You are:

+ Share | Print page

Chapter 16. Safety Research at the Transport Research Laboratory

Volume 1 Part 6
Brief résumé of Chapter 16:
Safety Research at the Transport Research Laboratory

D. M. Colwill, OBE, BSc, FIHT, MIAT, FInst.Pet
first recipient of the Paviours' Medal 2000, Worshipful Company of Paviours
formerly Senior Research Fellow, Transport and Road Research Laboratory
Chairman of the Technical Committee on Flexible Roads

Preface

Life and limb were far from safe from accident in the environment of the roads even when the horse provided most of transport's motive power. As the twentieth century came to depend for transport on the internal combustion engine, and then to exploit it more and more in a physical environment little changed except for the addition of hard top, the rate at which accidents occurred and their severity increased hugely. The essence of the problem was to safely accommodate the instantaneous choices of millions in charge for their own purposes of vehicles the momentum of which could far exceed that required for lethality. The problems mounted as vehicles went faster.

But, when the prospect could be descried of superimposing on the road environment a system of motorways, it was seen from the outset that at least where they would run that there was an opportunity to have traffic flowing freely and rapidly with much of the risk of accidents eliminated. So with huge relative benefit in human, social and economic terms, that safety benefit constituted the heart of the case for the investment. In the same spirit John Cox provided lighting for the major motorway intersections at the ends of Preston’s By-Pass in 1958 and at Walton and Samlesbury but the Ministry sadly refused to pay for that lighting. Years later the RRL’s scientific findings and Bird’s economic analysis led the Ministry to reverse their odd view of motorway lighting

This chapter shows that after 1958 many more opportunities to reduce the risk of accidents were identified and confirmed by research at the Transport and Road Research Laboratory - work which won it a premier place among road research institutions internationally. How its findings entered into the detail of motorway design and use, is the subject of this chapter. Its author is a leading exponent of the research process who has also been actively engaged in consultancy, for example for the Automobile Association’s Road Safety Research Foundation.

Doug Colwill writes of the origin of this work:

Introduction

"The Roads Research Laboratory (RRL) was created in 1933 as an institution within the National Physical Laboratory Between 1933 and 1939 the Laboratory produced 230 reports covering materials used in road building, including a joint study with the Forest Products Research Laboratory, the Ministry of Transport and the Timber Development Association into the use of wood blocks and their effect in reducing risk of skidding. The connection with safety was a prime consideration. The work formed the basis of evidence to the Alness Committee in 1938 on prevention of accidents. Dr E. V. Appleton, Secretary of the Department of Scientific and Industrial Research (DSIR), recommended to the Alness Committee that the Laboratory should be equipped with a division devoted to matters of safety on the roads. In 1943 Sir William Glanville, who had held office as Director of the RRL since 1939, asked Dr Appleton whether a Department other than the DSIR could "initiate inquiries into and to judge the necessity to undertake research of national problems, the administrative responsibilities for which rest with another Department". This very opportunity would come about in 1946.

Early international research engagements

Chapter 1 of this Volume has outlined the contacts made by various British institutions, but hardly at all by British governments, with other countries which had been investing in motorways before World War II, and in the USA even before World War I. The developments in Germany in the 1930s had the most important influence on ideas for motorways in the UK. These roads, of which 1,115 miles had been built between 1935 and 1940, were built with dual carriageways, each 7.5m wide, with a central reserve 3.5 to 5m wide planted with shrubs to reduce dazzle. Intersections with other roads were limited and traffic speeds in excess of 100 km/h were envisaged.

Formal British government involvement came with its financing the Laboratory's study of the autobhanen in 1946 under a Co-operative Committee composed of two nominees of the Ministry of Transport, two nominees of the Cement and Concrete Association, three County Surveyors and two members of the staff of the Laboratory. The study was made under the auspices of the British Intelligence Objectives Sub-Committee of the British Control Commission and the DSIR in co-operation with the Ministry of Transport. The report of its findings was published under the title German Motor Roads 1946 as Road Research Technical Paper No 8 and British Intelligence Objectives Sub-Committee Overall Report No 5.

