Scholarly article on topic 'A Better Knowledge of Powered Two Wheelers Accidents'

A Better Knowledge of Powered Two Wheelers Accidents Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
Academic journal
Transportation Research Procedia
OECD Field of science
Keywords
{"Powered Two Wheelers" / accidents / database}

Abstract of research paper on Economics and business, author of scientific article — Nicolas Dubos, Berengère Varin, Olivier Bisson

Abstract In France, the knowledge about road accidents is mainly based on the data collection and subsequent feeding of the national road accident database by the police. This is a very comprehensive database from a quantitative point of view which needs to be improved from a qualitative point of view. Indeed, some data are necessary to have a very precise understanding of the accidents (accident mechanisms, vehicle maneuvers, etc.) but they are not present in the French national accident database. In order to get some qualitative data, the partners of VOIESUR project have analyzed fatal and injury accidents that occurred in the year 2011. This analyze is reported in a database. From this database, thorough analyses have been carried out on different topics (children, older road users, etc.). The present study addresses the powered two-wheelers (PTW). His objective is to strengthen the knowledge about PTWs accidents, in order to improve PTWs safety. This study has shown how specific were PTWs in terms of dynamics or other factors such as alcohol impairment, speed, short practice of vehicle. The analysis of the different accident factors, and in particular the new knowledge brought by VOIESUR (speed used in accidents, maneuvers performed in emergency situations, etc.) should enable the decision makers to implement concrete actions towards PTW users, for whom safety improvements are identified, and also for other users involved in PTW accidents. The elements provided in the present study should encourage the implementation of measures adapted to PTW diversity and usage, in terms of categories (customs, roadsters, etc.), and engine sizes (mopeds, light and heavy motorcycles). Lastly, the role of road infrastructure has also been scrutinised, and a greater vulnerability of PTW is put forward. The study shows a great potential that could be grasped by road managers to improve the safety of PTW users (roadside obstacle management, provision of hard shoulders in case of run-off-the-road situations, etc.). From a global viewpoint, the richness of the variables shed a new light on both the nature and complexity of PTW accidents in France.

Academic research paper on topic "A Better Knowledge of Powered Two Wheelers Accidents"

ELSEVIER

www.elsevier.com/locate/procedia

6th Transport Research Arena April 18-21, 2016

A better knowledge of powered two wheelers accidents

Nicolas Dubos a'*, Berengère Varina, Olivier Bissona

a Cerema-Center for studies and expertise on Risks, Environment, Mobility, and Urban and Country Planning (Cerema), France

The partners of VOIESUR project are LAB (Laboratoire d'Accidentologie et de Biomécanique), CEESAR (Centre Européen d'Etudes de Sécurité et d'Analyse des Risques), IFSTTAR (Institut Français des Sciences et Technologies des Transports, de l'Aménagement et des Réseaux), and Cerema. The project has gotten the support of Fondation MAIF and ANR (the French national research agency).

Abstract

In France, the knowledge about road accidents is mainly based on the data collection and subsequent feeding of the national road accident database by the police. This is a very comprehensive database from a quantitative point of view which needs to be improved from a qualitative point of view. Indeed, some data are necessary to have a very precise understanding of the accidents (accident mechanisms, vehicle maneuvers, etc.) but they are not present in the French national accident database. In order to get some qualitative data, the partners of VOIESUR project have analyzed fatal and injury accidents that occurred in the year 2011. This analyze is reported in a database. From this database, thorough analyses have been carried out on different topics (children, older road users, etc.). The present study addresses the powered two-wheelers (PTW). His objective is to strengthen the knowledge about PTWs accidents, in order to improve PTWs safety.

