Working PapersArbeitsberichte

How Can Car2X-Communication Improve Road Safety

A Statistical Based Selection and Discussion of Feasible Scenarios

Sebastian Röglinger | Christian Facchi

How Can Car2X-CommunicationImprove Road Safety

A Statistical Based Selection andDiscussion of Feasible Scenarios

Sebastian Röglinger | Christian Facchi

Abstract

A lot of applications have been sketched inthe area of Car2X-Communication. Safetyaspects are often mentioned among othersas a motivation for Car2X-Communicationsystems. We present a statistical based se-lection of feasible applications with the goalof increasing safety. Therefore the officialstatistical information of the year 2007 hasbeen considered.Additionally we will present some conse-quences for the introduction of Car2X-Communication systems including infras-tructure.

ArbeitsberichteWorking Papers

Heft Nr. 15 aus der Reihe„Arbeitsberichte – Working Papers“

ISSN 1612-6483Ingolstadt, im April 2009

How Can Car2X-Communication Improve Road Safety 2 Feasible Scenarios for Car2X

1 Introduction

Car2X-Communication is a field of researchwith high interest and considerable potentialfor improving road safety. Because of that,there are several organizations like the C2C-CC[1] and projects like SIM-TD [2], AKTIV [3, 4],Cognitive Automobiles [5, 6] or SAFESPOT[7] devoted to this research direction. Topicsof these projects are real life tests, increasingtraffic safety and traffic flow, the introductionof cooperative traffic systems and more.A lot of different applications have been de-

fined. Due to [8] they can be categorized by:

• safety applications

• transport efficiency applications

• infotainment applications

In [9] it has been stated that a significant re-duction of killed persons by road accidents is adeclared European goal. Therefore, we focus inour research project on applications increasingroad safety.In this paper, we show feasible scenarios with

potential for increasing road safety by intro-ducing Car2X-Communication. These scenar-ios are selected by high crash rates or a highamount of fatalities. Furthermore, we discusssolutions in which a communication systembased on Car2X-Communication can achieve areduction of accidents.The German Federal Statistical Office publi-

cizes statistical information [10] concerning theroad crashes on german roads annually. We usethis statistics for our considerations because ofthe high quality, completeness and the actual-ity. For this paper, the publication based onthe year 2007 is used because the report for2008 is not yet available. Some of these statis-tics are combined with the latest results of thegerman traffic census of 2005 [11] for consider-ations of the different road usage.To identify the scenarios with high crash

rates we have taken accident victims and fatal-ities into account. Accident victims are those

persons who are injured or killed as conse-quence of a crash. So, fatalities are a subsetof accident victims.

The remainder of the paper is organized asfollows: In Section 2 different scenarios arepresented by statistics and are discussed inthe context of Car2X-Communication. Relatedwork is mentioned in Section 3. The papersconclusion and our future work are provided inSection 4. Finally the acknowledgment is givenin Section 5.

2 Feasible Scenarios forCar2X

mainroads

highway

urbannon-urbanurban

non-urban

mainroads

highway

fatalitiesaccidentvictims

Figure 1: Tree with Categories

Figure 1 shows the fundamental separationof the major categories for crash scenarioswhich are used in this document. All presentedstatistical information in this paper is based onthat tree to make the conclusions comparable.We used a comparison between killed per-

sons and the total amount of accident victimsbecause this allows us to figure out where thescenarios with high crash rates are located.Furthermore, an evaluation of the relevance ofthose scenarios for road safety is possible.

2.1 General Survey

In this section we give a general survey ofthe statistics on road accidents. Even thisleads to first relevant information for Car2X-Communication.

- 1 -

How Can Car2X-Communication Improve Road Safety 2.1 General Survey

2.1.1 Statistics

To get an first idea where accidents happenand which of those accidents are really danger-ous we present an abstraction of [10]. Table 1shows a comparison between the major cate-gories. The raw values (1st column) of thattable are qualitatively depicted in Figure 2. InTable 2 the percentage shares of fatalities in thetotal amount of accident victims is presented.

