Identifying the Factors Contributing to Injury Severity in...

Research ArticleIdentifying the Factors Contributing toInjury Severity in Work Zone Rear-End Crashes

Kairan Zhang1 and Mohamed Hassan 2

1National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong UniversityChengdu Sichuan 610031 China2National United Engineering Laboratory of Integrated and Intelligent Transportation Southwest Jiaotong University China

Correspondence should be addressed to Mohamed Hassan abo hesenyyahoocom

Received 5 February 2019 Revised 31 March 2019 Accepted 15 April 2019 Published 2 May 2019

Academic Editor Jaeyoung Lee

Copyright copy 2019 Kairan Zhang and Mohamed Hassan This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

Egyptrsquos National Road Project is a large infrastructure project aiming to upgrade the existing network of 2500 kilometers as well asconstructing new roads of 4000 kilometers tomeet todayrsquos need Increasing highwaywork zones eventually direct the challenges fortraffic safety andmobility Realizing the need for mitigating the impact of such a challenging scenario this paper aims to investigateand identify the factors of work zone rear-end crash severity In this regard a random parameter ordered probit model was appliedto analyze data on the Egyptian long-term highway work zone projects during the period of 2010 to 2017 The factors of speedingand foggy weather conditions are found to be the key indicators for modeling the random parameters Besides during the weekendand at nighttime there is a higher risk of rear-end crash in work zones while heavy and passenger vehicles are at greater risk in thisregard It is anticipated that the findings of this study would facilitate transport agencies in developing effective measures to ensuresafe mobility across work zones

1 Introduction

Approximately 10466 people in Egypt died in road crashes inthe year 2013 Such a higher proportion of road traffic fatalityhighlights the critical nature of road safety in Egypt Youngand middle-aged individuals have been the most vulnerableage groups in this regard which eventually renders severeimpact on the society and on the emerging national economyof Egypt as well [1 2]

In theNational Roads Project more than 4000 kilometersof new roads are currently being constructed to strengthenthe Egyptian road network In addition another 2500 kilome-ters of existing road networks are reportedly being upgraded[3] which has directly led to an increase in the numberof work zones Consequently these work zones impede thetraffic and create conflicting situations for traffic flow andconstruction activities

Rear-end crashes are among the most common crashtypes on the highways and the most concerning scenariois the alarming number of the corresponding injuries and

fatalities For instance it has been showed that rear-endcollisions constitute 30 of all injuries and 297 of allproperty damage in the USA [4] Additionally it is alsoargued that rear-end crashes mostly occur on highway workzones rather than nonwork zones [5ndash9]

It is observed that Egypt significantly lacks in publisheddata that would define the severity of rear-end crashesand also the relationship of these crashes with work zonesAccordingly the academicians practitioners and the govern-ment agencies need to put collaborative efforts in this regardas identification and investigation of the critical factorscontributing to work zone rear-end crashes will facilitatedeveloping the appropriate and effective countermeasuresthat would serve the purpose of controlling the increasingrate of road safety issues across the highway

2 Literature Review

Having a good knowledge of the factors related to workzone rear-end crashes is essential to reassure effective and

HindawiJournal of Advanced TransportationVolume 2019 Article ID 4126102 9 pageshttpsdoiorg10115520194126102

2 Journal of Advanced Transportation

efficient work zone safety It is carried out by reviewing thefindings of previous studies focusing the influential factorsof the work zone and rear-end crashes with respect to theirseverity and frequency and also the statistical injury severitymodels used by the researchers were also studied in order todetermine an appropriate injury severity model The sectionbelow underpins the findings from past studies in this regard

In terms of work zone crashes Zhang et al developeda hybrid approach that combines a factor analysis methodand an ordered probit model to carry out a comprehensiveanalysis of work-zone crashes The results showed that thecrash type factor was significantly associated with work-zone severity [10] With a similar approach Osman et alused the ordered probit and logit models to identify factorscontributing to large truck crash injuries in work zonesand discovered that daytime speeding and rural areas wereassociated with more severe injuries [11] Also the orderedprobit model was developed by Ghasemzadeh and Ahmed toinvestigate the effect of weather on the severity of work zonecrashes The researchers concluded that weather and lightingconditions were the most important factors influencing crashseverity at work zones [12] In another study by Bharadwajet al driving behavior was found to be the most criticalrisk factor in work zone crashes [13] Wei et al investigatedwork zone crash severity under different light conditionsshowing that the combination of factors nighttime highspeed and driving under the influence or in poor lightingconditions leads to an increase in the injury rate of 727[14] On the other hand Sze and Song examined the levelof association between crash severity and common factors ofwork-zone-related crashes by applying a multinomial logisticregression model The authors concluded that the factorsof the vulnerability of road users heavy vehicles and thedaytime were significantly related to the severity of injuriesin work zone crashes [15] Long et al conducted a study toexamine the major factors contributing to work zone crashesBased on the study results the rear-end crash type was foundto be the most significant factor as it tends to intensifythe crash severities [16] According to Harb et al the roadtype age gender weather and lighting condition and drugsand alcohol involvement were substantial risk factors thatinfluence work-zone crashes [17]

The multiple logistic regressions were used by Yan etal to investigate risk factors of rear-end crashes on majorroads having signalized intersections The result identifiedseven environmental factors strongly associated with rear-end crash risk [18] In the research work ofWu et al speedingof the following and the leading vehicles differing amounts ofheadway and the density of the fog were significantly relatedto risks in rear-end collisions [19] Mohamed et al inferredthat there are seven variables causing substantial risk factorsthat influence rear-end crashes ie speed driver experienceroad type a number of lanes etc [20] On the same noteYan and Radwan concluded that rear-end crashes occurringat signalized intersections are associated with higher speedlimits daytime and wet and slippery road surface conditions[21] A study by Li et al was conducted to evaluate howexplanatory variables affect collision risks for three differenttypes of collisions at diverging freeways The researchers

indicated that the outcomes of rear-end crashes were moreserious than other types of collisions [22]