This study has shown how specific were PTWs in terms of dynamics or other factors such as alcohol impairment, speed, short practice of vehicle. The analysis of the different accident factors, and in particular the new knowledge brought by VOIESUR (speed used in accidents, maneuvers performed in emergency situations, etc.) should enable the decision makers to implement concrete actions towards PTW users, for whom safety improvements are identified, and also for other users involved in PTW accidents. The elements provided in the present study should encourage the implementation of measures adapted to PTW diversity and usage, in terms of categories (customs, roadsters, etc.), and engine sizes (mopeds, light and heavy motorcycles). Lastly, the role of road infrastructure has also been scrutinised, and a greater vulnerability of PTW is put forward. The study shows a great potential that could be grasped by road managers to improve the safety of PTW users (roadside obstacle

Available online at www.sciencedirect.com

ScienceDirect

Transportation Research Procedia 14 (2016) 2274 - 2283

Procedía

TRANSPORT RFSÉARCH ARENA

* Corresponding author: E-mail address: Nicolas.Dubos@cerema.fr

2352-1465 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of Road and Bridge Research Institute (IBDiM) doi: 10.1016/j.trpro.2016.05.243

management, provision of hard shoulders in case of run-off-the-road situations, etc.). From a global viewpoint, the richness of the variables shed a new light on both the nature and complexity of PTW accidents in France.

©2016 The Authors.Publishedby ElsevierB.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of Road and Bridge Research Institute (IBDiM) Keywords: Powered Two Wheelers; accidents; database

1. Introduction

As a result of recent improvements in road safety in France, our country has one of the lowest accident rates in Europe, having fallen from 8,000 to 4,000 road deaths in 10 years. These improvements demonstrate the importance of ensuring that all stakeholders (politicians, businesses, associations, public, etc.) work together in pursuit of the ultimate goal, i.e. saving lives. However, this goal can only be achieved if the effort undertaken is genuinely aligned with actual needs. As such, it is important to conduct an analysis of the current situation. This, in turn, will enable us to establish whether the proposed measures are suited to the reality on the ground, and to ensure that both current and future safety systems are better targeted, in line with improvements in technology.

Various road safety data collection systems were implemented in France in the 1990s and 2000s, with a view to analysing the road safety situation during these periods. These systems focused primarily on fatal accidents, and therefore failed to provide a comprehensive overview of reality. The aim of the VOIESUR project was to conduct a new situation analysis for 2011 and to extend the scope of observation beyond fatal road traffic accidents, to include personal injuries.

The project also included targeted studies on specific populations, especially vulnerable users such as pedestrians, children and older road users, or powered two-wheelers (PTWs) which are the object of this study. The main objectives of the VOIESUR project were as follows:

• To provide a comprehensive road traffic accident information system.

• To conduct an assessment of current road safety problems.

• To study, in detail, certain particularly vulnerable road user groups (pedestrians, PTWs, children and the older road users).

• To identify potential improvements to infrastructure, road behavior and vehicles.

• To provide a system capable of anticipating road safety problems over the next 10 years.

• To continue the work conducted in the 1990s and 2000s.

2. Innovative approach

2.1. Content of the VOIESUR accident database and data reliability

Unlike other projects, the VOIESUR project was based on a detailed analysis of road traffic accident reports -paper records produced by the police covering fatal accidents and personal injuries. The aim of VOIESUR was to capitalise on the wealth of data contained in these reports. In this project, all fatal accident reports were collected and coded. A different approach was adopted for personal injury accident reports, with 1/20th of these reports selected at random and coded (due to cost and time imperatives). A total of 8,500 accident reports were coded, of which 946 PTWs fatal accidents and 1095 PTWs personal injury accidents. For each accident, there were 350 variables, and each variable can have several values.

It is important to stress that, overall, the quality of the data used is better for fatal accidents than for personal injuries, since these data depend, in part, on the quality of the information recorded by the police. Naturally, the police tend to record more detailed information in relation to a fatal accident. Quality of information coded in VOIESUR database depends on policy data. Then, a quality control (for example consistency control between several variables) has been realized for all coded accidents, which assume a good reliability of the coded

information. As a consequence, we suppose that there was not unclear information used. But some variables were particularly unknown, as mechanical modification of moped (80% of unknown), or as lights of PTW switched on (82% of unknown). On the contrary, some variables have a good rate of known values, as motorcycle driving licence (90%), wear of crash helmet (92%) or blood alcohol level (85%). Overall, rate of known values was better in fatal accidents than in personal injury accidents.

