This report reveals the potential road safety impacts of a possible permanent implementation of automated speed camera systems in a larger scale. The report also includes recommendations on a possible implementation. Firstly, the report presents international experiences with automated speed camera systems from a number of countries that have evaluated the system as well as the results from the Danish trial with ten fixed position speed cameras carried out from January 2009 to January 2010. Afterwards, estimates of the probable effects on accidents with personal injuries in case of a general introduction of automated speed cameras in Denmark are calculated based on statistical modelling of Danish accident data. Based on international best practices and knowledge about Danish conditions, two alternative scenarios for a general implementation of automated speed cameras in Denmark are presented. Finally, a socio-economic analysis that partly calculates the benefit-cost ratio for the installation and operation of automated speed cameras, partly the traffic fine revenue in case of installation of automated speed cameras is carried out to assess the total effects for the Government finances. Fixed position speed cameras consist of individual poles or series of poles along the road on which the cameras are mounted. The pole is connected to coils in the road that measure the speed of the passing vehicles at a specific point. In Denmark, as in Norway, Sweden and Finland, both the licence plate and the driver violating the speed limit are photographed, since the driver is legally responsible for the speeding violation. In other countries (e.g. France, the Netherlands, Belgium and Great Britain) the owner will be held legally responsible for the speeding violation, and it is therefore only necessary to photograph the licence plate and send the fine notice to the owner. Point-to-point speed cameras (also called section control) use the driver’s average speed over a given road section to assess whether a speeding violation has been committed. The poles with the cameras are placed at each end of the section and a photo of the car’s licence plate is taken, and when two matching licence plates have been registered, the average speed is calculated based on the time that has passed between the two pictures. All the passing vehicles are photographed. International experiences with automated speed cameras particularly comprise fixed position speed cameras and to a smaller degree point-to-point speed cameras. The experiences are positive, both with regard to reducing the speed and the number of accidents. The following countries are included in this overview of the evaluation of fixed position speed cameras: Norway, Sweden, Finland, the Netherlands, Great Britain, France, Belgium, Australia (NSW), Australia (VIC). All the countries mentioned have documented a positive effect on the cars’ mean speed, and all have documented a positive effect on the number of accidents with personal injuries. The average effect is a decrease in the number of accidents with personal injuries by 20-25%. The reduction of the mean speed lies around 7-10% on road sections with automated speed cameras. At the same time, most evaluations show that the highest speeds are reduced most, especially the speeds exceeding the speed limit. Moreover, the standard deviation of the mean speed is reduced. All reductions contribute to enhanced traffic safety. Point-to-point speed cameras prevent the so-called “kangaroo driving” meaning that the driver reduces the speed considerably when he approaches an automated speed camera and immediately accelerates again when he has passed it. In this way, with point-to-point speed cameras the speed reduction is more even resulting, among other things, in a smoother traffic flow. As point-to-point speed cameras are a relatively new phenomenon, there are no well-documented scientific studies of the effect on speed and accidents, but preliminary studies seem to show a positive effect. Compared with fixed position speed cameras the reduction of the mean speed seems to be smaller when passing the first pole (3-4%), but on the other hand the lower speed is apparently maintained along the entire road section. The effect on accidents with personal injuries seems promising for point-to-point speed cameras, maybe even better than for fixed position speed cameras, but the material is still too limited to be used for drawing final conclusions. Point-to-point speed cameras are under evaluation in several countries, and the coming years will provide documentation and clarification of the exact magnitude of the effects. In Denmark, the police normally carry out speed control using laser measurements (manual control), radar (mobile automated speed control) or time measurement (control performed from a vehicle). In case of mobile automated speed control, the speed measurements are made either from a vehicle or using a tripod. In addition to the traditional speed control, the National Police (Rigspolitiet) and the police districts “Midt- and Vestsjællands Politi” and “Nordsjællands Politi” carried out a trial with fixed position speed cameras in 2009. Ten poles were installed, six on main roads and four in urban zones. Six cameras were moved around between the poles, so that all measurement points were assigned the same measurement time to the greatest possible extent. During the trial period almost 20,000 fine notices for speeding violations were issued. The installation of automated speed cameras clearly reduced the speed of the vehicles when passing an automated speed camera. The reduction both applies to the mean speed and the standard deviation of the