amou

nt(r

awva

lues

)

urba

n/no

n-ur

ban

with

/wi

thou

tHW

all

leav

es

fatalitiesfatalitiesfatalities 4949 100 % 100 %urban roads 1335 27 % 27 %non-urban 3614 73 % 100 %main roads 3012 83 % 61 %highway 602 17 % 12 %

victimsvictimsvictims 436368 100 % 100 %urban roads 278307 64 % 64 %non-urban 158061 36 % 100 %main roads 126119 80 % 29 %highway 31942 20 % 7 %

Table 1: Comparison between Categories

The first column ’amount’ of Table 1 includesthe raw values for each leaf of the tree pre-sented in Figure 1. For example there were436 368 people injured or killed by road acci-dents in 2007 in Germany. 4 949 of them werefatalities.

fatalities victims

urban roads

main roads

highway

27 %

61 %

12 %

64 %

29 %

7 %

Figure 2: Qualitative Comparison between Cate-gories

The second column of Table 1 includes thevalues for urban and non-urban roads. As itcould be expected, the values show a change ofrelevance between fatalities and accident vic-tims. On urban roads the rate of accident vic-tims is higher than on non-urban roads. How-

ever, the rate of fatalities on urban roads is lessthan on non-urban roads.

The third column of Table 1 compares mainroads (non-urban roads without highways)with highways (in tables sometimes abbrevi-ated to ’HW’). As result this column pointsout that highways are less dangerous than mainroads. Both aspects (fatalities and victims)show that with nearly the same distribution.Main roads and highways are in that point di-rectly comparable because on both road typesapproximately the same traffic volume takeplace which is shown later in this section (Ta-ble 3 and Figure 3).

The most interesting values of Table 1 arepresented in the last column which includes thevalues for all leaves (also depicted in Figure 2).Highways are once more in the lower range.Only 12% of the fatalities and only 7% ofthe accidents with victims happen on highways.Beside this, urban roads and main roads switchroughly their values by changing the point ofview from fatalities to accident victims. Thechange of relevance by the urban/non-urbancomparison (2nd column) is therefore almostindependent of highways.

Table 2 shows the percentage share of fatal-ities in the total amount of accident victims(see column 1 of Table 1). Despite the factthat most accidents with victims take place onurban roads, those roads take by far the leastshare of fatalities. On the second place high-ways can be found. On highways there are byfar the least shares of fatalities and accidentvictims in their respective total amounts butthe high traveling speed cause accidents witha higher mortality rate than on urban roads.The highest share of fatalities in accident vic-tims take main roads. The reason for thatcould be seen by the values Table 2 is calcu-lated from. On main roads the raw value foraccident victims is in the middle (lower than onurban roads) but the raw value for fatalities isthe greatest one (greater than on urban roads).

To characterize the situation on non-urbanroads more in detail, the results from the ger-man traffic census from 2005 [11] can be usedto take the different road usage into account.

- 2 -

How Can Car2X-Communication Improve Road Safety 2.2 Characteristics of Crashes

fatalitiesvictims

· 100 %

total 1.13 %

urban roads 0.48 %non-urban 2.29 %main roads 2.39 %highway 1.88 %

Table 2: Comparison between Killed and Total

This traffic census is updated every five years.In Table 3 and Figure 3 the road length and theannual mileage is presented. One of the resultsof [11] is that 53.0% of the non-urban traffictakes place on highways which only have a partof 11.9% in the non-urban road net. Interest-ing is that especially on highways occur only17% of the non-urban accidents with fatalitiesand only 20% of the non-urban accident withvictims.

road length annual mileagein in

103 km 109 km

highway 12.3 11.9 % 213.3 53.0 %main roads 90.9 88.1 % 189.5 47.0 %

103.2 100 % 402.8 100 %

Table 3: Comparison between road length and an-nual mileage

highway main roads

road length

annual mileage

Figure 3: Qualitative Comparison between roadlength and annual mileage

2.1.2 Relevance for Car2X

Based on the accident statistics [10] and on thetraffic census [11], the highest influence for in-creasing road safety could be reached by im-proving the situation on main roads. Main

roads show the highest rate of fatalities and theamount of accident victims is also quite high.Taking into account that only 47% of the non-urban traffic takes place on main roads, thefact that the amount of fatalities is five timeshigher than on highways is significant. For theintroduction of a Car2X-Communication sys-tem, it is therefore important that the full func-tionality of the system is guarantied especiallyon main roads. Useful systems for main roadscould be Cooperative Collision Warning, Elec-tronic Brake Lights, Glare Reduction, Coop-erative Adaptive Cruise Control and more asdescribed in [1].