The approach adopted for the VOIESUR project represents a half-way point between using raw data recorded and reported by the police, and using detailed accident studies produced by accident research specialists. The benefit of this approach is that it provides sufficiently detailed information while covering a large number of accidents.

2.2. Resources

The establishment of the VOIESUR database involved an extensive and cumbersome process, including the following tasks:

• The initial creation of road traffic accident reports by the police, which were then submitted to insurance companies.

• The collection of 8,500 reports from insurance companies.

• General coding of the variables contained in the accident reports (e.g. the age of the people involved, vehicle types, road types), requiring the recruitment of 10 full-time coders over a period of several months.

• Expert coding for specific points (vehicle speeds, accident scenarios).

• Quality control of the database throughout the project.

A multidisciplinary approach was adopted for this project, involving infrastructure specialists, vehicle specialists and user behaviour experts. This, in turn, made it possible to determine infrastructure-environment-vehicle interactions and to identify specific problems and counter-measures, via an integrated approach to safety.

2.3. Data example: speed

The accident report coding process produced a range of qualitative data. For example, by reconstructing the accidents from these data, it was possible to determine the speed of the vehicles both before the accident and at the moment of impact. In order to do this, the accident reconstruction experts involved in the project used the various items of information contained in the reports to establish the speeds of the majority of the vehicles involved:

• A scale plan containing: the final position of the vehicles, the impact zones, the marks left by the vehicles (where applicable) and the origin of the vehicles.

• A description (road condition, surface condition) of the infrastructure (roadway, verge) on which the impact took place.

• Photos providing information about the nature and severity of the impacts.

• Witness reports about the sequence of events leading up to the accident.

In addition to this information contained in the accident reports, the experts also based their analyses on the laws of physics (conservation of kinetic energy and momentum, etc.), as well as their own knowledge. Their ability to establish vehicle speeds was therefore dependent on the volume of information contained in the accident reports, as well as the specialisation and experience of the experts concerned.

2.4. PTW study: objectives and methodology

For PTWs, the specific objectives of the research were to study road safety risks among these users and to improve current knowledge of accident mechanisms. This study was unique in terms of both its multiple objectives and its scale.

The overall objectives were as follows:

• To enhance current knowledge of PTW users: knowledge of accident risks and occurrence, assessed against user socio-demographic variables, infrastructure and vehicle characteristics and risk exposure.

• To identify accident mechanisms among PTW users.

Finally, results should allow:

• to reinforce the safety equipments worn by PTW users,

• to suggest ideas on how the public authorities could raise awareness among the PTW users and develop repressive policies towards them,

• to give road managers some concrete views about the influence of road infrastructure on PTW accident causation and severity.

What was unique about this work was that it considered the specific characteristics of PTWs from the perspective of multiple user types. Indeed, the PTW category covers a range of different practices (journey type, distance, type of network used, etc.) and often atypical driving styles (risk-taking). It also covered a diverse range of users (age bracket, licence requirement, specific driving training). Finally, it covered an array of different PTW types, each with its own unique dynamic behaviour.

The methodology therefore involved differentiating between fatal accidents and personal injury accidents, which differ substantially in nature as demonstrated, in France, by Van Elslande et al. (2008). The analysis therefore involved assessing PTWs as a whole, and dividing PTWs into the three categories defined at national level, i.e. mopeds, light motorcycles and heavy motorcycles (as indicated in the ONISR report, 2012). Finally, for each theme, a comparison was drawn with light vehicles in order to highlight the specific characteristics of PTWs. Similarly, PTW accidents were compared with national surveys (CGDD, 2013), providing information about the types of PTW on the road and driver profiles (age, driving experience, etc.).