mean speed as well as the 85%-fractile4. On the main roads, the mean speed dropped by 9.1 km/h on average in the enforcement direction on weekdays (12%) and 12.1 km/h during weekends (14%). On urban roads the reduction in mean speed was not so big, namely 10% and 13%, respectively. The share of passenger cars and mini vans driving at a self-chosen speed and exceeding the speed limit was determined at three poles: two in a rural zone (speed limit of 80 km/h) and one in an urban zone (speed limit of 50 km/h). Before the installation of automated speed cameras this share, with figures for weekdays first and figures for weekends in brackets, was 80% (81%) and 65% (76%) at the two poles in the rural zone and 52 (64%) in the urban zone. After the installation of automated speed cameras the same share of offenders was 15% (21%) and 6% (10%) in the rural zone and 22% (28%) in the urban zone. The mean speed also dropped in the opposite direction of the trial; this decrease was between one third and half the decrease found in the enforcement direction. The automated speed cameras tested in the trial seem to have registered other speeding offenders than those registered by the methods of speed enforcement used so far. The most important observation was that the share of local offenders was considerably smaller in case of automated speed cameras as compared to mobile automated speed cameras and traditional methods of enforcement. This indicates a behavioural regulation among particularly the local road users in the areas where the automated speed cameras were placed. Finally the speeding violations committed during the weekends were more often registered by automated speed cameras than by other methods of speed enforcement. The decreases in the mean speeds found in the Danish trial are bigger (10-14%) than those found internationally (7-10%). The overall purpose of automated speed cameras is to lower the number of speed-related accidents with personal injuries by reducing the number of drivers exceeding the speed limit In principle, it can be assumed that the effects of automated speed cameras will be biggest on sections where the cars exceed the speed limit, where the number of persons killed and injured in traffic accidents is big and where it can be assumed that the number of accidents is affected by the speed. The literature study shows that the choice of the places where the automated speed cam eras are to be installed is generally based on knowledge about an increase in accident density or in the number of killed and seriously injured persons per km and year. The biggest effect is expected on roads where the speeding violations are significant. To assess the consequences of a general introduction of automated speed cameras in Denmark, this report outlines two scenarios. The scenarios describe two different ways of large-scale imple mentation of fixed position speed cameras in Denmark. The scenarios are general and operate with road types and not individual road sections. When installing automated speed cameras it is always necessary to carry out an assessment of the road section in question with respect to accident frequency, speed, congestion situation and technical suitability. However, other kinds of considerations may also be taken into account, e.g. aspects of regional coverage. Both scenarios focus on main roads, since the trial with automated speed cameras was primarily carried out on main roads and since these roads are characterised by a high accident density and a high share of serious speed-related accidents. In addition, both scenarios suggest that a small number of automated speed cameras is installed on urban roads and motorways (cf. Table 0.1). When the experiences from these cameras have been gathered, the number of cameras may be increased. n the calculations it has been assumed that automated speed cameras are installed in series of 3- 4-5 poles with an average distance of approximately 5 km between the individual poles placed in a series on main roads (so-called serial speed cameras). Both Sweden and Finland have had good experiences with this procedure. It is a prerequisite that the camera is only active in one of the poles of the series. In this way the advantages of fixed position speed cameras and point-to-point peed cameras are combined into one system in which speeding violations are registered once in a relatively simple way, but resulting in a smooth traffic flow with reduced speed and only minor “kangaroo driving”. The scenarios operate with the possibility to install automated speed cameras on either one side of the road or on both sides of the road. If the poles are only installed on one side of the road, a larger number of line kilometres can be covered by automated speed cameras with the secondary effect that the speed in the opposite direction of travel is reduced. If the poles are installed on both sides of the road, maximal effect on speed and accidents with personal injuries in both directions is obtained, but along less line kilometre (provided the number of poles is constant). It must therefore be assessed on an ad hoc basis which of the systems is preferred for implementation. When installing the automated speed cameras it should also be decided how many hours per day the cameras installed should be working, as there must be concordance between the number of registered speeding violations and the administrative capacity available so that the drivers are not photographed without receiving a fine notice. In the three outlined scenarios main roads were chosen for the installation of automated speed cameras, i.e. state roads that are not motorways or dual carriageways, and major municipal