Nearly the same importance take urbanroads. On urban roads occur about two thirdsof all accidents with victims. As to be ex-pected, the lower travel speed is the reasonfor the lower mortality rate of this accidents.A suitable system for increasing safety in thiscases could be a intersection assistance, e.g.the PReVENT subproject INTERSAFE [12]was working on.

As previously mentioned highways take onlya low share in accidents which could also beseen easily in Figure 2. Because of the highvelocity a high share in accidents ends up withfatalities. This is significant to any realizationof the Car2X-Communication, because a suit-able system should also be able to cope withhigh vehicle speed. Furthermore, the systemshould handle time critical messages in appro-priate time. To deal with all such requirements,the PHY- and the MAC-layer (OSI ReferenceModel for network stacks [13]) have to be de-veloped accordingly [14, 15].

2.2 Characteristics of Crashes

Another point of view for the definition of fea-sible scenarios to improve road safety providesthe breakdown of the different crash types. Inthis section we split up the different crash typesmore in detail.

- 3 -

How Can Car2X-Communication Improve Road Safety 2.2 Characteristics of Crashes

2.2.1 Statistics

The fatality rates derived from this drill-downreflects the different types of roads very well.So it is sufficient, only taking fatalities into ac-count. An abstraction of the accident statistics[10] is presented in Table 4 and Table 5.

urbannon-urban

wo. HW highway

crash between carsstart, stop,standing

2.8 % 0.6 % 3.0 %

ahead or waiting(rear end crash)

1.6 % 2.6 % 35 %

besides withsame direction

2.9 % 1.8 % 8.0 %

oppositedirection

8.2 % 33 % 3.0 %

turn or junction 20 % 14 % 0.5 %

othersrun-off-road(road departure)

18 % 39 % 40 %

accidents w.pedestrian

36 % 5.6 % 3.3 %

impact withobstacle

0.5 % 0.5 % 1.3 %

other 11 % 3.0 % 6.8 %

100 % 100 % 100 %

Table 4: Distribution of Crash Types

In Table 4 the distribution of crash types onthe special road types is given. E.g. 20% ofthe accidents on urban roads happen in turnsor junctions. The crash types presented in Ta-ble 4 emphasize some important cases whichare highlighted. Such cases are for urban roads’turn or junction’, ’run-off-road’ and ’accidentswith pedestrian’ which sum up to 74%. Onmain roads there are the cases ’opposite direc-tion’ and ’run-off-road’ which sum up to 72%and for highways there are ’ahead or waiting’and ’run-off-road’ relevant which sum up to75%. It could be seen that the row ’run-off-road’ is relevant on every road type.In Table 5 are crash characteristics and

specifics given. These attributes are secondaryconditions and can therefore not be summedup to 100%. As for the crash types in Ta-ble 4 there are also some outstanding cases

urbannon-urban

wo. HW highway

characteristicsintersection 16 % 6.9 %T-junction 17 % 12 % 2.3 %estateentrance

6.4 % 1.7 %

steep upwards 4.1 % 6.7 % 7.3 %steepdownwards

8.2 % 13 % 8.5 %

turn 16 % 40 % 18 %

specificsgrade crossing 4.6 % 0.8 %crosswalk 1.6 %crosswalk attraffic lights

5.6 % 0.2 %

bus stop 3.2 % 0.5 %working areas 0.9 % 0.4 % 3.8 %

Table 5: Distribution of Crash Characteristics andSpecifics

highlighted. These cases are for urban roads’intersection’, ’T-junction’ and ’turn’. For mainroads the high rated cases are ’T-junction’,’steep downward’ and ’turn’. Highways showonly for ’turn’ a significant value in respect ofcrash specifics. The case ’turn’ is emphasizedon every road type and takes therefore a specialrole.