3. Results

3.1. General snapshot of PTW accidens

The investigations conducted during this study confirmed the difference between fatal accidents and personal injury accidents among PTWs, along with the associated accident factors.

16.5% 13% 8.1%

Fig. 1. Three main PTW fatal accident configurations (PTWs shown in green).

The most common fatal accident configuration (16.5% of fatal accidents) takes place in open countryside (69%) for heavy motorcycles and in urban areas for light motorcycles and mopeds. Among heavy motorcycles, the majority of accidents involving the vehicle leaving the roadway take place on a left-hand bend (59%). However, this is not the case for mopeds, which generally leave the roadway on a straight section of road. For this accident configuration, it comes as no surprise that user-related factors are particularly important. Since this type of accident involves a loss of control, initial risk-taking (especially alcohol use and speed) is particularly important. Indeed, in

the most common fatal accident configuration (i.e. first picture of figure 1), some 39% of heavy motorcycle users had a blood alcohol level exceeding 0.5 g/L. Meanwhile, more than three quarters of light motorcycle and moped users had a blood alcohol level exceeding 0.5 g/L. In terms of speed in the most common fatal accident configuration (i.e. first picture of figure 1), the PTWs were exceeding the speed limit in more than half of all cases, with heavy motorcycle users exceeding the speed limit by a significant margin. Poor knowledge of the PTW also plays an important factor, with the user having owned the PTW concerned for less than a year in the majority of cases. However, the high proportion of users using a PTW without a motorcycle driving licence also has a major impact (20% of heavy motorcycle users and 33% of light motorcycle users). In the majority of cases, a failure to execute a driving task is the most common cause of PTW faults. This proportion is particularly high among mopeds, which suffer from significant general fault rates.

Details of PTW personal injury accidents are not given here, since the accident rates for this type of accident are much less clear.

Overall, and when compared with other vehicle categories, PTWs are more frequently involved in accidents:

• at intersections

• at car parks (entering or leaving a car park)

• following overtaking.

Generally speaking, the majority of PTW accidents are multiple vehicle accidents (81%). However, this figure should be assessed with caution due to the fact that, at national level, single vehicle accidents involving PTWs may be under-reported in relation to other categories of user. Nevertheless, this remains an important trend. Even for fatal accidents (where the issue of under-reporting is less significant), the figures show that more than 65% of PTW accidents are multiple vehicle accidents, which is significantly higher than for light vehicle accidents (55%). This means that interaction with other vehicles is therefore a significant factor in PTW accident rates, more so than for other types of user. Although single vehicle PTW accidents are less frequent than multiple vehicle accidents, these single vehicle accidents are nevertheless the most serious accidents (6.7 deaths per 100 PTW users involved).

There are significant differences in PTW accident rates by severity. The data surrounding personal injury accident rates are much less clear than those for fatal accidents. The six main fatal accident configurations account for 56% of all fatal accidents, whereas the six main personal injury accident configurations account for just 35% of these accidents in total. This lack of clarity also extends to the environment in which the accident takes place (urban area/open countryside). Fatal accident rates are the same in urban areas and in open countryside, whereas personal injury accident rates differ significantly depending on the environment concerned. This significant diversity among personal injury accidents reflects the inherent problems in characterising the issues relating to this type of accident, particularly given that PTW personal injury accidents are 24 times more common than fatal accidents.

In fatal accidents, PTW user task execution problems are to blame in 30% of cases. In personal injury accidents, however, the most common problems are situation assessment (23.6%) and detection of other vehicle or element (21.5%). Furthermore, when compared with other vehicles, diagnosis, assessment, decision-making and execution faults are higher among PTW users than among other vehicle types. However, detection and general fault (global loss of driving ability) problems are lower among PTWs than among other types of vehicle. Moreover, the faults differ according to the category of PTW concerned. In both fatal accidents and personal injury accidents, general faults affect mopeds more than motorcycles. This is particularly interesting when compared with the generally higher alcohol usage rates among moped users.