In terms of work zone rear-end crashes a plethoraof literature report the observation of an increased rateof rear-end crashes in work zones compared to nonworkzones [5ndash9] Qi et al investigated rear-end crashes in thework zones and utilized the ordered probit model Howeverthe research method is noted to have certain lacking likethe consideration of the driversrsquo gender and age vehiclecharacteristics and weather and lighting conditions [23]Silverstein et al conducted a study using the regressionmodel of Negative binomial (NB) and also the model ofmultinomial logit (MNL) to estimate how different factorscause fatal crashes in both the work and the nonwork zonesThe findings of work showed that rear-end crashes havea higher probability of causing death in work zones thannonwork zones [24] Likewise the comparison of work zonerear-end crash scenario for Singapore and Beijing revealedthat trucks were at a higher risk of suffering from rear-endcrashes in particular when the heavy vehicle is leading [25]Meng and Weng also suggested that the percentage of heavyvehicles influences the frequency of rear-end crashes in workzone [26]

In predicting the injury levels for collisions the orderedprobit model was utilized to examine the different factorsthat contribute to severe injury crashes [15 27 28] Abdel-Aty applied the same technique to discover the relationshipbetween the critical factors causing injury severity in crasheson different roadway sections The author stressed the signif-icance of this model to measure injury severity from crashessince it produced the best resultswhilemaintaining simplicity[16] Another study signified the efficacy of this model toinvestigate the severity of crashes [17] Similarly anotherresearch work also employed the ordered probit model toinvestigate the different risk factors and also the severity levelsof injury sustained in single and two-vehicle collisions [18]

The random parameter ordered probit model is a gen-eralization of the traditional ordered probit model allowingrandom regression coefficients thereby capturing effectscaused by differences in unobserved variables The randomerrors of the regression parameters are assumed distributedaccording to a priori distribution often chosen to be uniformtriangular or normal Predictions based on random param-eter ordered probit regression can be expected to be moreaccurate and statistically superiority than the results from thestandard model [19ndash23]

Research results based on data from western countriesmay not be directly applicable to developing countries suchas Egypt due to differences in roadway designs traffic charac-teristics and driver behaviorThereby the current study aimsto identify the factors that have a significant impact on theinjury severity of vehicle occupants that are involved in workzone rear-end crashes Moreover the impact of the identifiedfactors on injury severity based on available Egyptian trafficdata is also investigated by utilizing a random parameterordered probit model To the authorsrsquo knowledge no researchonwork zone rear-end crashes in Egypt has been published todate The present study is therefore an attempt to bridge thisknowledge gap

Journal of Advanced Transportation 3

3 Data Collection

In Egypt theMinistry of the Interior has a traffic departmentwhose key role is managing the database of national roadcrashes For crashes that occur on federal highways theMinistry of Transport regularly collects crash data The paperinvestigates work zone rear-end crashes that have occurredin 12 highway maintenance and rehabilitation long-termprojects (with a duration greater than one-year) during theperiod of 2010 to 2017 In this regard a total of 1045 crashreports were identified within the studied period Crashvariables extracted from the database were classified intosix categories including information of the driver vehicleinformation time of the crash characteristics of the roadwork-zone information and environmental conditions Sincethe level of injury is ordinal in nature the injury severityvariable was classified into three categorical levels includingno injury injury and fatal crashes In the current study theseverity of the crash was identified on the basis of the highestinjury severity sustained For instance in the case of onefatality (at least) it is termed as a fatal crash Similarly aninjury (one at least) resulting from a crash is classified as aninjury Closing off highways to traffic while maintenance andrehabilitation work is ongoing is very difficult Sometimeshalf of the road has to be open to traffic during workingon another half Since this situation is inevitable this paperhas taken into account the types of surface construction touncover which surface conditions contribute to the rear-end crashes in work zones In this regard the type ofsurface construction for each crash is divided into five cat-egories reflecting the situations of highway surfaces (AsphaltMilling Concert Removing Asphalt and Base) surface Thedescriptive statistics and frequency distribution of the factorsincluded in the analysis are reported in Table 1

4 Methodology

The random-parameter ordered probit model is especiallyappropriate for investigating how levels of injury dependon circumstantial factors The randomness of the param-eters provides compensation for unknown latent variablesaccounting for heterogeneity in the predictions of the fixed-parameter model In order to study the rear-end crash datawe apply the model

ylowast119894 = x1015840119894 lowast 120573119894 + 120598119894 (1)

120573119894 sim 119892 (120573 120579) (2)

where i = 1 2 n is the index of observations ylowast119894 isthe dependent variable for observation i xi is a vector ofcovariates 120573 are the mean parameter values and 120598119894 is an errorterm assumed to be distributed as a standard normal randomvariable 120598119894 sim N(0 1)

We note that when ylowast119894 is a binary variable and theparameters 120573119894 = 120573 are fixed (nonrandom) we have thetraditional probit model when ylowast119894 is an ordered variable with119868 categories and 120573119894 are fixed the model is an ordered probit

The probability density function for the ordered probit modelis

119891 (ylowast119894 | x119894120573119894)

=119868

prod119869=1

(Φ (120583119895 minus x1015840119894120573119894) minus Φ(120583119895minus1 minus x1015840119894120573119894))119910119894119895(3)

where 120583119895 are the threshold values for the ordinals In (2)the parameter vector 120573119894 is allowed to be different for eachobservation i so that the marginal effects on the dependentvariable differ in the sample The general assumption on theparameter vector is that it is drawn from some distribution119892(120573 120579) where the vector 120579 are the parameters of the aprioridetermined distribution often chosen as uniform triangualor normal We here assume that 120573119894 is normally distributedthat is 120573119894 sim 119873(120573 1205902119894 ) for each component 120573119894119896 in 120573119894which generalizes the model to a random-parameter orderedprobit model In case all 120590119896= 0 the model reduces to thefixed-parameter ordered probit The estimation of the fixed-parameter vectors in the ordered probit model is performedby likelihood maximization (ML) In the random parametercase it is necessary to resort to the simulated maximumlikelihood (SML) method

In the random-parameter ordered probit model we needto estimate the two parameter vectors 120573119894 and 120579 Since 120573119894is not observable we integrate out 120573119894 from the conditionaldistribution (2) to obtain

P119894 (120579) = int120573119894

P (y119894 | x119894120573119894) 119892 (120573119894) 119889120573119894 (4)

However (4) has no closed-form solution and so is solvedby Monte Carlo integration yielding an approximation P120484(120579)used as the factor in the maximum likelihood function Forany given parameter vector 120579 a sample value 120573119894119903 of theparameter vector is obtained in draw r from the assumed dis-tribution with density 119892(120573 120579) from which P120484(120579) is calculatedfor observation i using