roads with regional traffic and high traffic density (AADT above approximately 2,000 cars). A considerable part of these roads is also covered, i.e. from 9% to 43% in the two scenarios (automated speed cameras in one direction). The main roads included are very different, both with respect to traffic intensity, accident density, congestion as well as technical suitability. Therefore, it is crucial for judging an optimal effect to carry out an assessment of each section before large-scale implementation of automated speed cameras. Furthermore, the tendency towards minor speed reductions between the poles (“kangaroo driving”) which is the weakness of serial automated speed cameras should be met by accompanying measures. It is recommended to continue the use of mobile automated speed control, even in road sections with serial automated speed cameras in order to benefit from the unforeseeable effect of this measure. It is also recommended to ensure that the poles as far as possible are prepared for future use for point-to-point speed cameras. If experience with point-to-point speed cameras shows that the “kangaroo driving” reduces the effect of fixed position speed cameras considerably, it will be possible to carry out trials with point-to-point speed cameras on the included road sections. The equipment can be transferred to other sections which in itself will probably have an effect comparable to mobile automated speed control. Both calculated scenarios have a clear positive effect on road safety due to the reduction of accidents with personal injuries, cf. Table 0.2. In these calculations the savings from the trials on roads in cities, dual carriageways and motorways are not included, although they will contribute with a minor reduction of accidents with personal injuries. In addition to this, conservative assessments in the scenarios are calculated using international effect assessments. Finally, calculations based on installation of poles in both directions placed ”face to face” are included, covering half of the number of line kilometres. The bold figures in the table have been calculated using the effect assessments from the Danish trial with fixed position speed cameras and with Danish figures for the relationship between speed and accidents with personal injuries. Furthermore it is assumed that the automated speed cameras are only installed in one direction. Since all scenarios have a positive effect on road safety, the choice of effect level will determine how to carry out a general implementation of automated speed cameras. In the economic analysis the costs for establishing and operating automated speed camera systems have been included, as well as administrative costs to collect fines. On the benefit side the reduced costs of traffic accidents, fuel savings and reduced CO2-emissions have been included as savings. The case study for the economic analysis is a typical major trunk road section with an AADT of 7,500 vehicles and an average accident density of 0.22 personal injury accidents per kilometre road. By installing automated speed cameras placed in one direction with an average distance of 5 km the benefit-cost ratio is calculated at 1.4 under the basic assumptions. Hence, the installation of automated speed cameras is therefore probably economic feasible, however, this conclusion is very dependent on uncertain assumption. The benefits more or less balance the costs no matter whether the automated speed cameras are placed in one or two directions. However, the result becomes negative if international effect assessments are used. It should be noted that two thirds of the total yearly operating costs under the given prerequisites are made up by administrative processing costs for the collection of the fines issued for speeding violations. The administrative processing costs increase linearly with the part of the year during which the automated speed cameras are active. In the case study this has been set at 15% weighing on the one hand minimising the administrative costs against on the other, consideration for a speed effective control. If the automated speed cameras are active 100% of the time, the administrative costs becomes greater than the saved accident costs, and the benefit-cost ratio is changed to 0.3. The speed fine revenue is not included in the cost-benefit analysis, as fines are not considered as a benefit in an economic perspective, but as a transfer of means from the citizens to the State. However, if the issue is considered from the point of view of government budget, it can be concluded that the fine revenue is much bigger than the costs for establishing, operating and administrating the automated speed cameras. With an active enforcement time of 15% of the automated speed cameras, the yearly fine revenue per line kilometre is approximately 200,000 DKK. If the cameras were active 100% of the time, the fine revenue would increase to 1,300,000 DKK per line kilometre. The overall conclusion from the study is that a general introduction of automated speed cameras in Denmark is expected to have a positive impact on the speed and will result in a yearly reduction of about 50 accidents with personal injuries for the scenario with 500 poles. A possible general introduction of automated speed cameras will have a reasonable economic feasibility, and the traffic fine revenue to the State is expected to be far bigger than the costs for establishing and operating the systems. Apart from the effects included in the assessment in this report, a high coverage of automated speed control may result in systemic effects in terms of a general speed reduction on the road types in question. (Author/publisher)
Abstract