The specific characteristics of each PTW category become even clearer when it comes to emergency manoeuvres, with higher-powered vehicles involving harder braking, and with users of these high-powered PTWs taking more noticeable evasive action. Among mopeds, the figures reveal a low braking rate (23% in fatal accidents and 32% in personal injury accidents), as well as a low reaction rate (63% and 55% respectively). Similarly, avoidance is more common among heavy motorcycle users than among mopeds users. These figures show that reactions differ according to the PTW category. Moreover, where the reaction figures are identical, the underlying explanations are different. For example, when a moped user fails to react, in the majority of cases this is because the user has not spotted the hazard. However, a lack of reaction by a motorcycle user is generally due to a lack of space or time.

However, questions remain over the benefits of undertaking an emergency manoeuvre, since the severity of accidents in which the PTW made an emergency manoeuvre is higher than for those accidents where the user did not make such a manoeuvre. As well as questioning the benefits of such a manoeuvre, it is also important to address the success with which such a manoeuvre is made, as well as the initial driving conditions (potentially making such a manoeuvre ineffective). For PTWs, braking remains a key issue. The injury-related consequences of braking are twice as high as the consequences of avoidance, and are significantly higher than the consequences of braking among cars.

3.2. The PTW user

The analyses conducted have revealed the importance of the user's role and the user's faults in PTW accident rates. They have also revealed potential areas for improvement in road safety, concerning both prevention and traffic regulation enforcement.

• In terms of speed, the results show the benefits of reducing driving speeds among PTWs, and particularly among heavy motorcycle users. Both average speeds and speeding rates were analysed.

Fig. 2. Average speed of light vehicles and heavy motorcycles involved in fatal accidents (first picture) and in personal injury accidents (second picture), according to speed limit.

For both fatal accidents and personal injury accidents involving PTWs, the initial average speeds of heavy motorcycles are significantly higher than those of light vehicles. However, the extent of this difference varies between fatal accidents and personal injury accidents. Overall, the average difference in speed between heavy motorcycles and light vehicles involved in personal injury accidents stands at 10 km/h (figure 3). For fatal accidents (figure 2), this difference is significantly higher, standing at between 19 and 50 km/h according to the speed limit. The most striking figure concerns the average speeds of heavy motorcycles involved in fatal accidents on roads where the speed limit is 50 km/h, where the average speed is 30 km/h above the speed limit. Some 76 of the 83 accidents (92%) take place in urban areas. When compared with light vehicles, the speed of heavy motorcycles appears to be a major factor in PTW accident rates, with a particularly significant impact on severity. This observation may be confirmed by analysing average speeds, irrespective of the speed limit or type of road concerned.

Table 1. Comparison of average initial speeds of heavy motorcycles and light vehicles.

Average initial speed Light vehicles Heavy motorcycles Delta

Fatal accidents 72.9 km/h 93.3 km/h 20.4 km/h

Personal injury accidents 55.2 km/h 59 km/h 3.8 km/h

This tendency for higher speeds is also revealed in an analysis of speeding trends. Speeding is more common among heavy motorcycles (involved in 62% of fatal accidents) than among light vehicles (26%). Heavy motorcycles also tend to run over the speed limit by a higher margin. The fatal accident figures reveal that, in 17% of cases

where a heavy motorcycle was exceeding the speed limit, it was exceeding this limit by a significant margin (more than 50 km/h above the limit). This figure is almost twice the rate found among light vehicles.

These results show the benefits of reducing driving speeds and PTWs. The public authorities in France have recently decided to introduce stricter regulations concerning the size of PTW registration plates, to ensure that these types of vehicle cannot evade detection by automatic speed cameras at the roadside.