P120484 (120579) = 1119877119877

sum119903=1

P120484 (y119894 | x119894120573119894119903) (5)

for a total number of samples R The simulated maximumlikelihood estimator 120579119878119872119871 is chosen as

120579119878119872119871 = arg max120579120598Θ

119873

sum119894=1

log P120484 (120579) (6)

It can be shown that the SML estimator is consistent andasymptotically normal under some regularity conditions Theperformance of simulated maximum likelihood is dependenton a large number of samples which can be very time-consuming In order to keep the number of draws reasonablylow the points are drawn from a Halton sequence which hasbetter coverage than pseudo-random number generators Inthe Simulated Maximum Likelihood 119877 = 200 Halton drawswere used which have been shown to give accurate parameterestimates [20 21]


Table 1 Summary of descriptive statistics

Variable Proportion Mean SDAt-Fault DriverYoung If driver is underlt 35 years = 1 otherwise = 0 700 070 0457Middle If driver is between 35 - 50 years = 1 otherwise = 0 252 025 0434Old If driver is above gt 50 years = 1 otherwise = 0 45 004 0207Male If driver is male= 1 otherwise = 0 909 091 0288Road GeometryCurve If crash occurred in curve section = 1 otherwise = 0 72 007 0258Straight If crash occurred in straight section = 1 otherwise = 0 430 043 0495U- Turn If crash occurred in U-turn section = 1 otherwise = 0 217 022 0413Grade If crash occurred in straight amp grade road = 1 otherwise = 0 281 028 0450Crash InformationHeavy vehicle If a heavy vehicle was involved = 1 otherwise = 0 461 046 049Passenger If a passenger car was involved = 1 otherwise = 0 749 075 043Environmental FactorsWeekdays If crash occurred on a weekday = 1 otherwise = 0 410 041 049Daytime If crash occurred during daylight = 1 otherwise = 0 476 048 050Fog If crash occurred under foggy weather condition = 1 otherwise = 0 347 035 047Winter If winter season (DecndashJanndashFeb)= 1 otherwise = 0 406 041 049Summer If summer season (JunndashJulndashAug)= 1 otherwise = 0 154 050 036Rain If crash occurred under rainy weather condition = 1 otherwise = 0 67 007 025Work zone InformationRural If crash occurred in rural area = 1 otherwise = 0 665 067 047Speeding If exceeded posted speed limits = 1 otherwise = 0 587 059 049N lane closures If more than one lane closures = 1 otherwise = 0 623 062 048Type of constructionAsphalt If surface construction is asphalt = 1 otherwise = 0 226 023 041Base If surface construction is base = 1 otherwise = 0 150 015 035Remove Asphalt If surface construction is removing asphalt = 1 otherwise = 0 304 030 046Milling If surface construction is milling= 1 otherwise = 0 102 010 030Concrete If surface construction is concrete = 1 otherwise = 0 217 022 041Crash SeverityNo injury 168Injury 596Fatal injury 236

The I categories are determined from the thresholds andthe probabilities of the ordered responses are given by thethresholds and the standard normal cumulative distributionfunction Φ as

119875119894 (119910 = 0) = Φ (minusx1015840120573)119875119894 (119910 = 1) = Φ (1205831 minus x1015840120573) minus Φ(minusx1015840120573)119875119894 (119910 = 119868) = 1 minus Φ(120583119868minus1 minus x1015840120573)

(7)

The marginal effects are computed as follows with thesample mean for each category 119895 as an argument

120597119875119894 (119910 = 119895)120597x = (120601 (120583119895minus2 minus x1015840120573) minus 120601 (120583119895minus1 minus x1015840120573))120573 (8)

where 120601 (sdot) is the probability density function of the standardnormal distribution

5 Results and Discussion

51 Model Specification Tests In this study the statisticalsoftware R with the package Rchoice was used for modelparameter estimation A total of 1045 observationswere takenfor the respective 25 independent variables Each explanatoryvariable of the data set was first tested for multicollinearityon the basis of the Variance Inflation Factor test (VIF) VIFbasically quantifies the change in variance or the extent ofcorrelation among the predictors in a model If the value ofVIF is in the range of 5-10 the predictors are affirmed to havea high correlation between them and if VIF value gt10 and


there seems to exist multicollinearity affecting the estimationof regression coefficients [24]

In the current study the VIF values were acquired in therange of 103-35 which informed that the explanatory vari-ables had no concerns regarding multicollinearity Accord-ingly the least significant variable was removed using theprocedure of backward eliminationTheprocedure continueduntil a final model was achieved Thus 13 variables from themodel were having statistical significancewith the confidenceinterval of 95 while lsquodaytimersquo was the only variable havingstatistical significance with a confidence interval of 90Afterwards the likelihood ratio test was used for testing thevalidity of the null hypothesis ie the fixed-parameter modelhas statistical equivalence to the random parameters modelThemethod is illustrated in the following section as adoptedfrom the study of Washington et al [25]

1198832 = minus2 [119871119865119894119909119890119889 (120573) minus 119871119877119886119899119889119900119898 (120573)] (9)

where 119871119865119894119909119890119889(120573) = log-likelihood convergence (For Fixedmodel) and 119871119877119886119899119889119900119898(120573) = log-likelihood convergence (ForRandom model) The resulting value of chi-square statisticie (X2 = 414) with two degrees of freedom and over 9999distribution confirmed the statistical significance and dom-inance of random-parameter model in comparison to thefixed-parameter model Besides the researchers have alsoused other methods for comparing the performance of thetwo models These methods included ldquoBayesian informationcriterion ndash BICrdquo ldquoAkaike information criterion ndash AICrdquo andldquoPseudominusR2 taking into account that lower values of AIC andBIC are good while a higher value of Pseudo-R2 indicate abetter model fit Accordingly the AIC and BIC values of therandommodelwere relatively lower than the fixedmodel andthe random model also acquired dominance over the fixedmodel with a relatively higher value of Pseudo-R2 (ie fixedmodel = 0203 and the random model = 0224) [26] Theseresults indicate that the random parameter model describesthe outcomes better than the fixed-parameter model