2.2.2 Relevance for Car2X

The distribution of accidents depends on theroad type as shown in Section 2.2.1 for fatalitiesby crash types, characteristics and specifics.Because of that a Car2X-Communication sys-tem could improve its performance if it dynam-ically adapts to the road type the car is driv-ing on. So, the complexity of the whole sys-tem could be reduced to two or three majorcrash types depending on the road type withan at least 72% coverage of all accidents withfatalities. The crash type ’run-off-road’ andcrash characteristic ’turn’ show on every roadtype an increased significance. This leads tothe fact that a Car2X-Communication systemfor increasing road safety needs also some ba-sic functionalities like beaconing, forward col-lision warning, hazardous location notification

- 4 -

How Can Car2X-Communication Improve Road Safety 2.3 Driving Mistakes

or post crash warning [1]. However, it couldbe an advantage to use a different parameter-ization for the applications depending on theroad type. With such an situation based ad-justment the system could improve the over-allperformance of the safety applications.

To improve the situation on urban roads isa hard challenge for Car2X-Communication.Because of radio obstacles (e.g. buildings) incities, a two-dimensional routing is required[16, 17]. Furthermore, pedestrians and inter-sections are especially in cities relevant whichcan be seen in Table 4 and Table 5. Those situ-ations could not only be satisfied with Car2X-Communication. However, Car2X could playan important role in systems which combinethe special abilities of different techniques. Forexample, on intersections with traffic lightsthe Car2X system could exchange informationabout the switching timing [1, 18] betweencars and traffic lights. Furthermore, a trafficlight communication system could support therouting for emergency vehicles or detect haz-ardous situations with involved cars which arenot yet equipped with a Car2X system. Toprotect pedestrians, cars or infrastructure ele-ments which detects a hazardous situation be-tween a person and another car with its ownenvironment sensors (e.g. cameras, radar, li-dar, ...) could warn at least the involved car orthe surrounding environment.

On non-urban roads the crash types differfor main roads and highways but they have incommon that the increased values results fromcrashes with cars driving ahead or coming fromopposite direction. So, a system which ob-serves and interprets the traffic in front of thecar should be able to detect and so to preventsome crashes or at least weaken the impacts ofa crash. Once more, a Car2X-Communicationsystem on its own could not achieve a satisfyingsystem. But in combination with environmentsensors of cars and the infrastructure these sit-uations with increased crash values could beaddressed [19].

2.3 Driving Mistakes

In this section we take driving mistakes intoaccount because those are part in the rea-sons of almost all crashes. A future Car2X-Communication system has to address thisalso.

2.3.1 Statistics

Driving mistakes are among the main rea-sons for crashes in [10]. For accidents withvictims there were 351 186 drivers faults andfor accidents with fatalities there were 5 730drivers faults counted. Taking into accountthat there were in total 436 368 accident vic-tims and 4 949 fatalities, as presented in Ta-ble 1, the importance of addressing driving mis-takes is clearly shown. The numbers for driversfaults which cause victims or fatalities and thetotal amounts for victims or fatalities differ be-cause they are not correlated (e.g. the faultof one driver could cause one, two ore moredeath or two drivers faults could cause onlyone death).

The german crash statistics [10] splits theinformation about driver faults in several di-mensions. In this paper we selected the valuesfor the reasons which cause fatalities which arepresented in Table 6. Some cases like cargo se-curing and others have been omitted and someothers have been combined for clarity. The case’way of driving’ is a combination of speed, dis-tance, side by side driving and lane change.The case ’traffic coordination’ is a combinationof right of way, turn off and reverse driving.

The two columns including the distributionof drivers faults in Table 6 refer to the amountof the addressable cases. Every case shows amore or less relevant share in the total amountso none could be omitted. The most impor-tant case is ’way of driving’ followed by ’trafficcoordination’. The case ’wrong road usage’ ismostly import on non-urban roads in contrastto the case ’behavior towards pedestrian’ whichis mostly important on urban roads.