• In terms of crash helmets, the effectiveness of the fastening system has also been addressed.

Table 2. Rate of crash helmet-wearing after impact, where the crash helmet was worn and properly fastened prior to impact (n=662 users for fatal accidents and 444 users for personal injury accidents).

Crash helmet-wearing after impact Fatal accidents involving a PTW Personal injury accidents involving a PTW

Yes 87% 99%

No 13% 1%

Where the PTW wore a properly fastened crash helmet prior to the accident, the crash helmet remained in place after the impact in 87% of cases. This figure is higher than the results observed in older studies (in particular ACEM, 2004). This may be due to the effects of improved fastening system technologies in recent years. In personal injury accidents, the crash helmet remained on the user's head in 99% of cases, indicating the highly effective nature of the fastening system. However, where the crash helmet was not fastened prior to the accident, it did not remain on the user's head after the accident in almost all cases.

Table 3. Rate of crash helmet-wearing after impact, where the crash helmet was worn but not properly fastened prior to impact (n=53 users for fatal accidents and 17 users for personal injury accidents).

Crash helmet-wearing after impact Fatal accidents involving a PTW Personal injury accidents involving a PTW

Yes 4% 0%

No 96% 100%

• As well as the effectiveness of the fastening system, these figures also reveal the general limitations of the protection that the crash helmet provides - limitations that had already been revealed previously in ACEM (2004). Indeed, some 19% of fatal accidents involving a PTW are accidents in which the PTW user does not hit a vehicle or fixed obstacle at the roadside. In these cases, the user simply falls onto the road or verge. Among these 19% of cases (175 fatal accidents involving PTW users), almost 80% of users were wearing a crash helmet correctly but were nevertheless killed. This observation highlights the importance of wearing other safety equipment in addition to the crash helmet (back protection, airbag jacket, protective jacket, etc.). The use of this type of equipment was extremely limited in the accidents studied.

Table 4. Taux de port des équipements complémentaires (gants, blouson, bottes, pantalon, dorsale, équipement retro réfléchissant, airbag) au casque dans les accidents mortels (n=497).

Rate of use of other protective equipment in addition to the crash helmet PTWs Heavy motorcycles Light motorcycles Mopeds

No equipment worn 34.2% 17.4% 67.3% 74%

At least one item of equipment worn 65.8% 82.6% 32.7% 26%

Overall, around 66% of PTW users involved in fatal accidents wore at least one item of equipment. This rate was significantly higher among heavy motorcycle users (83%) than among light motorcycle and moped users (33% and 26% respectively). Furthermore, this rate was also higher among drivers (68%) than among passengers (38%). Of the 13 light motorcycle and moped passengers, none wore any item of equipment other than a crash helmet. This difference between the behaviour of passengers and drivers, and between heavy motorcycle users and light motorcycle and moped users, is also reflected in the crash helmet wearing rate. It is therefore necessary to raise

awareness about the importance of wearing additional protective equipment, and to promote examples of good behaviour among heavy motorcycle users to users of light motorcycles and mopeds.

3.3. What types of PTW are involved in accidents?

An analysis was conducted to reveal the main types of PTW involved in accidents. Each PTW model was allocated to a specific PTW type (roadster, racer, trail bikes, custom, scooter, etc.) then compared with all types of PTW on the road (based on the results of a national survey conducted by the Ministry of Transport).

This analysis revealed the particularly high rate of roadsters involved in accidents. A comparison between the accident data from the VOIESUR database and data from the national surveys on all types of PTW on the road (CGDD, 2013) revealed differing exposure rates according to the type of PTW concerned. In terms of heavy motorcycles, roadsters (figure 4, first picture) are particularly exposed. The raw figures show that this type of PTW is involved in more fatal and personal injury accidents than any other (almost half of all accidents), despite accounting for just 28% of all PTWs on the road. Other types of PTW, although less commonly represented in overall accident rate figures, nevertheless have concerning exposure rates. This is particularly true of racers (figure 4, second picture). Racer users account for 14% of all heavy motorcycle users involved in accidents.