Followed by acquiring these empirical findings for therandom model the subsequent analysis is focused on theresults summarized in Table 2 Consequently Table 3 presentsthe marginal effects or the instantaneous rates of change forthe study variables ie the additional information about thecategories of injury severity the likeliness of occurrence andalso the change in the corresponding categories

52 Model Estimation Results The results of the modelindicate that under the normal distribution lsquoSpeedingrsquo andlsquoFoggy weatherrsquo variables were found to be best modeled asrandom having statistically significant standard deviations

Table 2 illustrates the mean and standard deviation valuesof these parameters

For the lsquoSpeedingrsquo parameter the normal distributionis confirmed from the mean value of 154 and 169 asthe measure of its standard deviation It informs that 18distribution is less than 0 It can be interpreted that 82 of thehigh-speed vehicles across the work zone tend to increase therate of fatal and injury crashes while 18 of vehicles passingthrough the work zone and involved in rear-end crashes with

high speed are less likely to sustain injury crashes The otherparameter of lsquofoggy weather conditionrsquo has secured the meanvalue of -098 with 104 value of standard deviation Thisindicates that 827 of rear-end crashes that occurred duringfoggy weather conditions result in a decrease in possibleinjury crashes while 173 of the crashes result in an increasein fatal and injury crashesThe following section discusses thefindings specific to the categorized study variables

521 At-Fault Driver The findings specific to the age groupreflect that young drivers tend to have a higher probabilityof encountering fatal impact of rear-end crashes An expla-nation might be that the young drivers tend to have lesserexperience of driving along with having high-speed drivingattitude which eventually leads them to experience the mostsevere crashes [29] Based on crash severity model by genderit is noted thatmale drivers travelling throughwork zone tendto be more involved in rear-end fatal crashes as compared tofemale driversThis finding is deemed to be reasonable for theunique attitude of male drivers to take more risks drive overthe speed-limit and drive more aggressively Accordingly anumber of previous studies [16 30 31] are in agreement withthese findings However some studies [32 33] have arguedthis aspect while claiming that similar circumstances tend toaffect female drivers more than male drivers

522HighwayGeometry Rear-end crashes have greatly beenassociated with road geometry The positive coefficient ofcurved sections implies that rear-end crashes occurring oncurved sections tend to render high injury severityThis resultis in line with the previous research findings [34 35] Inthe same context a considerable amount of disagreement isalso observed as the researchers have reported the horizontalcurves on roads were significant towards decreased injuryseverities across work zones than the nonwork zones [9 36]However the study by Katta found that the horizontal curveshave insignificant impact on injury severity [37]

523 Crash Information lsquoWeekendrsquo and lsquoNighttimersquo arefound to be the key factors related to rear-end crashes ashigher severity is observed for the occurrences across all workzones as compared to those occurred during the weekdaysand daytime It may be supported by the fact that workzones are inactive during weekends which motivates thedrivers to drive at high speed especially in nighttime since thework zones are expected to be usually not operational duringthe weekend As a result the probability of experiencing aneventful crash is higher in this case The changing drivingconditions at night with lower visibility and higher speedsmade possible by lighter traffic should be considered as anumber of factors jointly increasing the fatality rates andinjury severity on the Egyptian highways

On the other hand the drivers commuting daily onweekdays during daytime have a better visual impressionmore time to recognize work zones and to react accordinglyThese drivers tend to be more aware of the danger and arebetter prepared to slow down or take other measures toreduce the crash risk These findings are consistent with a


Table 2 Rear-end injury severity model results

Variables Fixed-parameters model Random-parameters modelCoefficient t-stat P-value Coefficient (standard deviation) t-stat P-value

Young 0566 6577 lt0001 0675 6156 lt0001Male 0361 2751 000595 0392 2344 00191Curve 1188 6567 lt0001 1529 7069 lt0001Heavy 0211 2717 0006 0288 3017 00025Passenger 0358 3559 lt0001 0296 2119 00340Weekday -0773 -7857 lt0001 -0844 -699 lt0001Daytime -0138 -1759 007849 -0183 -1897 00578Fog -0660 -6756 lt0001 -0988 (1047) -678 lt0001Summer 0303 2797 000515 0329 2544 00109Rural 0296 3262 00011 0480 414 lt0001Speeding 0993 7293 lt0001 1541 ( 1695) 6265 lt0001N lane closures 0540 4475 lt0001 0785 5143 lt0001Asphalt 0753 6079 lt0001 1126 6843 lt0001Milling -0321 -2079 0037 -0419 -2335 00195Threshold 11205831 0558 244 00146 0626 2128 00333Threshold 21205832 2135 29424 lt0001 2684 2095 lt0001Log-Likelihood at zero -9925 -9925Log-Likelihood at Convergence -7905 -7698Number of Observations 1045 1045AIC 1612918 157551BIC 1667153 1664642McFadden Pseudo R2 0203 0224

Table 3 Marginal effects associated with the random-parametersmodel

VariableMarginal Effects

(Random-parameters model)No injury Injury Fatal

Young -01710 01085 05048Male -00994 00631 02934Curve -03875 02458 11438Heavy -00751 00477 02218Passenger -00730 00463 02155Weekday 02139 -01357 -06314Daytime 00466 -00296 -01375Fog 02504 -01589 -07393Summer -00835 00530 02466Rural -01218 00773 03594Speeding -03904 02477 11525No laneclosures -01989 01262 05873

Asphalt -02854 01811 08424Milling 01062 -00674 -03135

number of previous researches [31 38ndash41] However Zhaoand Garber found no major differences between the day andnighttime crashes in the work zone [42]

Aside from the consideration of crash time the factorof lsquovehicular typersquo is also crucial with regards to work zonerear-end crashes The results show that the heavy and pas-senger vehicles are positively associated with injury severity

in rear-end crashes Meanwhile it is further realized thatthe involvement of passenger vehicle in a rear-end crashdirects increased severity of injuries for the drivers and theoccupants More specifically the outcomes of heavy vehiclesinvolvement in rear-end crashes are more fatal since suchcrashes lead to multiple-vehicle crashes as well It leads tosevere driverrsquos injuries and multiple fatalities at the workzones simply because of reduced braking system capability

The impact of heavy duty and passenger vehicles on workzone crash severity is found to be consistent with the findingsof several earlier studies [43ndash45] It can be explained on thebasis of the fact that Egyptrsquos trucks characterize large volumeand excessiveweight as there ismore than 96 transportationof good by trucks [2]