- 5 -

How Can Car2X-Communication Improve Road Safety 2.4 Time of Day

amount distribution

urbannon-

urbannon-

urban urban

wrong roadusage

58 516 6.0 % 18 %

way ofdriving

367 1561 38 % 54 %

overtaking 34 344 3.5 % 12 %trafficcoordination

294 440 31 % 15 %

behaviortowardspedestrian

208 54 22 % 1.9 %

961 2915 100 % 100 %

Table 6: Fatalities as Sequence of Misbehavior

2.3.2 Relevance for Car2X

Unfortunately, all the above mentioned majorcases differ to much to cope with them witha single and simple system. Because of thatthere has to be a combination of systems. Suchcombined systems should change its behaviordepending on the situation.In the following, some approaches to cover

the major cases are discussed. The ’usage ofwrong roads’ could be addressed by a navi-gation system including a road map in com-bination with a higher localization accuracy[20]. The higher localization accuracy could beachieved with a Car2X-Communication basedcollaboration. The benefits of that are the ac-curate detection of the road lane the car isdriving in or the positioning on a intersection.The case ’way of driving’ could not be inhib-ited but reduced by beaconing (periodic send-ing) the position, driving direction and vehicletype to other vehicles. With those informa-tion a system trying to interpret the drivingsituation could be significantly improved. Thecase ’overtaking’ might be addressable witha collaborative system between cars which isalso a possibility for the cases ’traffic coordina-tion’ and ’behavior towards pedestrian’. Sucha collaborative system could include a curvespeed warning, pre-crash sensing, cooperativeforward collision warning or lane-change warn-ing [1, 21].

2.4 Time of Day

The fatality rates and the traffic density differover the time of day. So, a comparison betweenboth point of views will provide some interest-ing facts a Car2X-Communication system willbe confronted with.

2.4.1 Statistics

Table 7 and Table 8 show the distribution ofthe fatalities over the time of day. Table 7 hasa higher resolution of time and is depicted inFigure 4. Table 8 includes additional informa-tion about the traffic density (also depicted inFigure 6) consisting of average values of roadsections which are equipped with counting de-vices. Data recordings for the traffic densityare only completely available on state roads(germ.: Bundesstraßen) and highways [11].

time amount distribution

0 - 1:59 213 4.30 %2 - 3:59 183 3.70 %4 - 5:59 298 6.02 %6 - 7:59 412 8.32 %8 - 9:59 331 6.69 %

10 - 11:59 439 8.87 %12 - 13:59 484 9.78 %14 - 15:59 652 13.17 %16 - 17:59 700 14.14 %18 - 19:59 536 10.83 %20 - 21:59 373 7.54 %22 - 23:59 328 6.63 %

4949 100 %

Table 7: Distribution of Fatalities over Time [10]

time

dist

ribut

ion

0-1:59

2-3:594-5:59

6-7:598-9:59

10-11:59

12-13:59

14-15:59

16-17:59

18-19:59

20-21:59

22-23:59

Figure 4: Fatalities over Time

In Table 7 and Figure 4 it could be seenthat the fatality rate increases from 2 am to

- 6 -

How Can Car2X-Communication Improve Road Safety 2.5 Age of Car

6 pm and decreases afterwards. There is onlyone little peak between 6:00 - 8:00 am whichprobably is caused by the rush-hour traffic.The comparison in Table 8 figures out that thedistributions for fatalities and traffic densityare not correlated. For example, during nighttime the fatality rates on highways are higher(18.4% in amount) than in the evenings (9.6%in amount) or by day (10.8% in amount). De-spite, the traffic density is at night much lowerthan in the evenings or by day. Furthermore,Figure 5 shows a normalization of the fatalitiescolumns of Table 8 to one hour (values dividedby the length of the time interval). This figureshows only a small decrease of fatalities perhour over time of day which not decreases likethe traffic density (see Table 8 and Figure 6).Highways cause even an increase of fatalities atnight despite an decrease.

time

fatalities traffic densityØ in vehicle/h

amount HWstate

HWroad

6 - 17:59 3018 327 694 299418 - 21:59 909 87 462 221622 - 5:59 1022 188 117 664

4949 602

Table 8: Fatalities and Traffic Density over Time[10, 11]

highwaysamount6 - 17:59

18 - 21:59

22 - 5:59

252

227

170

27.3

21.8

31.3

Figure 5: Normalized Fatalities per Year overTime

highways

6 - 17:59

18 - 21:59

22 - 5:59

694

462

117

2994

2216

664

state roads

Figure 6: Traffic Density over Time

2.4.2 Relevance for Car2X

The non-correlation between fatality rates andtraffic density is an important information forthe design of upcoming Car2X-Communicationsystems. This fact shows that it is necessary fora Car2X system to cope with an arbitrary num-ber of communication partners. So, e.g. thesesystems have to work during nighttime with asmall number of cars in the surrounding andduring rush-hours too. Also, if there will be ahigh market penetration in future there will besituations with no or just a few communicationpartners.