The report included an explanation of the role of the Roads Research Laboratory in the following terms:

"THE ROAD RESEARCH LABORATORY of the Department of Scientific and Industrial Research is a government organisation for studying problems that arise in designing, building and maintaining public highways. The primary objectives of the work, which is undertaken in close co-operation with the Roads Department of the Ministry of Transport, are to improve the road as a channel for traffic, to reduce the overall costs of construction and maintenance and to promote safety and comfort in travel. The work is organised in two divisions: Materials and Methods of Construction are (a) soils and stones, (b) concrete roads, (c) bituminous surfacings, (d) road plant and machinery. Road safety services include (a) road lay-out, (b) surface characteristics, (c) traffic flow and movement control, (d) vehicle characteristics, (e) road user characteristics, (f) accident statistics. The Medical Research Council is collaborating in the work on road user characteristics."

The conclusions of the report were as follows:

"(1) The layout and alignment of the German motor roads are of a sufficiently high standard to suggest a degree of safety and sense of security above that obtained on ordinary main roads in Britain.

(2) Factors thought to contribute to this sense of security are : -

(a) Absence of cyclists, horse-drawn vehicles and pedestrians.

(b) Prohibition of access except at specified and comparatively infrequent junctions.

(c) Reduced fatigue in driving resulting from the absence of hazards.

(d) The uniformity of design of the roads and especially of the width of carriageway.

(3) Except on some long straight sections of the earliest roads built, the view changes at sufficiently frequent intervals to prevent excessive boredom.

(4) Although the standards of lay-out and alignment adopted for the roads appear to be adequate from the point of view of safety, the rigid adherence thereto does not necessarily result in a road of good appearance, especially where horizontal and vertical curves appear together.

(5) In order to avoid faults of appearance in roads of which long lengths are visible at one time, it appears desirable to make models or perspective drawings before fixing final lines and levels.

(6) The designs of the junctions and access points on the motor roads are satisfactory and do not appear complicated from the driver's point of view.

(7) The use of a wide bankette or edge strip to mark the side of the carriageway has the advantage that it provides a space on which vehicles can stop in an emergency without obstructing the carriageway, but, with the surface used in the German roads, it may be mistaken for a part of the carriageway and vehicles may travel on it by accident and so collide with stationary vehicles or with other obstructions. The degree of visual contrast provided by differences in colour alone, such as those used on the German motor roads, is not sufficient. A difference in texture is also required.

(8) If a bankette is used it should be wide enough to allow vehicles on it to be quite clear of the carriageway (a minimum width of 7ft 6 in is suggested) and its surface should be finished in such a way that the bankette cannot be mistaken for the carriageway.

(9) The dark-coloured carriageways, made with concrete containing a black pigment or with asphalt surfacings, with bankettes of light colour were thought to have a better appearance than those with the reverse arrangement of colours, but it should be mentioned that the roads were seen only in daylight.

(10) The normal riding surface of the concrete motor roads in Germany is as good as, but not better than, the best examples of British roads.

(11) A still higher standard of riding quality is desirable for really high-speed traffic. The most likely way of achieving this would be by improvements in jointing methods and materials.

(12) The design of the roads and the treatment of the foundations has generally eliminated cracking due to subgrade failure, but in this connection it must be mentioned that nearly all the roads seen were on sand or gravel soils, or were built on a layer of imported granular material, a fact which may account for the small amount of cracking.

(13) The frequency of cracking in cuttings suggests that better performance would have been obtained if the practice of laying the road on imported granular material had been used in cuttings as well as on embankments, or perhaps if as much attention had been given to the compaction of the subgrade in cuttings as in embankments."

From this first involvement specifically in motorway design the Laboratory went on in the next decades to publish the results of extensive further research projects bearing on safety. For example, work concerned with skid resistance and surface drainage is collected in Concrete Roads: Design and Construction published in 1955. More concise comparative data arising from safety appraisals of British and American motorways appears in the study which the RRL published jointly with the University of Birmingham in 1960 entitled The London and Birmingham Motorway, Traffic and Economics, RRL paper No 46, pp 36-42. Much more detail on safety matters was drawn together in two volumes - Research on Road Safety and Research on Road Traffic, first published by HMSO in 1963 and 1965, respectively. The policy implications of much of the research are recorded in George Charlesworth’s History of British Motorways (Thomas Telford, London,1984).

Statistics of relative safety on classes of road

From the outset, Britain's motorways were conceived as a fast, efficient and safe means of travel. Research on road safety provided basic information on the relative importance of the many factors which contribute to accidents and, hence, a useful guide to methods of preventing accidents. The annual vehicle-mileage on all roads in Great Britain during 1957 was 51,000 million and there were 21,800 injury accidents, an accident rate of 4.3 per million vehicle-miles. The range varied from about 13 in urban centres to about 1.5 on lightly trafficked roads. With 5,500 fatalities on the network, the fatality rate was about 10.9 per 100 million vehicle-miles. Many of these accidents occurred at junctions and on lengths with restricted visibility. Therefore it was expected that motorway with controlled access, shoulders and flyover junctions, largely eliminating the hazards associated with right turns, pedestrians and parked vehicles, would be significantly safer. This was confirmed by information from the United States where fatality rates of about 3 per 100 million-vehicles were occurring on their motorways.

