Characteristics and Contributory Causes Related to Large Truck

Crashes

Sunanda Dissanayake, Ph.D., P.E.Associate Professor

Kansas State University

Disclaimer• The Contents of this report reflect the view of

the authors, who are responsible for the facts and the accuracy of the information presented herein. This document is disseminated under

the sponsorship of the Department of Transportation University Transportation

Center Program, in the interest of information exchange. The U.S. Government assumes no

liability for the contents or use thereof.

Outline

• Introduction

• Objectives

• Methodology

• Data

• Results

• Conclusions

Introduction

• One ninth of all traffic fatalities in US involved a large truck.

• However, large trucks accounted for only 3% of registered vehicles and 7% of vehicle miles traveled.

• Truck crashes tend to be more severe than other crashes.

• Important to identify characteristics and what leads to increased severities.

Trucks?

For the purpose of this study:Large trucks: Trucks with gross weight of 10,000 pounds or more.

Objectives

• To identify characteristics and contributory causes related to fatal truck crashes and all truck crashes.

• To compare circumstances more common in fatal truck crashes as compared to fatal non-truck crashes.

• To identify the factors that are contributing to/related with increased severity of truck crashes.

Methodology and Data

• Objectives achieved by analyzing crash data related to large trucks.

•Two phases of the study: –One focused on fatal truck crashes from the whole country–Second focused on all truck crashes from Kansas

•Statistical Modeling techniques used.

Analysis of Fatal Truck Crashes

• FARS database.

• Includes all police-reported fatal crash data from the whole country.

• Very detailed data with many coded variables.

• Fatality occurred within 30 days of the incident.

Question………

• How many fatal truck crashes in the United States each Year?

Analysis of Fatal Truck Crashes

0

1000

2000

3000

4000

5000

6000

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Year

No.

of

Fat

alit

ies

No. of fatalitiesin trucks

No. of fatalitiesin Non-trucks

Total

Particularly devastating to the occupants of the other vehicle.

Vehicle Occupants killed in Large Truck Crashes

• Sad fact: Majority of the damage is to the occupants of the other vehicle.

Source: DRIVECAM Website

Analysis of Fatal Truck Crashes

Based on manner of collision – fewer single vehicle truck crashes

Analysis of Fatal Truck Crashes –Bayesian method – crash related

Ex. Construction/work area LR = 2.77 Fatal truck crashes are 2.77 more likely in construction/work areas

Analysis of Fatal Truck Crashes –Bayesian method –vehicle related

Ex. Defective Brake Systems LR = 8.22 Fatal truck crashes are 8.22 times more likely to have defective brake systems

Analysis of Fatal Truck Crashes –Bayesian method-driver related

Ex. Following Improperly LR = 3.7 Fatal truck crashes are 3.7 more likely to have a driver that was following improperly.

Phase II – All Crashes• Data from Kansas• KDOT’s Kansas Accident Reporting System

database• Data related to truck crashes occurred

between 2004 and 2008 considered.• 18,919 observations.• Characteristics and Contributory causes

identified; severity modeling carried out.

BINARY LOGISTIC REGRESSION

• Variables were redefined in binary form; 1 or 0.• Variables checked for multicollinearity.• Binary Logistic Regression model developed• Dependent Variable crash severity was redefined as: = 1, if the occupants involved in the truck crash sustained injury of any severity level; = 0, otherwise• Sign of the variable important, Odds Ratio used to quantify the level of importance

RESULTS AND DISCUSSION

CHARACTERISTICS OF TRUCK CRASHES

Road Surface Type

Concrete (30.6%)

Blacktop (64.4%)

Gravel (3.1%)

Dirt (1.1%) Brick

(0.5%)

More truck crashes on Blacktop –makes sense !

Dry(79.2%)

Wet (10.3%)

Snow and Slush(3.5%)

Ice, Snow packed(6.2%)

Mud, Dirt or Sand (0.5%)

Straight and level

(67.3%)

Straight on Grade (19.3%)

Straight at hillcrest (1.7%)

Curved and Level

(5.5%)Curved on Grade

(5.3%)

More Truck Crashes under Dry road

surface condition

More Truck Crashes on straight and level

Road Surface Condition Road Surface Character

Lane Class

Two Lane Undivided (38.8%)

Four Lane Divided(35.3%)

Six Lane Di-vided (14.1%)

Four Lane Undivided

(9.3%)

Eight Lane Divided(2.3%) Two Lane Divided (0.2%)

2-lane undivided very critical

Light Condition

Daylight Dawn Dusk Dark-Street Lights on

Dark-No Street Lights

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0% 76.0%

2.3% 1.5%

7.6%12.5%

Light Condition

Per

cen

tage

of

Tru

ck C

rash

es

Majority under daylight conditions –exposure ?