2.5 Age of Car

In this section we present statistical informa-tion about the distribution of the age of thecars involved in accidents. Such information isimportant to identify the ramp-up time whichis necessary for a new technology to show re-sults.

2.5.1 Statistics

Table 9 includes the age of the passenger carswhose drivers manly caused the accident [10].The information is presented for the accidentvictims and the fatalities. We use this informa-tion about the main originators and passengercars because of the lack of information aboutthe involved vehicles. So the numbers are notdirectly comparable to the others in that pa-per. However, we include these numbers be-cause they will contribute relevant informationfor Car2X-Communication systems. Addition-ally, it should be considered that the rows havedifferent time intervals. Figure 7 shows a qual-itative comparison between the distributionswhereas the area of each bar equates to thecorresponding value according to the time in-terval.In Figure 7, nearly an equipartition is re-

flected, especially for accident victims. Thedistributions for the fatalities shows a little in-crease over time with one peak between 8 and10 years. Both distributions exhibits lower val-ues for the first year.

- 7 -

How Can Car2X-Communication Improve Road Safety 3 Related Work

Age of Carvictims fatalities

in years

< 1 9240 741 - 2 18575 1282 - 3 17582 1313 - 4 16325 1334 - 5 16504 1415 - 6 16752 1616 - 8 35048 3298 - 10 41513 474

10 - 12 37980 382> 12 77630 956

unknown 24205 219

311354 3128

Table 9: Victims and Fatalities over Age of Car

<1

1-2 2-3 4-53-4 5-6 6-8 8-1010-12 age

dist

ribut

ion

accident victims

fatalities

<1

1-2 2-3 4-53-4 5-6 6-8 8-1010-12 age

dist

ribut

ion

Figure 7: Distribution of Age of Cars

2.5.2 Relevance for Car2X

The market introduction of a Car2X-Communication system will show effectson the crash statistics very slowly. Thereare two main reasons for that. At first, themarket introduction will be a slow processwhich will take several years. In [1] it is shownthat a market penetration of 50% wouldtake 6 years. This calculation is based onthe assumption that each new car would beequipped with a Car2X system. This exampleis not realistic but at least the upper boundof the time for an feasible market penetrationcould be estimated. At second, the accidents

are caused by cars at any age which is shown inSection 2.5.1. So, there will be hazardous sit-uations and accidents for a long time betweennewer and older cars which are equipped withand without Car2X-Communication systems.These two reasons in combination will slowdown the impact of Car2X-Communicationsystems on the crash rates.Therefore, the most important point for

the design of a realistic Car2X-Communicationsystem is a long term compatibility. It is ab-solutely necessary that the first equipped carsprovide benefits for the customer without ahigh market penetration. In addition, such carsmust be capable to operate with future versionsof the network.

3 Related Work

A lot of research is done or validated on high-way scenarios [22, 23]. This work is importantbecause on highways occur high traffic densities(see Table 3). Therefore, it must be ensuredthat Car2X-Communication systems work wellin such scenarios. But for safety research thepotentials of highways are limited as we haveshown.In [24] the german accident occurrence is

presented. This document shows an histori-cal analysis of the german and european acci-dents. Furthermore, the accident statistics arediscussed in the context of the different federalstates, place, month, reason, traffic contribu-tion and some special groups of persons. How-ever, this document shows only a general anal-ysis of the accident numbers and is thereforenot related to Car2X-Communication.

4 Conclusion and FutureWork

In this paper we analyzed several statisticalinformation and combined these to determinefeasible scenarios for improving road safety byan introduction of a Car2X-Communicationsystem.

- 8 -

How Can Car2X-Communication Improve Road Safety 5 Acknowledgment

The most important result is that accidentswith victims and fatalities show the lowestrates on highways. Because of that, highwaysare not so important for the development ofsafety applications. On main roads and incities occur more accidents which should there-fore be addressed by safety applications.By reviewing the distributions of the crash

types, characteristics and specifics of the differ-ent road types, we detected differences whichshould be considered:

• The crash type ’run-off-road’ and thecrash characteristic ’turn’ are relevant forany road type.