524 Environment Related The impact of foggy weatherconditions is found to be significant in terms of causing workzone rear-end crashes The result further indicates that thecrashes occurring during foggy weather are not as severe asduring other weather conditions It is interpreted based onthe fact that the reckless attitude of drivers is not notableduring adverse weather period as compared to clear anddry weather ie the drivers are intrinsically cautious whiletravelling in adverse weather conditions However somestudies report the contrary findings that foggy conditionincreases the rate of fatal crashes [31 36]

In addition to the impact of foggy weather condition thesummer season has also acquired a significant associationwith the injury severity causing an increase in the possibilityof the injuries These findings are supported by the factthat the number of vehicles on the road is higher in the


summer season that eventually makes the passenger vehiclesvulnerable to experience crash on the highways A numberof studies are found to be in agreement with this finding[46 47]

525 Work Zone Information Advanced information avail-ability to drivers ahead of work zones like the speed limitconstruction type and the number of lane closures arealso found to exhibit a significant relationship with theseverity of work zone crashes The respective variables areobserved to have statistically positive values of the coef-ficients that inform that having inadequate informationregarding work zones leads to increasing the likeliness ofrear-end fatal injury crashes In this context Green andSenders presented the foremost reason causing road trafficcrashes to be linked with the information flow Like whena driver does not perceive a critical situation earlier thedelay in driver perception-reaction time is a definite outcome[48]

In addition the positive coefficient value of the variableldquolane closuresrdquo reveals that the closure of many lanes willbe associated with an increased rate of occurrence for fatalcrashes A likely reason is forced merging of traffic whichcould be attempted late due to poor visibility thereby leadingto increased severity In particular heavy vehicles can bedifficult to maneuver in work zones if changing of lanes iscommenced late Consistent resultswere obtained in previousstudies [38 44 45 49 50]

Another key element of consideration is constructiontype ie during asphalt surface construction vehicle occu-pants are more likely to be involved in rear-end work zonefatal crashes High speeding has been known to be a strongcontributory factor for high frequency of crash severity Theanalysis results found a clear trend that the risk of rear-endcrash is increased if the speed limit is exceeded It can beexplained based on the fact that high-speed driving combinedwith the scenario of following other vehicles too close tendsto be the major contributory factor for work zone rear-endcrashes This finding is supported by a number of previousstudies [38 39 44 51 52]

Hence it can be asserted that the estimated coefficientvalues of the study variables facilitate determining the impactof the respective variables on the likeliness of the rear-endcrash In this regard it can be inferred that lsquospeedingrsquo factoraffects the injury level with the greatest coefficient for a rear-end crash that occurs at a work zone with 120573 = 154 whereaslsquoheavy vehiclersquo factor seems to render the lowest risk (120573 =0288)

Besides Table 3 also presents the marginal effects foreach factor with regards to the specific level of injury severityand provides additional information to confirm the previousfindings Comparing the results for driversrsquo gender if themale drivers are travelling across work zones a significantdecrease is observed for lsquoNo Injuryrsquo category (ie 99)whilethe probability of lsquoInjuryrsquo and lsquoFatal Injuryrsquo is increased by63 and 293 respectively In addition to this the resultsof road class are also comparable ie the likeliness of noinjury crashes is decreased by 121 and the probability

of injury and fatal crashes increased by 77 and 359respectively if rear-end crashes occurred in work zone ruralareas

6 Conclusion

The current study presents an analysis of injury severityof work zone rear-end crashes using a random-parameterordered probit model This approach allows for unob-served heterogeneity in the data typically for the fac-tors related to drivers vehicles and weather conditionsBased on Egyptian work zone crash database from 2010to 2017 the modeling procedure indicates that the factorslsquoSpeedingrsquo and lsquoFoggy weatherrsquo are best modeled as randomparameters

On investigating the impact of different factors onseverity of rear-end crashes this study reveals importantfindings passenger or heavy vehicles driving at high speedthrough work zones during nighttime with regards to theprospect of lane closures and also affirms that unexpectedmaneuvers tend to have a greater chance of being involvedin fatal rear-end crashes In addition the study also showsthat young male drivers who travel at nighttime duringthe weekends tend to suffer more from fatal injuries It isfollowed by another assertion that vehicle occupants aremorelikely to be involved in injury and fatal rear-end crashesin rural work zone area and horizontal curves In terms ofhighway construction the injury severity was higher duringasphalt surface construction than during milling surfaceconstruction

The finding from this analysis can be expected tofacilitate transport agencies in the development of effi-cient measures to ensure safety at work zones mitigaterisk factors and increase the traffic safety on Egyptrsquoshighways

Since most crashes can be attributed to human errorFor this reason it is recommended to develop effectivedriver training programmes to offer regular skills trainingto prelicensed and post licensed drivers specifically instill-ing the attitude of being a responsible driver across workzones

Furthermore ITS technologies such as variable speedlimit (VSL) and dynamic message signs (DMS) at an appro-priate distance ahead of the work zone can provide efficientmeans to provide drivers with updated information Provid-ing dynamic information can mitigate the risks of adversedesign human factors and roadway conditions as well asbalancing the traffic volumes between lanes thus reducinglane changing maneuvers

Other more conventional traffic engineering solutionsinclude increasing the upstream distance from the workzone where illumination or fluorescent devices conesand barrels are placed as well as enhancing visibilityand reducing the speed limit Also the use of flashinglights can be efficient in making drivers reduce theirspeed

The outcome of the current study is limited by availabledata which may affect results and their interpretations An


example hereof is that police reports do not include detailedcrash location within work zones (whether the incidentoccurred in the advancewarning area transition area activityarea or termination area) A limitation of the model in thisstudy is that some potentially crucial information (such astraffic volume number of vehicles) has not been consideredsince the data are missing from the database It is thereforeadvisable that more detailed information be collected andentered into the database so that it could be used for furthermodel calibration and a more detailed analysis of risk factorsin different crash scenarios

Data Availability

The data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

The authors thank the Ministry of Commerce of the PeoplesRepublic of China (MOFCOM) for funding of PhD doc-toral program at Southwest Jiaotong University Dr EssamRadwan from the University of Central Florida for valuablecomments and suggestions that have helped improve thisarticle