Weather Condition Time of Day

No Adverse Condition (81.8%)

Rain, Mist or Drizzle (6.9%)

Snow(4.4%)

Strong Winds(1.7%)

Freezing Rain (1.0%) Snow and Winds

(1.4%)

00:0

1am to

3:00

am

3:01

am to

6:00

am

6:01

am to

9:00

am

9:01

am to

12:0

0noo

n

12:0

1pm to

3:00

pm

3:01

pm to

6:00

pm

6:01

pm to

9:00

pm

9:01

pm to

0000

hrs

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

3.6% 4.8%

15.3%

21.0%22.5%

18.8%

8.8%5.1%

Time of the Day

Perc

enta

ge o

f C

rashes

82% -Under no adverse weather

78% - 6 am to 6 pm

Age of the Truck Driver Gender of the Driver

16-20 21-40 41-60 61-80 >80 Others0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

9.3%

48.9%

31.5%

4.3%0.1%

5.7%

Age of the Truck Driver (Years)

Per

cen

tage

of

Cra

shes

Majority – middle aged Majority - male

Male Female Unknown0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%78.7%

16.9%

4.4%

Gender of the Truck Driver

Per

cen

tage

of

Tot

al T

ruck

Cra

shes

Truck Maneuver

Straight-follow-ing road

Right turn Left turn Backing Changing lanes Others0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%54.6%

8.0% 7.7%5.3% 5.3%

19.1%

Truck Maneuver

Per

cen

tage

of

Cra

shes

Distribution of Truck Crashes based on Truck Maneuver

Sing

le Veh

icle

Angle-

Side

Impa

ct

Rear E

nd

Side

swip

e (Sa

me Dire

ction

)

Backe

d int

o

Side

swip

e (Opp

osite

Dire

ction

)

Head O

n

Unkno

wn0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%35.2%

19.7%16.5% 15.3%

4.1% 3.3%1.4% 1.8%

Manner of Collision

Per

cen

tage

of

Cra

shes

Manner of Collision

More single vehicle crashes than in fatal truck crashes

Vehicle Body Type Accident Class

Automobile (56.0%)

Van(8.6%)

Pickup Truck (20.4%)

Sport Utility Vehicle (12.5%)

Others (1.6%)

Distribution of Two-Vehicle Truck Crashes Based on Body Type

Collision with other motor vehicle

(63.2%)Collision with fixed object

(13.3%)

Overturned(7.6%)

Collision with an-imal

(7.1%)

Collision with parked motor ve-

hicle(3.8%)

Other non-collision(3.6%) Other Collisions

(1.4%)

Distribution of Truck Crashes based on

Accident Class

Road Function Class

Rural p

rincip

le ar

terial

Urban

inter

state

Rural i

ntersta

te

Rural m

inor a

rteria

l

Urban

prin

ciple

arter

ial

Urban

free

way/ex

pressw

ay

Rural m

ajor c

ollecto

r

Urban

min

or arte

rial

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

25.1%

21.9%

15.8% 14.9%13.1%

5.8%

2.7%0.6%

Road Function Class

Per

cen

tag

e o

f C

rash

es

More in rural areas

Accident Location

Non In

terse

ction

Inter

secti

on

Inter

secti

on re

lated

Inter

chan

ge ar

ea

Roads

ide-in

cludin

g sho

ulder

Parking

lot-d

rivew

ay ac

cess

Med

ian0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%49.4%

16.6%12.4%

8.0% 6.9% 5.6%0.9%

Accident Location

Per

cen

tage

of

Cra

shes

50% - at non-intersection locations

Average Annual Daily Traffic

<10,000 10,001-20,000

20,001-30,000

30,001-40,000

40,001-50,000

50,001-60,000

60,001-70,000

>70,000 0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

70.00%63.74%

14.29%

6.98%4.95%

2.84% 3.33% 2.97%0.90%

Average Annual Daily Traffic (AADT)

Per

cen

tage

of

Cra

shes

Distribution of Truck Crashes based on Average Annual Daily Traffic

Contributory Causes of Truck Crashes

Question:• What is the most common type of contributory

cause?Driver related?Roadway? Environment related?Vehicle defects?Other?