By developing in-depth knowledge of the types of PTW involved in accidents, it is therefore possible to fine-tune awareness-raising initiatives targeting their users.

Fig. 3. Representation of heavily exposed PTW types - roadsters (first picture) and racers (second picture).

3.4. The role of infrastructure in PTW accidents

One of the main aims of our work was to address a topic often raised by motorcyclist associations in France, i.e. road surface maintenance failures and their impact on PTW accidents. Available data have shown that road surface maintenance problems are only rarely the cause of fatal or personal injury accidents. In fact, these elements (potholes, gravel, oil slicks, etc.) have been established as the root cause of an accident in fewer than 2% of cases. These figures are significantly lower than those quoted in the literature (particularly ACEM, 2004) and can reflect the particular situation in France, or an improvement of road maintenance conditions over the last 10 years, or a information not available in road traffic accident reports. It is important to note, however, that these two studies did not cover the same scope. While poor road surface maintenance has a limited impact on personal injury accident rates, it nevertheless has an impact on PTW user comfort and on material accidents - topics not covered in this

In order to provide national public authorities and local managers with the tools they need to improve the quality of France's road network, emphasis was also placed on the link between obstacles and PTW accidents.

This is a topic of particular concern for PTWs. According to ONISR (2006), PTWs accounted for 18% of all vehicles that collided with fixed obstacles in France in 2004. The most severe accidents also occurred among PTWs (14.3 deaths per hundred victims among motorcyclists, compared with 11 for car drivers, according to ONISR, 2012). In order to fine-tune the data available on obstacles, the statistics were initially broken down to reveal the number of PTW accidents involving fixed obstacles. For the purposes of this analysis, the term "fixed obstacle" relates to any item at the side of the road that, if impacted, may aggravate the consequences of a vehicle leaving the

road. Pedestrians and moving vehicles are excluded from the scope of this definition. The ground and the road surface itself are not considered obstacles, in line with the French technical literature on this subject (SETRA, 2002).

Fig. 4. Proportion of accidents involving obstacles, by accident type (n=946 fatal accidents and 22,780 personal injury accidents).

The analysis reveals that more than one third of fatal accidents involving a PTW, and more than 10% of personal injury accidents, are accidents involving a fixed obstacle. The involvement of fixed obstacles is naturally higher in single vehicle PTW accidents. Unsurprisingly, this phenomenon is less regularly observed in accidents involving a collision with another vehicle. It is important to note that, proportionally speaking, fatal accidents involving fixed obstacles are more common among cars than among PTWs. In fact, according to ONISR (2012), some 48% of fatalities in car accidents involved a collision with a fixed obstacle, compared with 36% of fatalities in motorcycle accidents. However, this initial analysis only provides a general overview of the situation, and these figures do not reveal whether the fixed obstacle was hit by the PTW vehicle or by the PTW user. Further in-depth analysis was therefore conducted to identify the exact nature of the collision with the obstacle (user, vehicle or both).

Fig. 5. Point hitting the obstacle in accidents involving a collision between a PTW and an obstacle (n=389 fatal accidents and 3,392 personal injury accidents).

In around half of cases, the user collided with the obstacle (either the user only, or the user at the same time as the PTW vehicle). This figure does not negate the need to combat the presence of obstacles at the roadside. Nevertheless, it provides a more accurate insight into the actual role played by obstacles in PTW accidents, especially given the tendency, until now, for national data to over-state this risk (by aggregating all types of obstacle impact, i.e. user, vehicle or both).

Finally, an analysis of the type of fixed obstacle involved shows that crash barriers are most often involved in fatal accidents involving PTWs. The figures also reveal that crash barriers have an extremely high severity index number of users killed/number of PTW users colliding with the obstacle), at the same level as lighting columns (i.e. 28).