References

[1] WHOrganization ldquoGlobal Status Report onRoad Safety 2015rdquo2015

[2] A F Elshamly R A El-Hakim and H A Afify ldquoFactors affect-ing accidents risks among truck drivers in Egyptrdquo inProceedingsof the 6th International Conference on Transportation and TrafficEngineering ICTTE 2017 vol 124 p 04009 China July 2017

[3] Dailynewsegypt ldquoEgypt to add 4000 km to roads networkin 2016rdquo httpsdailynewsegyptcom20151108egypt-to-add-4000-km-to-roads-network-in-2016 2015

[4] S Singh ldquoDriver Attributes and Rear-end Crash InvolvementPropensityrdquo Tech Rep DOT-HS-809 540 National HighwayTraffic Safety Administration Washington DC USA 2003

[5] A J Khattak A J Khattak and F M Council ldquoEffects ofwork zone presence on injury and non-injury crashesrdquoAccidentAnalysis amp Prevention vol 34 no 1 pp 19ndash29 2002

[6] N M Rouphail Z S Yang and J Fazio ldquoComparative study ofshort- and long-term urban freeway work zonesrdquo Transporta-tion Research Record no 1163 pp 4ndash14 1988

[7] J Wang W Hughes F Council and J Paniati ldquoInvestigationof highway work zone crashes what we know and what wedonrsquot knowrdquo Transportation Research Record vol 1529 pp 54ndash62 1996

[8] H Yang K Ozbay O Ozturk and K Xie ldquoWork zone safetyanalysis and modeling a state-of-the-art reviewrdquo Traffic InjuryPrevention vol 16 no 4 pp 387ndash396 2015

[9] J Daniel K Dixon and D Jared ldquoAnalysis of fatal crashes ingeorgia work zonesrdquo Transportation Research Record vol 1715no 1 pp 18ndash23 2000

[10] K Zhang M Hassan M Yahaya and S Yang ldquoAnalysis ofwork-zone crashes using the ordered probit model with factoranalysis in Egyptrdquo Journal of Advanced Transportation vol 2018Article ID 8570207 10 pages 2018

[11] M Osman R Paleti S Mishra and M M Golias ldquoAnalysis ofinjury severity of large truck crashes in work zonesrdquo AccidentAnalysis amp Prevention vol 97 pp 261ndash273 2016

[12] A Ghasemzadeh and M M Ahmed ldquoExploring factors con-tributing to injury severity at work zones considering adverseweather conditionsrdquo IATSS Research 2018

[13] N Bharadwaj P Edara and C Sun ldquoRisk factors in work zonesafety events a naturalistic driving study analysisrdquo Transporta-tion Research Record vol 2673 no 1 pp 379ndash387 2019

[14] XWei X Shu B Huang E L Taylor and H Chen ldquoAnalyzingtraffic crash severity in work zones under different light condi-tionsrdquo Journal of Advanced Transportation vol 2017 Article ID5783696 10 pages 2017

[15] A J Khattak R J Schneider and F Targa ldquoRisk factors in largetruck rollovers and injury severity analysis of single-vehiclecollisionsrdquo Transportation Research Record vol 40 2003

[16] M Abdel-Aty ldquoAnalysis of driver injury severity levels atmultiple locations using ordered probit modelsrdquo Journal ofSafety Research vol 34 no 5 pp 597ndash603 2003

[17] M Abdel-Aty and J Keller ldquoExploring the overall and specificcrash severity levels at signalized intersectionsrdquo Accident Anal-ysis amp Prevention vol 37 no 3 pp 417ndash425 2005

[18] K M Kockelman and Y-J Kweon ldquoDriver injury severityan application of ordered probit modelsrdquo Accident Analysis ampPrevention vol 34 no 3 pp 313ndash321 2002

[19] E Dabbour M Haider and E Diaa ldquoScience direct usingrandom-parameter and fixed-parameter ordered models toexplore temporal stability in factors affecting driversrsquo injuryseverity in single-vehicle collisionsrdquo Journal of Traffic andTransportation Engineering (English Edition) vol 6 no 2 pp132ndash146 2018

[20] M Islam and S Hernandez ldquoLarge truck-involved crashesExploratory injury severity analysisrdquo Journal of TransportationEngineering vol 139 no 6 pp 596ndash604 2013

[21] S Islam S L Jones and D Dye ldquoComprehensive analysisof single- and multi-vehicle large truck at-fault crashes onrural and urban roadways in Alabamardquo Accident Analysis ampPrevention vol 67 pp 148ndash158 2014

[22] P C Anastasopoulos and F L Mannering ldquoAn empiricalassessment of fixed and random parameter logit models usingcrash- and non-crash-specific injury datardquo Accident Analysis ampPrevention vol 43 no 3 pp 1140ndash1147 2011

[23] F Chen and S Chen ldquoInjury severities of truck drivers insingle- andmulti-vehicle accidents on rural highwaysrdquoAccidentAnalysis amp Prevention vol 43 no 5 pp 1677ndash1688 2011

[24] D W Marquaridt ldquoGeneralized inverses ridge regressionbiased linear estimation and nonlinear estimationrdquoTechnomet-rics vol 12 no 3 pp 591ndash612 1970

[25] S P Washington M G Karlaftis and F L Mannering Statisti-cal and Econometric Methods for Transportation Data AnalysisChapman amp Hall Boca Raton Fla USA 2010

[26] Y Xiong J L Tobias and F LMannering ldquoThe analysis of vehi-cle crash injury-severity data A Markov switching approachwith road-segment heterogeneityrdquoTransportation Research PartB Methodological vol 67 pp 109ndash128 2014


[27] A Khattak and F Targa ldquoInjury severity and total harm intruck-involved work zone crashesrdquo Transportation ResearchRecord vol 1877 no 1 pp 106ndash116 2004

[28] A J Khattak M D Pawlovich R R Souleyrette and S LHallmark ldquoFactors related to more severe older driver trafficcrash injuriesrdquo Journal of Transportation Engineering vol 128no 3 pp 243ndash249 2002

[29] A Borowsky D Shinar and T Oron-Gilad ldquoAge skill andhazard perception in drivingrdquo Accident Analysis amp Preventionvol 42 no 4 pp 1240ndash1249 2010

[30] I M Dias Work Zone Crash Analysis and Modeling to IdentifyFactors Associated with Crash Severity and Frequency KansasState University 2015