• Urban roads show additional relevance forintersections and accidents with pedestri-ans.

• Non-urban roads without highways showadditional relevance for crashes with carsdriving in the opposite direction, T-junctions and on road segments which aresteep downwards.

• Highways show additional relevance forcrashes with waiting cars or cars drivingahead.

• Driving mistakes are part in the reasons inalmost all accidents.

• The time of day of accidents and trafficdensities are not correlated.

To cope with these points a Car2X-Communication system needs some basicfunctionalities, some special road dependingfunctionalities and must be able to handlehigh and low traffic densities.The last point serves the identification of

ramp-time for market introduction. The dis-tribution of the accidents over the cars agesis quite balanced. So, effects for road safetywill take a long time. The slow market intro-duction and the occurring of accidents by lowtraffic densities show the necessity for applica-tions and systems which work without director many communication partners.In our future work, we intent to:

1. expand the statistical analysis to euro-pean scope to check whether the resultsare comparable,

2. conduct technical feasibility studies toidentify applications and communicationstrategies with the biggest impact on roadsafety and

3. put one of those feasible applications orcommunication strategies into practice.

5 Acknowledgment

For careful reading and providing valuablecomments on draft versions of this paper theauthors would like to thank Andreas Hermann,Marinus Luegmair and Peter Trapp. Fur-thermore, the authors acknowledge the finan-cial support for this research which has beengranted by the research program at Bavarianuniversities of applied sciences.

References

[1] “Car 2 Car Communication Consortium- Manifesto: Overview of the C2C-CCSystem.” http://www.car-to-car.org/,2007.

[2] “Sichere Intelligente Mobilität TestfeldDeutschland.” http://www.sim-td.de/.

[3] “Adaptive und kooperative Technologienfür den intelligenten Verkehr.” http://www.aktiv-online.org/.

[4] “Aktiv - experiencing the future together.”http://www.aktiv-online.org/.

[5] “Cognitive Automobiles.” http://www.kognimobil.org/.

[6] R. Nagel, S. Eichler, and J. Eberspacher,“Intelligent Wireless Communication forFuture Autonomous and Cognitive Au-tomobiles,” in IEEE Intelligent VehiclesSymposium, 2007.

- 9 -

How Can Car2X-Communication Improve Road Safety References

[7] “The SAFESPOT Integrated Project- Co-operative systems for roadsafety.” http://www.safespot-eu.org/documents/SF_Presentation_TRA2006_Goteborg.pdf.

[8] H. Hartenstein and K. P. Laberteaux, “ATutorial Survey on Vehicular Ad Hoc Net-works,” IEEE Communications Magazine,2008.

[9] “White Paper – European TransportPolicy for 2010: Time to Decide.” Com-mission of the European Communities,http://ec.europa.eu/transport/strategies/2001_white_paper_en.htm,2001.

[10] German Federal Statistical Office,Wießbaden, “Verkehr – Verkehrsunfälle2007.” http://www.destatis.de/, 2008.

[11] T. Kathmann, H. Ziegler, and B. Thomas,Straßenverkehrszählung 2005, Ergebnisse,Bericht der Bundesanstalt für Straßenwe-sen. Wirtschaftsverlag NW, 2007.

[12] “PReVENT - INTERSAFE.”http://www.prevent-ip.org/en/prevent_subprojects/intersection_safety/intersafe/.

[13] A. S. Tanenbaum, Computer Networks.Pearson Studium, 2003.

[14] K. Bilstrup, E. Uhlemann, and E. Ström,“Medium Access Control in Vehicular Net-works Based On the upcoming IEEE802.11 p Standard,” in World Congress onITS, 2008.

[15] Y. Wang, A. Ahmed, B. Krishnamachari,and K. Psounis, “IEEE 802.11 p Perfor-mance Evaluation and Protocol Enhance-ment,” in IEEE International Conferenceon Vehicular Electronics and Safety, 2008.

[16] C. Lochert, H. Hartenstein, J. Tian,H. Fussler, D. Hermann, and M. Mauve,“A Routing Strategy for Vehicular Ad HocNetworks in City Environments,” in IEEEIntelligent Vehicles Symposium, 2003.