This trend was confirmed in the first year of operation of the London to Birmingham motorway. In Table 1 below it is compared with experience on the A5, which ran nearly parallel to the M1:

Accidents in the 1950s on motorways and other trunk roads between London and Birmingham.

 

M1/M10/A45
Nov 1958 to Oct 1960

A5
1956

Number per 100 million vehicle-miles

Casualties :    Fatal

7

15

Serious  

43

120

Slight  

69

210

Injury accidents

70

220

 

The injury accident rate on the motorway was lower than that recorded on any all-purpose road in Great Britain. On motorways, accidents due to stationary vehicles are almost entirely eliminated due to the provision of hard shoulders.

A similar review was conducted by Lancashire County Council, who carried out extensive studies of traffic and accidents on the existing road system in order to provide a factual basis for planning an improved system. The injury accidents occurring in 1946-7 on 205 miles of the major network in Lancashire were analysed and it was found that 45 per cent of the accidents occurred at intersections. It was estimated that, had the traffic in this survey been travelling on a motorway, the number of accidents would have been reduced by 70 per cent. This investigation also showed that 20 per cent of accidents on the rural part of the network would have been prevented if stationary vehicles had not been permitted on the carriageway. This trend in accidents was confirmed at a very early stage in the construction of the motorway network by data from the highway network in Lancashire, an example of which is given below.

The information on accidents is provided by the police using a standardised report form (STATS 19), which has been revised at various intervals since 1949. The police complete a report form for every personal-injury accident and this information is collected by the transport department of government and is analysed by various organisations. This has provided a long-term database, which has been most useful in establishing trends in accident frequencies and severities as the motorway network has evolved.

Accidents before and after the opening of the M6 between Preston and Lancaster by-passes.

Severity

For 12 months before opening. Route A6

For 12 months after opening

A6

M6

Fatal

8

4

0

Serious

29

15

4

Slight

106

41

19

 

Much greater reductions in accidents are expected in areas where motorways effectively by-pass routes which pass through more densely built-up areas.

In order to assess the benefits of accident reduction, a method of quantifying the cost was developed in Professor J.H. Jones' report on road accidents to the Minister of Transport in 1946. The approach taken was to divide the costs into two groups: those that cause a diversion of current resources, and those that cause losses of future output because of death and injury. The diversions of current resources mainly involve repairing the damage to vehicles and other property, medical treatment, and the administration of insurance, the police and the law. Another item, which is difficult to measure, is the cost of the delays to other road users that results from road accidents; in the early years this item was not taken into account but, in later years, ways were found of quantifying it.

The measure used to assess loss of future output is its net present-day value, which is the difference between the future loss of output of those killed and injured and the future consumption of those killed, assuming a normal expectation of life. Both factors are discounted to give present day values. Generally, the figure used to assess the value of savings in accidents is the average cost of all accidents per personal-injury accident. In calculating this value, it is assumed that there are two damage-only accidents for every personal-injury accident. This average for 1961 was £625. One use of these values is in the economic assessment of road improvements, which can be based on the number of personal-injury accidents per year before and after the improvement.

(The history of valuation of life in assessing costs of road accidents is traced more fully in chapter 17 of this Volume.) Sir Frederick Cook was much exercised about the trends identified below, suggesting they were sufficient cause to design and build a safer type of road, namely the motorway system.

Design

The concept of a motorway as a dual carriageway road with grade separation, completely fenced in, normally with hard shoulders, exclusively for the use of prescribed classes of motor vehicles dates back to a draft Parliamentary Bill in 1906 (Goldstein 1980)

The Restriction of Ribbon Development Act 1935 included a section on standard widths for roads:

* The unit width of each traffic lane should be 10 ft. For two-lane carriageways carrying a large proportion of heavy vehicles the width should be 11 ft.

* Dual carriageways were desirable where traffic in peak hours was expected to reach 400 vehicles/h. Where more than two lanes were necessary, dual carriageways were preferred to widening single carriageways. Dual carriageways were often justified solely on grounds of public safety.