Contributory Causes of Truck CrashesTruck Crashes Based on Type of Contributory Cause

Type of Contributory

Cause

Number of Truck Crashes

% of Crashes Involving

Contributory Cause

Driver Related 13,260 73%Environment Related

2,36013%

Road Related 1,409 7.8%

Vehicle Related 1,112 6.1%

Pedestrian Related 30 1.7%

Truck Crashes Based on Driver Contributory Cause

Driver Related Contributory Cause Number of Crashes Percentage of Crashes Involving Driver Related Cause

Failed to give time and attention 6,458 35.50%

Too fast for conditions 1,962 10.80%Failed to yield right of way 1,644 9.00%

Improper lane change 1,196 6.60%

Followed too closely 1,178 6.50%Made improper turn 1,016 5.60%

Disregard traffic signs, signal 770 4.20%

Avoidance or evasive action 742 4.10%Improper backing 726 4.00%

Improper passing 487 2.70%

Wrong side or wrong way 337 1.90%

Distraction in or on the truck 327 1.80%

Fell asleep 307 1.70%

Under influence of alcohol 250 1.40%

Reckless/careless driving 197 1.10%

Ill or medical condition 105 0.60%

Exceeded posted speed limit 101 0.60%

Did not comply with license restriction 91 0.50%

Improper or no signal 77 0.40%

Impeding traffic, too slow 76 0.40%

Under influence of drugs 66 0.40%

Aggressive, antagonistic driving 46 0.30%

Improper parking 46 0.30%

Vehicle Related Contributory CauseNumber of

CrashesPercentage of Crashes

Involving Vehicle Related Cause

Falling Cargo 389 34.0%Defective Tires 220 19.2%Defective Brake System 175 15.3%

Defective Wheel(s) 128 11.2%

Trailer-coupling related 85 7.4%

Other lights 48 4.2%

Unattended or driverless (not in motion) 41 3.6%

Unattended or driverless (in motion) 22 1.9%

Defective Windows-windshield 18 1.6%

Defective Exhaust System 12 1.0%

Headlights 5 0.4%

Truck Crashes Based on Vehicle Contributory Causes

Environment Related Contributory CauseNumber of

Crashes Percentage of Crashes Involving Environment

Related Cause

Animal Related 966 37.8%Rain, mist or drizzle 388 15.2%Falling snow 352 13.8%

Strong winds 336 13.2%Sleet, hail, freezing rain 185 7.2%

Vision obstruction - glare 93 3.6%

Vision obstruction - cultural 77 3.0%

Fog, smoke or smog 75 2.9%

Blowing sand, soil, dirt 39 1.5%

Vision obstruction - vegetation 26 1.0%

Reduced visibility due to cloud cover 17 0.7%

Truck Crashes Based on Environment-Related Contributory Causes

Truck Crashes Based on Road-related Contributory Causes

Photo credit: Iowa State University

Road Related Contributory CauseNumber of

CrashesPercentage of Crashes

Involving Road Related Cause

Icy or Slushy 686 45.7%Wet 281 18.7%Snow-packed 239 15.9%

Debris or Obstruction 113 7.5%

Road Under Construction/Maintenance 79 5.3%

Shoulders 69 4.6%

Ruts, Holes, Bumps 20 1.3%

Inoperative Traffic Control Device 14 0.9%

Truck Crashes Based on Road-related Contributory Cause

Binary Logistic Regression

Variable Mean Standard Deviation Description

ALCOHOL 0.01586 0.1249 =1 if the truck driver is under the influence of alcohol, =0 otherwise

BRAKES 0.03547 0.18496 =1 if the crash occurred due to brakes, exhaust, headlights, windows-windshield, cargo or tires, =0 otherwise

CARELESS 0.01813 0.13342 =1 if the truck driver is distracted or is too aggressive, =0 otherwise

CC_DR 0.69898 0.45871 =1 if the crash occurred has driver related contributory cause,

CC_ENV 0.12464 0.33032 =1 if the crash occurred has environment related contributory cause,

CC_RD 0.07448 0.26255 =1 if the crash occurred has road related contributory cause,

CC_VEH 0.0583 0.23432 =1 if the crash occurred has truck related contributory cause,

CLASS 0.63169 0.48236 =1 if the crash involves collision with a motor vehicle in transport, =0 otherwise

COLLISION 0.17929 0.38361 =1 if the crash involved a head-on collision, =0 otherwise

CONSTR_MAINT 0.05872 0.23511 =1 if crash occurred in construction, maintenance or utility zone,

CONTROL 0.81077 0.3917 =1 if the crash site has a traffic control device, =0 otherwise

DAMAGE 0.86432 0.34246 =1 if the truck had a damage, =0 otherwise

DAY 0.87774 0.32759 =1 if crash occurred during weekdays, =0 otherwise

DRUGS_ALCOHOL 0.01617 0.12615 =1 if the truck driver is influenced with drugs or alcohol, =0 otherwise

TRAPPED 0.0195 0.13829 =1 if truck driver was trapped, =0 otherwise

EVASIVE 0.0481 0.21398 =1 if the truck driver took evasive action or is too slow, =0 otherwise

GENDR 0.78699 0.40945 =1 if the driver of the truck was a male, =0 otherwise

IMP_MAN 0.1313 0.33773 =1 if the truck driver made improper maneuver, =0 otherwise

INOPERATIVE 0.00476 0.06881  =1 if the crash occurred at construction site or has inoperative traffic control device, =0 otherwise

