deactivated and entry is made through the door, a simple switch operated by the door or door latch could activate the intruder alarm. Arming might be delayed to allow a legitimate user to deactivate the system using, say, the steering lock unit as a switch. Several variations have been proposed, or used in accessory alarms. Mostly they have used the standard courtesy light switches on the door as the primary sensor, responding for 17

example to change in battery voltage or current. Present-day door switches may be insufficiently reliable for this task. More reliable switches are now being fitted to certain models as sensors for vehicle condition monitoring (in this case indicating ‘door open’ on the instrument panel), and these could be used for the alarm systems. With electrically operated locks, the switch could be built into the lock mechanism. Design assessment: detecting entry (i) Breaking and entering User acceptance depends critically on avoiding inconvenience and false alarms. Current accessory intruder alarm systems are prone to both problems as they require some additional action by the driver. (Nevertheless, one luxury car manufacturer suggests that ultrasonic accessory systems are easily accepted by the customer and have a low false alarm rate; there is little information on these aspects of passive infra-red.) Linking the device with a high security door lock avoids these problems by incorporating the arming and disarming function within the normal action taken on entering and leaving the car. Costs will be increased as the sensors are extra items, and suitable specialised control circuitry must be provided. The effect on manufacturing effort to install either type of system is minimal and integration should present few problems. Performance should be better if the sensors are designed into the vehicle as they can be optimally located and mounted for sensitivity and signal-to-noise ratios. Reliability should be high as solid-state electronic units are used; piezo-ceramic transducers are well-proven in service. Maintenance should be simple; repair would be by replacement. There are no special failure or safety requirements and EMC requirements will be no greater than for other vehicle electronic systems. Loss of power will deactivate the intruder alarm, and accordingly the supply must be protected. (ii) Entry via the door On user acceptance, convenience is high as the door sensors could be armed and disarmed through the normal sequence of entry into the car. Cost and manufacturing effort are minimised by using door sensors and the horn which are already fitted; the alarm electronics are extra. A voltage drop door sensor system could cost around £25 in the showroom; since it is very simple it should be highly reliable in service. There are no special safety implications and failure is not critical. EMC requirements are as for other car electronics. The system will be affected by power failure in the same way as other alarms; it can, though, be wired so loss of switch power alone sets off the alarm. It is more suitable for integration into new designs of car. That it is now offered as an optional extra on some expensive models demonstrates its compatibility with other vehicle systems. 18

Vehicle immobilisation This is generally easier to accomplish than preventing entry into the vehicle. Solutions fall into two categories determined by the complexity of the vehicle’s electronic systems. The majority of current vehicles can be categorised as ‘simple’, with minimal electronics and no integration of systems. For the simple vehicle, electrical and electronic immobilisation methods that cut out the starter, ignition or fuel pump are never likely to be really satisfactory. The first two could readily be bypassed, and all three would require the addition of a high security lock and perhaps an anti-tamper alarm, increasing cost and complexity. Mechanical methods of immobilisation are thus more appropriate. Only today’s most expensive cars can be called ‘advanced’: these will have at least one major sophisticated electronic system fitted, most likely associated with ignition control or engine management (the electronic control of such functions as carburation and ignition timing), and based on microprocessors. For advanced designs, high security can be achieved at little cost. As time goes by progressively more vehicles will fall into the advanced group and within 10 years this will probably become the norm for the middle-market car (see Appendix 5 of the full report). Extra electronic systems will be successively added to advanced vehicles until a step increase in sophistication occurs: the various systems are integrated, and the multi-wire loom is supplanted by multiplexing techniques using single wires under microprocessor control carrying coded information as well as power. In the ‘highly integrated’ advanced vehicle the security concepts would be the same as for the merely advanced, but the implementation would probably be different. With the exception of the sensors, if considered at the design stage the security functions could ‘piggy-back’ on the standard on-board electronics, saving cost and complexity and opening up a wide range of options.

Immobilisation in advanced vehicles With a car incorporating an engine management system the ignition control is extremely hard to bypass. When the engine management controller is armed and refuses to respond to illegal commands, a complete substitute controller would have to be fitted. For arming/disarming the controller, the solutions are similar to those for access to door locks. As current regulations require a mechanical immobilisation device they would probably only supplement the steering column lock for some time, with the advantage of preventing a car being towed away without a special truck. Keypads: the keypad is in a milder environment inside the car than when it serves as the means of access to the door lock: it also eliminates the problem of drivers leaving the car with the keys in the ignition. For cost saving, the keypad fitted to a trip computer could be used. 19

Hand-held controllers: these could be serious contenders although their performance has not been assessed in this application. Keys: by integrating mechanical and electrical systems a very secure ignition lock with a mechanical key can be produced and perhaps incorporated within a steering column lock assembly. This circuitry, when activated by the key, would set the engine management system into operation by transmitting a trigger signal; once functioning it would carry on working. When the ‘ignition’ is switched off and/or the key removed, the engine management controller would reset; to activate it again another trigger signal is required. Integration of mechanics and electronics is vital to avoid simple bridging of the ‘switch’ to fool the transmitter into sending a trigger signal. The key unit would communicate with the engine management controller by a serial digital link, for which fibre optics could provide ultra-high security. The overall security level achieved would ultimately depend on the security of the key itself. The controller could be re-programmable to allow resetting of codes, to enable ‘changing’ of the lock should the key fall into the wrong hands and to allow replacement of the lock or the controller should either fail. With this method there is no need to retain the steering lock.