* The standard widths adopted should provide for future widening of the carriageways with cycle paths and footpaths as might be necessary. For this purpose there should be ample central reservations, margins and verges. Dual carriageways should be separated by a central reservation of the maximum width consistent with the layout of the road.

A list of minimum standard widths was drawn up ranging from 60 ft for single carriageways to 140 ft for dual (more than two-lane) carriageways with footpaths, cycle tracks and wide verges. A significant change at this time was that several dual carriageway schemes were included in the Report of the Road Fund for 1935/36 rather than the more usual widening of single carriageways. In 1945, H. E. Aldington (Deputy Chief Engineer in the Ministry of War Transport) presented a paper on the design of motorways (Institution of Civil Engineers, 1948) in which he set out a number of principles for the design of new roads for through traffic, including:

* Dual carriageways, separating the traffic in the two directions;

* The elimination of the crossing of all types of traffic at the same level;

* The elimination of stationary traffic;

* The gradual feeding-in of one traffic stream into another where they join, in order to reduce longitudinal traffic disturbances;

* The accurate alignment with regular longitudinal curves eased into the tangents by transition curves;

* The curves to be suitably super-elevated for the design speed;

* The careful and regular grading with vertical curves to give proper sight lines;

* A simple system of direction and warning signs;

* A uniform running surface.

These principal objectives were quantified by Aldington into a draft standard for motorways:

Design speed

75 miles per hour

Formation width

Two-lane dual - 93 ft.

Three-lane dual - 109 ft.

Marginal strip

1 ft wide at each side of the carriageway, flush with it and of a contrasting colour.

Carriageway

Dual two-lane - 22 ft excluding marginal strip.

Dual three-lane - 30 ft excluding marginal strip.

Verge

Normally 15 ft and clear of obstructions, but may be reduced. to 5 ft at bridges.

Central reserve

Not less than 15 ft; width to be maintained at bridges.

Curves

Radius not less than 3000 ft.

Gradients

Normal maximum 1 in 30 but up to1 in 20 permitted in hilly areas.

Lay-bys

To be provided at intervals to enable drivers to draw off the carriageway to rest or make minor repairs.

Roadside facilities

No frontage access but places to be provided for the supply of petrol and for refreshment.

Bridges

Modern designs with appropriate forms of construction.

Pavement design and surfacing

Attention to foundations is essential. All road surfaces should, as far as practicable, be non-skid.

 

When the decision was made to construct motorways, the Ministry of Transport had to decide on the standards of layout and construction to be used.

The considerable attention given earlier to this aspect, as described above, provided a basis and the following main features were adopted at that time (Baker 1960):

Carriageway

Dual 24 ft (two-lane) or 36 ft (three lane) with 1 ft marginal strips of contrasting colour on both sides of each carriageway.

Central reserve

13 ft wide.

Hard shoulders

8 ft wide (later increased to 10 ft).

Verges

3 ft 6 in.

Design speed

70 mile/h.

Horizontal Curvature

Minimum radius 2,865 ft (2° curvature).

Gradients

Normal maximum 3 per cent; 4 per cent in hilly country.

Speed change lanes

12 ft wide, 400 - 800 ft long.

Vertical clearance

16 ft 6 in.

 

Several junction layouts were devised by the Ministry and it was considered that the frequency of junctions should be one about every 10 miles in rural areas but more frequent in developed areas, with a minimum spacing of 3 miles.

Service areas were to be provided in areas set back from the motorway with access at the rear to the minor road network for staff and supplies to these areas. Parking space was to be provided and the services were to include fuel supplies and refreshments.

Motorways were to be equipped with an emergency telephone system located on the hard shoulder with telephones at about 1 mile intervals, connected directly to a police station.

In 1966, the standard cross-section for motorways was reduced by 13 ft from 129 ft to 116 ft overall. This was achieved by reducing the width of each verge by 5 ft 6 in reduction in the width of each hard shoulder and the omission of a separate 1 ft marginal strip adjoining the central reserve. It was estimated that these changes would reduce construction costs by about 3 per cent. They were, however, of concern to the local authorities and to the motoring organisations. In particular, they were concerned about the effect that the reduced distance between opposing carriageways might have on road safety. It was recommended that a detailed study of the likely effect of the reduction in motorway standards on accident rates should be made by the Laboratory and that any future decision on this matter should be taken in the light of this study. However, the Ministry considered that their decision to install safety barriers on the central reserves of motorways had obviated the need for this study.