LIGHT 0.75961 0.42733 =1 if the light condition is daylight, =0 otherwise

Example of Variables Considered in the Model

*- Significant at 0.05 level

Parameter Estimates and Odds Ratio of Large truck Crashes in the Model

Variable Estimate

Standard Error Pr > Chi-SqOdds Ratio

95% Wald Confidence

Limits For Odds Ratio

Intercept* -1.522 0.163 <0.0001    

ALCOHOL*

0.979 0.135 <0.0001 2.662.04,3.4

7CARELESS* 0.334 0.126 0.0078 1.40 1.09, 1.79CC_DR* 0.6 0.054 <0.0001 1.82 1.64, 2.02CC_RD* -0.332 0.084 <0.0001 0.72 0.61, 0.85CC_VEH -0.09 0.093 0.3329 0.91 0.76, 1.10CLASS 0.102 0.052 0.0509 1.11 1.00, 1.23COLLISION* 0.471 0.052 <0.0001 1.60 1.45, 1.77CONSTR_MAINT* -0.267 0.083 0.0013 0.77 0.65, 0.90CONTROL* 0.308 0.057 <0.0001 1.36 1.22, 1.52 DAMAGE* 1.116 0.083 <0.0001 3.05 2.60, 3.59DAY -0.003 0.058 0.9661 1.00 0.89, 1.12EVASIVE* 0.427 0.079 <0.0001 1.53 1.31, 1.79GENDR* -0.129 0.049 0.0079 0.88 0.80, 0.97IMP_MAN* -0.453 0.068 <0.0001 0.64 0.56, 0.73INOPERATIVE -0.247 0.328 0.4508 0.78 0.41, 1.48LIGHT 0.06 0.049 0.2209 1.06 0.96,1.17MANEUVER* 0.321 0.041 <0.0001 1.38 1.27, 1.49MIDDLE_AGED* 0.102 0.043 0.0166 1.11 1.02, 1.20OLD 0.092 0.14 0.5141 1.10 0.83, 1.44ONAT_TC* -0.521 0.054 <0.0001 0.60 0.53, 0.66

RAIN* 0.33 0.132 0.0124 1.391.07, 1.80

RUTS -0.148 0.224 0.5091 0.86 0.56, 1.34S_CHAR* -0.114 0.041 0.0051 0.89 0.82, 0.97S_COND* 0.256 0.056 <0.0001 1.29 1.16, 1.44S_TYPE* 0.132 0.04 0.0011 1.14 1.05, 1.24SAFETY_EQUIPT* -1.378 0.075 <0.0001 0.25 0.22, 0.29SMOG_SAND 0.355 0.218 0.1037 1.43 0.93, 2.19SNOW 0.151 0.099 0.1261 1.16 0.96, 1.41SPEED* 0.442 0.054 <0.0001 1.56 1.40, 1.73SPEED_LIMIT_1* -0.801 0.051 <0.0001 0.45 0.41, 0.50SPEED_LIMIT_2* -0.39 0.077 <0.0001 0.68 0.58, 0.79SPEED_LIMIT_3* 0.116 0.052 0.0252 1.12 1.01, 1.24TRAPPED* 4.417 0.344 <0.0001 82.81 42.21, 162.44UNATTND 0.483 0.329 0.142 1.62 0.85, 3.09VSN_OBSTRUCT* -1.326 0.132 <0.0001 0.27 0.21, 0.34WRONG 0.014 0.058 0.8034 1.01 0.91, 1.14

Examples

Findings and Conclusions• More than 80% of fatalities in large truck crashes

are occupants of the “other” vehicles.

• Relatively smaller % of single vehicle fatal truck crashes, as compared to SV fatal non-truck crashes.

• Many more…….

• Majority of all truck crashes in KS occurred on blacktops, in daylight conditions, under no adverse weather conditions.

• Contributory cause for most truck crashes- driver related. 73%

• Most common: Failing to give time and attention, driving too fast for conditions.

Findings and Conclusions

• Animal related causes and rain/mist/drizzle more common among environment related causes.

• Falling cargo, defective tires more common among vehicle related causes.

• Binary logistic regression provided a good means to identify the factors leading to increased severity of truck crashes.

• Odds Ratio shows the level of importance.• Highest odds ratio of 83 - when driver is trapped –

most likely to contribute to increased severity.• 2.7 times higher odds when driven by person

under the influence of alcohol. • Many more…• More focused and targeted countermeasure

ideas/programs developed based on the critical factors.

Slide design © 2009, Mid-America Transportation Center. All rights reserved.

MATCNishitha Bezwada and Siddhartha Kotikalapudi

Mr. Steven Buckley @ KDOTKDOT and NHTSA

Associate Director at KSU – Dr. Hossain

CREDITS

You can copy any of these graphics and paste them on other slides.