[31] A Ahmadi A Jahangiri V Berardi and S G MachianildquoCrash severity analysis of rear-end crashes in California usingstatistical andmachine learning classificationmethodsrdquo Journalof Transportation Safety amp Security pp 1ndash25 2018

[32] J Weng G Du and L Ma ldquoDriver injury severity analysisfor two work zone typesrdquo Proceedings of the Institution of CivilEngineers - Transport vol 169 no 2 pp 97ndash106 2016

[33] M Abdel-Aty and H Abdelwahab ldquoModeling rear-end col-lisions including the role of driverrsquos visibility and light truckvehicles using a nested logit structurerdquo Accident Analysis ampPrevention vol 36 no 3 pp 447ndash456 2004

[34] Y Li and Y Bai ldquoHighway work zone risk factors and theirimpact on crash severityrdquo Journal of Transportation Engineeringvol 135 no 10 pp 694ndash701 2009

[35] Q Wang ldquoStudy on crash characteristics and injury severity atroadway work zonesrdquo Civil Engineering 2009

[36] R Harb E Radwan X Yan A Pande and M Abdel-AtyldquoFreeway work-zone crash analysis and risk identificationusing multiple and conditional logistic regressionrdquo Journal ofTransportation Engineering vol 134 no 5 pp 203ndash214 2008

[37] V Katta ldquoDevelopment of crash severity model for predictingrisk factors in work zones for Ohiordquo 2013

[38] Y Qi R Srinivasan H Teng and R Baker ldquoAnalysis of thefrequency and severity of rear-end crashes in work zonesrdquoTraffic Injury Prevention vol 14 no 1 pp 61ndash72 2013

[39] C Chen G Zhang R Tarefder J Ma H Wei and H GuanldquoA multinomial logit model-Bayesian network hybrid approachfor driver injury severity analyses in rear-end crashesrdquoAccidentAnalysis amp Prevention vol 80 pp 76ndash88 2015

[40] Q Yuan M Lu A Theofilatos and Y Li ldquoInvestigating driverinjury severity in rear-end crashes involving trucks in beijingareardquo Chinese Journal of Traumatology 2016

[41] M Osman R Paleti and S Mishra ldquoAnalysis of passenger-car crash injury severity in different work zone configurationsrdquoAccident Analysis amp Prevention vol 111 pp 161ndash172 2018

[42] N J Garber andM Zhao ldquoCrash characteristics at work zonesrdquoTech Rep VTRC 02-R12 2002

[43] J Weng S Xue Y Yang X Yan and X Qu ldquoIn-depth analysisof driversrsquo merging behavior and rear-end crash risks in workzonemerging areasrdquoAccident Analysis amp Prevention vol 77 pp51ndash61 2015

[44] J Weng Q Meng and X Yan ldquoAnalysis of work zone rear-end crash risk for different vehicle-following patternsrdquoAccidentAnalysis amp Prevention vol 72 pp 449ndash457 2014

[45] Q Meng and J Weng ldquoEvaluation of rear-end crash risk atwork zone using work zone traffic datardquo Accident Analysis ampPrevention vol 43 no 4 pp 1291ndash1300 2011

[46] F Naznin G Currie and D Logan ldquoExploring the impacts offactors contributing to tram-involved serious injury crashes onMelbourne tram routesrdquo Accident Analysis amp Prevention vol94 pp 238ndash244 2016

[47] N V Malyshkina and F L Mannering ldquoMarkov switchingmultinomial logit model An application to accident-injuryseveritiesrdquo Accident Analysis amp Prevention vol 41 no 4 pp829ndash838 2009

[48] MGreen and J SendersHumanError in RoadAccidents VisualExpert CRC Press Canada 2004

[49] B Wu H Xu and W Zhang ldquoIdentifying the cause andeffect factors of traffic safety at freeway work zone basedon DEMATEL modelrdquo in Proceedings of the Second Interna-tional Conference on Transportation Engineering pp 2183ndash2188Southwest Jiaotong University Chengdu China 2009

[50] S A Mohamed K Mohamed and H A Al-Harthi ldquoInvesti-gating factors affecting the occurrence and severity of rear-endcrashesrdquo Transportation Research Procedia vol 25 pp 2103ndash2112 2017

[51] Q Meng J Weng and X Qu ldquoA probabilistic quantitativerisk assessment model for the long-term work zone crashesrdquoAccident Analysis amp Prevention vol 42 no 6 pp 1866ndash18772010

[52] J Weng G Du and L Ma ldquoDriver injury severity analysisfor two work zone typesrdquo Proceedings of the Institution of CivilEngineers Transport vol 169 no 2 pp 97ndash106 2016

International Journal of

AerospaceEngineeringHindawiwwwhindawicom Volume 2018

RoboticsJournal of

Hindawiwwwhindawicom Volume 2018


Active and Passive Electronic Components

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwwwhindawicom

Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of



Journal ofEngineeringVolume 2018

SensorsJournal of



RotatingMachinery


Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia

Submit your manuscripts atwwwhindawicom


efficient work zone safety It is carried out by reviewing thefindings of previous studies focusing the influential factorsof the work zone and rear-end crashes with respect to theirseverity and frequency and also the statistical injury severitymodels used by the researchers were also studied in order todetermine an appropriate injury severity model The sectionbelow underpins the findings from past studies in this regard

In terms of work zone crashes Zhang et al developeda hybrid approach that combines a factor analysis methodand an ordered probit model to carry out a comprehensiveanalysis of work-zone crashes The results showed that thecrash type factor was significantly associated with work-zone severity [10] With a similar approach Osman et alused the ordered probit and logit models to identify factorscontributing to large truck crash injuries in work zonesand discovered that daytime speeding and rural areas wereassociated with more severe injuries [11] Also the orderedprobit model was developed by Ghasemzadeh and Ahmed toinvestigate the effect of weather on the severity of work zonecrashes The researchers concluded that weather and lightingconditions were the most important factors influencing crashseverity at work zones [12] In another study by Bharadwajet al driving behavior was found to be the most criticalrisk factor in work zone crashes [13] Wei et al investigatedwork zone crash severity under different light conditionsshowing that the combination of factors nighttime highspeed and driving under the influence or in poor lightingconditions leads to an increase in the injury rate of 727[14] On the other hand Sze and Song examined the levelof association between crash severity and common factors ofwork-zone-related crashes by applying a multinomial logisticregression model The authors concluded that the factorsof the vulnerability of road users heavy vehicles and thedaytime were significantly related to the severity of injuriesin work zone crashes [15] Long et al conducted a study toexamine the major factors contributing to work zone crashesBased on the study results the rear-end crash type was foundto be the most significant factor as it tends to intensifythe crash severities [16] According to Harb et al the roadtype age gender weather and lighting condition and drugsand alcohol involvement were substantial risk factors thatinfluence work-zone crashes [17]