At the same time as there was inter-urban motorway construction there was discussion of motorway construction in urban areas, as well as the development of other forms of transport.

Typical standards for urban motorways were recommended (Department of the Environment, 1971) and are set out below:

Design Speed 80km/h
Carriageway width dual two lane 7.3m
dual three lane 11.0m
Verge (paved, no hard shoulders) Normal width 2.75m
Minimum width 2.0m
Clearance(back of verge to safety fence) 0.6m
Central reserve Normal width 3.0m
Desirable minimum width Without lighting columns 2.0m
With lighting columns 2.4m
Absolute minimum 1.8m
Clearance(edge of carriageway to safety fence) 0.6m
Design flow (peak hour in one direction) Dual two-lane 3,600 veh/h
Dual three-lane 5,700 veh/h

 

An urban motorway is essentially a dual carriageway road restricted to specified classes of motor vehicle with no intersection at grade and a somewhat lower standard than that of rural motorways with more frequent access points. However, there are relatively few miles of motorway in urban areas in Britain and even fewer within the heavily congested areas of cities. They were built for various reasons: the M5 and M6 near Birmingham, part of the national strategic network, were routed near to major sources of traffic; the M621 in Leeds links national routes to terminal areas in the city; and the Mancunian way in Manchester provides relief to congested areas.

Thus the cross-section for motorways has changed over the years as can be seen from the flollowing table. The change for lay-bys to hard shoulders took place before the first motorways were built and the use of 12 ft lanes was an early innovation. When standards were being formulated during the 1950s there was a proposal to use 4ft 6in wide hard shoulders constructed so that grass would grow on them.

However, shoulders of that width had been found to be dangerous in Germany so it was decided initially by the Ministry to adopt an 8ft width. Nevertheless, this was increased to 10ft in the light of the early experience on the Preston By-pass and the southern sections of the M1. The early standard of 11ft for the width of a lane was questioned when considering permitting goods vehicles and buses to be built up to 8ft wide. While there was little evidence on desirable lane widths from other countries but there seems to have been a consensus among highway engineers that 12ft was probably about right and that was adopted as the Ministry standard. The design speed for rural motorways is 70 mile/h and for urban motorways 50 mile/h. The rural standard is as the following table shows.

Rural motorway cross-sections

 

1948

1963

1968-9

1980

Verge 15 ft 10 ft 5 ft 1.5 m
Hard shoulder 4.5 ft with lay-bys 100 ft long 10 ft wide 10 ft 9.5 ft 3.3m (10.8 ft)
Nearside marginal strip 1 ft 1 ft 1 ft 0.2m (0.6ft)
Carriageways Normally 22 ft D2 : 24 ft
D3: 36 ft
Normally D2 : 24 ft
D3 : 36 ft
3 lane: 11m - (36 ft)
2 lane : 7.3m - (23.9 ft)
Offside marginal strip 1 ft 1 ft 6-8 in wide edge strip at carriageway surface 0.2m (0.6ft) on carriageway
Central reservation (half width) 7 .5 ft 6.5ft 6.5 ft 2m (6.5 ft)
Formation width dual 3 lane 109 ft 129 ft 116 ft 35.6m (117 ft)
Formation width dual 2 lane 93 ft 105 ft 92 ft 28.2m (92.5 ft)
Source Aldington Smith Select Committee DTp

 

Considerable importance has been attached to achieving a 'flowing alignment' (Spencer, 1948). While taking into account basic balances of cut and fill, there should be a phasing of horizontal and vertical curvature so as to present a pleasing appearance of the road, obviating the appearance of kinks or optical misalignments. This concept was subsequently set out in a Ministry of Transport memorandum (Department of Transport, 1975).

On rural motorways, horizontal curves have an absolute minimum radius of 510m and a desirable minimum of 960m. Transition curves are desirable on curves with a radius less than 3000m and are regarded as essential on curves with a radius less than 1500m. Vertical curves were to be provided at all changes of gradient and the curvature has to be large enough to provide safe stopping sight distances. At intersections horizontal and vertical curves were to give as generous a sighting distance as possible. Gradients were not normally to be more than 3 per cent but in hilly country 4 per cent might be acceptable.

Motorway interchanges were built to several designs, the most common being grade separated with a roundabout over the motorway. This type provides good visibility for drivers and helps to slow traffic leaving the motorway. One of the most economical types is the diamond arrangement where the slip roads to and from the motorway make simple junctions with the side roads but this type of junction is not regarded as being suitable when side roads carry appreciable volumes of traffic.