[17] A. Schmitz and M. Wenig, “The Effectof the Radio Wave Propagation Model inMobile Ad Hoc Networks,” in ACM in-ternational Workshop on Modeling, Anal-ysis and Simulation of Wireless and Mo-bile Systems, 2006.

[18] A. Benmimoun, J. Chen, and T. Suzuki,“Design and Practical Evaluation of an In-tersection Assistant in Real World Tests,”in IEEE Intelligent Vehicles Symposium,2007.

[19] S. Wender and K. C. J. Dietmayer, “Ex-tending Onboard Sensor Information byWireless Communication,” in IEEE Intel-ligent Vehicles Symposium, 2007.

[20] A. Boukerche, H. A. B. F. Oliveira, E. F.Nakamura, and A. A. F. Loureiro, “Vehic-ular Ad Hoc Networks: A New Challengefor Localization-Based Systems,” Com-puter Communications, 2008.

[21] Vehicle Safety Communications Project,“Final Report, DOT HS 810 591,” April2006.

[22] C. P. Young, B. R. Chang, S. Y. Chen, andL. C. Wang, “A Highway Traffic Simulatorwith Dedicated Short Range Communica-tions Based Cooperative Collision Predic-tion and Warning Mechanism,” in IEEEIntelligent Vehicles Symposium, 2008.

[23] M. Torrent-Moreno, Inter-Vehicle Com-munications Achieving Safety in a Dis-tributed Wireless Environment – Chal-lenges, Systems and Protocols. Univ.-Verl.Karlsruhe, 2007.

[24] German Federal Statistical Of-fice, Wießbaden, “Unfallgeschehenim Straßenverkehr 2007.” http://www.destatis.de/, 2008.

- 10 -

Biography

Dipl.-Ing. (FH) Sebastian Röglinger isa research assistant at the Institute of Ap-plied Research of the Ingolstadt Universityof Applied Sciences. His research interestsare in the area of communication systemsand automotive technologies and there-fore for Car2X-Communication. SebastianRöglinger was born in 1983 at Kösching,Germany. For his graduate in ElectricalEngineering and Information Technology atthe Ingolstadt University of Applied Sci-ences he was elected as the winner of theE.ON Bayern culture award for 2008. Af-ter his study he was employed at Elektrobitin Erlangen, Germany as Software Engineerin the Engineering Service Infotainment de-partment and tested and coordinated thesoftware further development of the Audibluetooth car telephone.sebastian.roeglinger@fh-ingolstadt.de

Dr. Christian Facchi is Professor forSW Engineering, Distributed Systems andMathematics at the Ingolstadt Universityof Applied Sciences, Germany since 2004.His major research interests are in the areaof performance analysis and modeling, em-bedded systems, communication systemsand networks, testing of Software and for-mal description techniques. Dr. Fac-chi was born in 1964 at Munich, Ger-many. He holds a doctoral degree and adiploma degree from the Technische Uni-versität München, Germany both for Com-puter Science. Before he changed to the In-golstadt University of Applied Sciences hehas been employed at Siemens in the Mo-bile Phones development department for 9years. During his employment at SiemensMobile Phones he was the head of world-wide strategy for SW development environ-ments.christian.facchi@fh-ingolstadt.de

Working PapersArbeitsberichte

How Can Car2X-Communication Improve Road Safety

A Statistical Based Selection and Discussion of Feasible Scenarios

Sebastian Röglinger | Christian Facchi

Impressum

HerausgeberDer Präsidentder Hochschule für angewandteWissenschaften FH IngolstadtEsplanade 1085049 IngolstadtTelefon: 08 41 / 93 48 - 0Fax: 08 41 / 93 48 - 200E-Mail: info@fh-ingolstadt.de

DruckHausdruckDie Beiträge aus der Reihe „Arbeits-berichte – Working Papers“ erscheinen inunregelmäßigen Abständen. Alle Rechte,insbesondere das Recht der Vervielfälti-gung und Verbreitung sowie der Über-setzung vorbehalten. Nachdruck, auchauszugsweise, ist gegen Quellangabe gestat-tet, Belegexemplare erbeten.

InternetDieses Thema können Sie, ebsenso wie diefrüheren Veröffentlichungen aus der Reihe„Arbeitsberichte – Working Papers“, unterder Adresse www.fh-ingolstadt.de nach-lesen.

ISSN 1612-6483