Design assessment: vehicle immobilisation From the driver’s perspective, operation is the same as with the present-day ignition key. There should be no problem, therefore, with user acceptance. The cost of the key electronics would be insignificant when compared with that of the main engine management system. No special manufacturing methods are needed and by definition the security function would be highly integrated with the rest. Reliability would be dominated by that of the main engine management system; m a i n t e n a n c e would be zero and repair b y r e p l a c e m e n t . O n s a f e t y , immobilisation prior to driving off is better than a delayed engine stall in traffic. Failure of the key unit is fail to safe. If the engine management system is already activated it will continue to operate normally until next switched off by the driver. EMC must be high, at a level equal to that of the engine management system. Should the power supply to the key unit fail, a bypass could be incorporated within the engine management system, for example by entering a code via a keypad, which would be acceptable in this fall-back mode, or a diagnostic socket. Preventing resale Vehicle identification Two possible approaches are outlined below. The first would only be suitable for the advanced models. 20

Information on the identity of the vehicle and its components is embedded within the on-board systems such as engine management and can be accessed through a diagnostic communications link fitted principally for use in servicing. While it provides a higher level of security than its mechanical equivalent — because the Vehicle Identification Number is simply stamped on — it too must be supported by administration of the numbering system and control of the replacement of defective units. The second approach, while offering rather less security, is simple and much cheaper. When a car is produced it can be allocated a code (unrelated to its registration number), entered on a strip securely welded on the body and perhaps on components. With bar coding and magnetic reading techniques (rather than optical ones), this method could be very cost-effective and the strip difficult for criminals to forge. The cost of electronic security devices: a final note Accurate estimates of the cost to the manufacturer of fitting electronic security functions are difficult to give. The cost will depend, for example, on the specific device fitted, the components employed and the extent to which existing parts of the car can be used within the security system. The security system can confer additional benefits such as increased convenience. With the help of a component supplier it has been calculated that a voltage drop alarm, factory fitted as standard to all vehicles, would add less than £50 to the showroom price of the car, taking account of component and installation costs, taxes and mark-up. The ‘optional-extra’ alarms of this type available now on a factory-fit basis are more expensive to produce and cost the motorist almost twice as much.

21

5 CONCLUSION The research suggests that to produce a car protected against the bulk of opportunist theft requires no major redesign of the vehicle or the addition of any expensive devices. There need be no inconvenience to users, no especial problems with reliability, and no compromise on safety. On low value cars, simple mechanical solutions may currently be the best approach. For models in the near future or of higher value, it should be possible to develop more sophisticated electrical and electronic solutions. While many mechanical solutions are available now with a minimal requirement for further work, the application of electronic security features largely depends on more research and development, although many prototype devices already exist. With design effort starting now, in five years’ time even the least expensive base models can be made far less vulnerable to theft of the car or its contents. A typical car could have the following standard, factory-fitted security package:

22

* * * * * * * *

Securely installed high security locks on all doors and boot No protruding sill buttons A high security, well installed steering column lock Laminated (or equivalent) side and rear window glass Protected bonnet release catches Shielded internal door latching components An audible reminder to remove the keys from the ignition The catches of tipping rear seats housed in the boot and not in the passenger compartment * The boot well of hatchbacks protected by a steel cover incorporating a lock, providing a secure place to leave valuables

More expensive models may also include: * Alarms * Central locking * Immobilisation of the engine through its electronic management system After a further five years, half the cars on the road could possess these features. In ten years’ time, mechanical features may still be the first line of defence, but electrical and electronic features will be more widely available. The reliability of electrical locking systems should be proven as central locking will have been in service for some years. They may therefore replace mechanical locks on base models, which may also be fitted with alarms and electronic engine immobilisation. New methods of gaining access to the car may be implemented. Multiplexing will be more widespread; this microprocessor-based system will provide significant opportunities to increase the sophistication of electronic engine immobilisation, alarms and door locking devices. Electronic systems providing for immobilisation, storage of vehicle identification information, and access to the car will only respond to electronic codes, and tampering will require a high degree of skill — cutting wires will achieve nothing.

23

Bibliography Southall, D. and Galer, M. (1985). The Crime-Free Car: A Feasibility Study. Loughborough: Institute for Consumer Ergonomics. (Available from British Lending Library). Altshuler, A., Anderson, M., Jones, D. and Womack, J. The Future of the Automobile. London: George Allen and Unwin. Burrows, J. and Heal, K. (1979). ‘Police car security campaigns’. In Burrows, J., Ekblom, P. and Heal, K. Crime Prevention and the Police. H o m e Office Research Study No. 55. London: HMSO. Henderson, J. R. (1985). Autocrime: Some Aspects of its Cost. N e w c a s t l e : Northumbria Police. Highway Loss Data Institute (1983). A comparison of the Theft Loss Experience of General Motors Passenger Cars with and without Factory-Installed Theft Deterrent Systems. Research Report HLDI A-19. Washington, D.C.: HLDI. Hough, J. M. and Mayhew, P. (1985). Taking Account of Crime: Key Findings from the Second British Crime Survey. Home Office Research Study No. 85. London: HMSO. Lee, B. and Rikoski, M. (1984). Vehicle Theft Prevention Strategies. Washington, D.C.: US Department of Justice. Mayhew, P., Clarke, R. V. G., Sturman, A. and Hough, J. M. (1 976). Crime as Opportunity. Home Office Research Study No. 34. London: HMSO. W h i c h ? (1984). ‘Autocrime’ W h i c h ? , August, pp. 342-345. London: Consumers’ Association.

25