The multiple logistic regressions were used by Yan etal to investigate risk factors of rear-end crashes on majorroads having signalized intersections The result identifiedseven environmental factors strongly associated with rear-end crash risk [18] In the research work ofWu et al speedingof the following and the leading vehicles differing amounts ofheadway and the density of the fog were significantly relatedto risks in rear-end collisions [19] Mohamed et al inferredthat there are seven variables causing substantial risk factorsthat influence rear-end crashes ie speed driver experienceroad type a number of lanes etc [20] On the same noteYan and Radwan concluded that rear-end crashes occurringat signalized intersections are associated with higher speedlimits daytime and wet and slippery road surface conditions[21] A study by Li et al was conducted to evaluate howexplanatory variables affect collision risks for three differenttypes of collisions at diverging freeways The researchers

indicated that the outcomes of rear-end crashes were moreserious than other types of collisions [22]

In terms of work zone rear-end crashes a plethoraof literature report the observation of an increased rateof rear-end crashes in work zones compared to nonworkzones [5ndash9] Qi et al investigated rear-end crashes in thework zones and utilized the ordered probit model Howeverthe research method is noted to have certain lacking likethe consideration of the driversrsquo gender and age vehiclecharacteristics and weather and lighting conditions [23]Silverstein et al conducted a study using the regressionmodel of Negative binomial (NB) and also the model ofmultinomial logit (MNL) to estimate how different factorscause fatal crashes in both the work and the nonwork zonesThe findings of work showed that rear-end crashes havea higher probability of causing death in work zones thannonwork zones [24] Likewise the comparison of work zonerear-end crash scenario for Singapore and Beijing revealedthat trucks were at a higher risk of suffering from rear-endcrashes in particular when the heavy vehicle is leading [25]Meng and Weng also suggested that the percentage of heavyvehicles influences the frequency of rear-end crashes in workzone [26]

In predicting the injury levels for collisions the orderedprobit model was utilized to examine the different factorsthat contribute to severe injury crashes [15 27 28] Abdel-Aty applied the same technique to discover the relationshipbetween the critical factors causing injury severity in crasheson different roadway sections The author stressed the signif-icance of this model to measure injury severity from crashessince it produced the best resultswhilemaintaining simplicity[16] Another study signified the efficacy of this model toinvestigate the severity of crashes [17] Similarly anotherresearch work also employed the ordered probit model toinvestigate the different risk factors and also the severity levelsof injury sustained in single and two-vehicle collisions [18]

The random parameter ordered probit model is a gen-eralization of the traditional ordered probit model allowingrandom regression coefficients thereby capturing effectscaused by differences in unobserved variables The randomerrors of the regression parameters are assumed distributedaccording to a priori distribution often chosen to be uniformtriangular or normal Predictions based on random param-eter ordered probit regression can be expected to be moreaccurate and statistically superiority than the results from thestandard model [19ndash23]

Research results based on data from western countriesmay not be directly applicable to developing countries suchas Egypt due to differences in roadway designs traffic charac-teristics and driver behaviorThereby the current study aimsto identify the factors that have a significant impact on theinjury severity of vehicle occupants that are involved in workzone rear-end crashes Moreover the impact of the identifiedfactors on injury severity based on available Egyptian trafficdata is also investigated by utilizing a random parameterordered probit model To the authorsrsquo knowledge no researchonwork zone rear-end crashes in Egypt has been published todate The present study is therefore an attempt to bridge thisknowledge gap


3 Data Collection


4 Methodology


ylowast119894 = x1015840119894 lowast 120573119894 + 120598119894 (1)

120573119894 sim 119892 (120573 120579) (2)




119891 (ylowast119894 | x119894120573119894)

=119868

prod119869=1




P119894 (120579) = int120573119894

P (y119894 | x119894120573119894) 119892 (120573119894) 119889120573119894 (4)


P120484 (120579) = 1119877119877

sum119903=1

P120484 (y119894 | x119894120573119894119903) (5)


120579119878119872119871 = arg max120579120598Θ

119873

sum119894=1

log P120484 (120579) (6)







(7)









1198832 = minus2 [119871119865119894119909119890119889 (120573) minus 119871119877119886119899119889119900119898 (120573)] (9)



















Asphalt -02854 01811 08424Milling 01062 -00674 -03135















6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



3 Data Collection


4 Methodology


ylowast119894 = x1015840119894 lowast 120573119894 + 120598119894 (1)

120573119894 sim 119892 (120573 120579) (2)




119891 (ylowast119894 | x119894120573119894)

=119868

prod119869=1




P119894 (120579) = int120573119894

P (y119894 | x119894120573119894) 119892 (120573119894) 119889120573119894 (4)


P120484 (120579) = 1119877119877

sum119903=1

P120484 (y119894 | x119894120573119894119903) (5)


120579119878119872119871 = arg max120579120598Θ

119873

sum119894=1

log P120484 (120579) (6)







(7)









1198832 = minus2 [119871119865119894119909119890119889 (120573) minus 119871119877119886119899119889119900119898 (120573)] (9)



















Asphalt -02854 01811 08424Milling 01062 -00674 -03135















6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia







(7)









1198832 = minus2 [119871119865119894119909119890119889 (120573) minus 119871119877119886119899119889119900119898 (120573)] (9)



















Asphalt -02854 01811 08424Milling 01062 -00674 -03135















6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





1198832 = minus2 [119871119865119894119909119890119889 (120573) minus 119871119877119886119899119889119900119898 (120573)] (9)



















Asphalt -02854 01811 08424Milling 01062 -00674 -03135















6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia










Asphalt -02854 01811 08424Milling 01062 -00674 -03135















6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia










6 Conclusion










Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




Data Availability




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


Identifying the Factors Contributing to Injury Severity in...

Documents

Transcript of Identifying the Factors Contributing to Injury Severity in...