Alignment

The geometric design of a motorway has to be harmonised with the terrain through which it passes so that it is pleasing to the user and is compatible with the landscape. The alignment of motorways may have an effect on the driver and, at an early stage, long straights were considered to invoke dangerous hypnotic and monotonous effects. A specific objection to long straights is that they increase an effect known as 'parallelism', which results from the impression given by several construction features running in parallel, for example, lane markings, the edges of the carriageway and the motorway fencing. The designer should aim to break-up the long-distance parallel effect that is not in accord with the surroundings. In general, it is better to have an alignment with a gentle curvature that presents an ever changing backcloth to the driver, thereby reducing the adverse effects of monotony. Horizontal and vertical curves should be as long as possible.

In setting standards for alignment 'sight distance' was a major consideration. 'Sight distance' was defined as the clear distance over which a driver would be able to see an obstruction from an assumed eye level of 3.5 ft above the surface of the road. The minimum sight distance on rural motorways was set at 950 ft in both the vertical and horizontal plans, which is equivalent to the minimum stopping distance at 70 mile/h. At interchanges, where visibility is of particular importance, longer sight distances should be provided, if possible.

On urban motorways, the minimum sight distance is set appropriately for the design speed. For example, at 50 mile/h the minimum sight distance is 425 ft, which is the minimum stopping distance at this speed.

The maximum longitudinal gradient for a motorway in rural areas is 3 per cent but in hilly country a 4 per cent gradient may need to be accepted. When commercial vehicles form a substantial portion of the traffic, an extra lane may be required on 3 per cent grades over 1500 ft in length and on four per cent grades over 1000 ft long.

The current alignment standards were implemented taking into account the following general principles: The horizontal and vertical curve should be as generous as possible; and, to relieve the monotony of driving on a road with good extensive forward visibility, long sections of the road should be aligned to give a view of some prominent feature ahead.

Changes arising from motorway accident analysis

Information on all personal-injury accidents is accumulated nationally through the STATS 19 procedure, which is held centrally on a database. In subjecting this information to analysis in the past 20 years TRL’s computing capacity has assisted greatly.

Integrated motorway traffic control centres

Chapter 16Working closely with the Police, the Highways Agency and other bodies like the Welsh Office for the M4, and the Northern Ireland Roads Service in Belfast, huge advances have been made in provision for improved motorway safety, early warning of hazards, controlled access and rescue after breakdown or accident, through control centres using the latest electronic and communications links.. Many more details of how that has been effected since 1958 appear in Volume 2.

Traffic control for Belfast’s motorways and the city bridges was introduced with the opening of the M2 Laggan Bridge in 1997. Photograph by courtesy of the Northern Ireland Roads Service.

 

Conclusions

Motorways have been designed as relatively safe roads. They have proved to be as safe as their designers hoped. Indeed only 6 per cent of the fatalities occurring on roads took place on motorways in 1999. Consequently motorways have engendered a downward trend in accidents, compared with the upward trtend so unpleasantly obvious from the 1920s into the 1960s. Safety analysis as indicated earlier, shows that accidents are attributable to a combination of four components, the driver, the vehicle, the road environment, and varying meteorological conditions by day and night. The TRL’s tradition of concern with these issues has done much to neutralise them.

Three are subject to various legal requirements for due care but, with the combination of the components varying with each accident, generalisations about the causes of accidents are of limited value. However, it is statistically significant that between 1986 and 1997 traffic flows on motorways have increased by 92% but only by 37% on all other categories of road. Concern about the severity of fog related accidents on motorways remains. There is also rightly a concern about the increasing severity of motorway collisions especially in the context described in a TRL Project Report of 1999 by M.J. Watkiss, Accidents on motorway hard shoulders and efforts to improve safety, PR/TT/082/99.

However, such work cannot meaningfully be considered alone. It needs to be seen alongside the accident records on and alongside high speed dual carriageways, which have been the subject of separate studies funded by the AA Road Safety Research Foundation on the dangerous non-motorway roads in Cambridgeshire. These roads compared poorly with the motorways in respect of safety, curvature, the constraint of excessive speed, and driver distraction.

It remains to be seen whether the safety record of London's former urban motorways will be higher or lower following the removal of safety features like roadside telephones and the usual motorway restrictions on their vehicular usership. Nonetheless the markedly lower accident rate on motorways indicates at least that, in terms of safety, there is definite positive value in the carefully considered design features and legal exclusions which are unique to them.