The user's ability to recover after leaving the road is also an area where significant improvements could be made in terms of PTW safety. According to an analysis of the VOIESUR database, the figures show that the ability to recover, on the same infrastructure, differs significantly according to the type of user. Light vehicle drivers are six or seven times more likely to be able to recover and return to the road than PTW users after leaving the road and mounting the verge. In this case, our analysis did not show that speed or alcohol had an impact on this capability. In

our view, the extremely low recovery rate among PTWs is due in part to the fact that a high percentage of accidents involving PTWs occur on bends, and in part to the technical complexity of the recovery manoeuvres necessary to return a PTW to the road. Whatever the reason, this observation should provide infrastructure managers with important information when it comes to establishing their verge investment and maintenance policy.

4. Conclusions and outlook

The wealth of data in the VOIESUR database provides detailed information about PTW accident rates in France. The findings of our analysis of this database reveal the need to give greater consideration to PTWs in infrastructure design and maintenance. The results in terms of obstacles and recovery ability show that major safety flaws remain. Moreover, these flaws also affect other types of user.

As well as the role of infrastructure, the data reveal the important role of changes in the behaviour of PTW users. The topics addressed in this respect reveal the potential for major safety improvements, both in terms of primary safety (speed, alcohol impairment, inexperience, etc.) and in terms of secondary safety (wearing and fastening crash helmets, expanding the use of additional protective equipment). These observations therefore provide initial pointers that should help with the long-term development of a national road safety policy in the coming years, focusing in particular on both PTW users and other users involved in PTW accidents.

The study also confirmed the specific characteristics of PTWs and the need to adopt a differentiated approach to this type of user. As well as differences between PTWs and other types of vehicle, there are also differences in problems, driver profiles and behaviour between different categories of PTW (heavy motorcycles, light motorcycles and mopeds). As our observations have shown, there are specific problems for each PTW category:

• speed for heavy motorcycles

• alcohol, failure to wear protective equipment for mopeds, or law rate of emergency manoeuvres.

We have also identified potential avenues for future, complementary research work, focusing in particular on PTW conspicuity problems (lights not switched on, effectiveness of lighting, jacket, prominence of the PTW), and on matters such as PTW dynamics when travelling around a bend and the need for more research into the link between driver training and emergency manoeuvres.

References

ACEM (2004), 'Maids: In-depth investigations of accidents involving powered two wheelers'. Association des Constructeurs Européens de Motorcycles, Brussels, Belgium

CGDD (2013). Les deux-roues motorisés au 1er janvier 2012. Chiffres et statistiques [en ligne], mars 2013, n°400. Disponible à l'adresse :

http://www.developpement-durable.gouv.fr/IMG/pdf/CS400.pdf. CGDD (2013). Les deux-roues motorisés : à chaque âge, son usage et ses dangers. Le point sur [en ligne], mars 2013, n°156. Disponible à l'adresse :http://www.statistiques.developpementdurable.gouv.fr/fileadmin/documents/Produits_editoriaux/Publications/Le_Point_Sur/2013/l ps156-deux-roues-motorises.pdf ONISR (2006). Les grands thèmes de la sécurité routière en France. Paris: La Documentation française, 2006. ONISR (2012). La sécurité routière en France - Bilan de l'année 2011. Paris: La Documentation française, 2012

SETRA (2002). Traitement des obstacles latéraux sur les routes principales hors agglomération. Service d'Etudes sur les Transports, les Routes et

leurs Aménagements, Bagneux, France. Van Elslande, P., Fournier, J.Y., Vincensini, M., Roynard, M., Nussbaum, F., & Clabaux, N. (2008). Tache 3 : analyse comparative de procédures d'accidents mortels et non mortels. Projet ANR/Predit « 2RM » 2006-2008 : Accidentologie, Usage et Représentation des deux Deux-Roues Motorisés. INRETS-LAB, LAB-GIE PSA/Renault, Août 2008.