ROAD TRAFFIC SAFETY

LECTURE 3: Crash Data

Dr. Madhumita Paul
Assistant Professor Department of Civil
Engineering IIT Kharagpur
ACCIDENT DATA

🞭 Different types of accident?

🞭 Why?
🞭 What are the sources?
🞭 India
🞭 Other countries
🞭 What is accident form?
🞭 Is it uniform?
ACCIDENT DATA

🞭 Who will fill it?

🞭 Minor accidents
🞭 Information included?
Accident data

 Prerequisite for each efficient road safety improvement

system
 Identification and definition of the relevant problem
together with knowledge of the data and parameters
describing this problem are essential for its successful
solution
 Comprehensive, up-to-date, accident data is needed for
recognition of the scope of road safety problems and for
raising public awareness

4
Road Safety Improvement System

Road Accident Investigation (RAI) by Road Engineers

 Requirement
 A certain amount of accidents and the accompanying data
according to the road engineer’s given task
 Example: Draft a preference list of improvement interventions or
black spot treatments
 Purpose
 Help road engineers detect the amount of road
infrastructure deficiencies that influence an accident’s
occurrence
 Guide road engineers in the implementation of appropriate
improvement measures

5
Road Safety Improvement System

Road Safety Inspection (RSI)

 RAI & RSI
 Both deal with existing roads
 RSI procedure
 Doesn’t require any data input
 Investigators need expert qualifications and experience in road
safety
 Investigators’ evaluation of “risk features” of roads and their
environs
 An identification of the hazardous situation
 Also draws attention to locations that need more detailed
investigation based on accident data analysis (RAI)

6
7
Road Safety Improvement System

 RAI reflects real occurrence and distribution of accidents and their

circumstances
 Has a high potential for identification of tailor-made safety measures and their
implementation.
 Quality and level of details of the relevant data are the prerequisite for the
application of appropriate countermeasures.

8
Application of Road Safety Analysis

a 500m
stretch
e.g. NH

Brannolte, U., Munch, A. (2009) Software-Based Road Safety Analysis in Germany, Paper
presented at the 4th IRTAD Conference, Seoul, Korea, September 17, 2009.

9
10
11
Data parameters and their quality

 Uniformity (to apply standard definitions)

 Vital importance for comparisons
 Even on the national level it is important that the local and
regional definitions comply with national ones
 There are different databases that often exist within one
country
 Data management
 Police
 Road administration
 Hospitals/health system
 Insurance companies

12
Any other aspect?

 Underreporting
 Takes place in a great amount
 Not all accidents are reported

 Intentionally unreported accident

 Issues
 Each database (police, hospital, road administration,
insurance company) has its own requirements on what and
how to report
 Only a careful comparison of the different sources can give a “true”
picture.

13
Data, data and data..

 It is quite evident that such complex monitoring of

accidents and traffic systems and their maintenance is very
expensive.
 Difficult for many countries to develop and operate such
systems to provide road engineers and decision-makers with
all the relevant and necessary information.
 Nevertheless, even a minimal amount of information can offer the road
engineer the ability to identify safety deficiencies in the road environment
and to design possible countermeasures
 Three levels of data sets are considered:
 Minimum data
 Road and traffic data
 Additional data

14
Minimum data
 Provide road engineers with relevant information necessary for basic
accident causation investigation.
 Example:
 Accident identification (a unique number-based system);
 Time (the date, hour, minute, and day of the week);
 Location;
 Accident type;
 Vehicles involved (number, type);
 Accident consequences (fatalities within 24 hours/30days,
injuries, material damage).
 Can be easily introduced in countries without any accident recording
system as an early step for a system-based reporting system.
 Doesn’t require huge financial resources, and only limited
human input is needed.
15
Road and traffic data

 Provide road engineers with relevant road infrastructure

information linked with the location of the accident and other
circumstances and factors contributing to the accident
occurrence.
 Road description (type of intersection, road number, road
category, cross-section, etc.);
 Specific places/objects (pedestrian crossing, rail crossing, bridge,
tunnel, bus/tram stop, parking place, petrol station, etc.);
 Road alignment (evident deficiency or not, slope, narrowing, etc.);
 Road surface (type, permanent state, actual conditions – e.g.
Snowy, wet, icy surface);
 Traffic control (traffic lights, road signs, policeman);
 Position of accident (travel direction of involved participants,
location - traffic lane, shoulder, roadside, etc.);
 Main causes of accidents (speeding, overtaking, right of way, etc.)

16
Additional data

 This set of information contains features related to the

vehicles and persons involved in the accident.
 Driver (category of license, driver experience, sex, age, nationality, education,
etc.);
 Impairment of the driver (alcohol, drugs, others, etc.);
 Condition of the driver (alert, tired, impulsive, sudden indisposition, suicidal,
etc.);
 Use of restraint devices (helmet, safety belt, child seat, etc.);
 Condition of the pedestrian (alert, impaired by alcohol/drugs, etc.);
 Behaviour of the pedestrian (proper, faulty, poor estimation of vehicle
movement, sudden entry to the road, etc.);
 License plate number;
 Brand make of vehicle;
 Vehicle operator (private, commercial, public transport, etc.);
 Year of production of the vehicle;
 Emergency service involvement.
17
Location information

 An accident blackspot is a term used in road safety

management to denote a place where accidents are
concentrated.
 Without a precise localization of road accidents, road
administrators cannot find and effectively treat the accident
blackspots on their road network.
 Inaccurate localizations mean misguided identifications
and result in the loss of financial means and time.
 The effective evaluation of implemented countermeasures
can also be influenced.

18
19
20
21
22
23
24
Other examples of crash forms
Crash report form – FRONT SIDE

26
Crash report form – BACK SIDE

27
STATS19 form
28
29
Filled out UD-10
crash report form
– FRONT SIDE

30
Filled out UD-10
crash report form
– BACK SIDE

31
32
33
https://www.michigantrafficcrashfacts.org/

34
35
Concerns

 Location error
 From nearest police station
 At Junction
 From center of intersection
 From edge of road
 Crash type error
 Side swipe as Right Angle
 Head on right turn as Right Angle

36
Crash Reporting Accuracy Issues

 Police reported data may not be accurate

 Errors in location
 Street name may be reported incorrectly or misspelt
 Incorrect estimate of offset distance or direction
 Some distances were measured from the centerline, while others
were measured from the curb
 Errors in type
 Officers often code crash based on actual impact angle
 Initial and intended final direction of involved vehicles are
more accurate indicators
 Improper typology leads to incorrect assessment as to cause of
crash
 Errors affect determination of probable cause of crashes and
subsequent countermeasure design

37
38
39
40
Crash Location Miscoding Example

 Suppose a police officer codes the location of a right-angle

crash as follows:
 Name of Road: First Street (E-W)
 At Intersection with: Main Street (N-S)
 10 m south on First Street
 Vehicle # 1 direction – SB
 Vehicle #2 direction – WB
 The crash may appear as shown in following figure based
on whether the distance was estimated from curb line or
centerline of intersection

41
Crash Type Miscoding Example

 Police officers often code sideswipe crashes as right angle

crashes
 Because the vehicles collide at a SLIGHT angle
 RIGHT ANGLE is reserved for vehicles originally traveling at
approximately right angles to each other
 Two vehicles traveling in the same direction do not collide at
a right angle

CORRECT CODING FOR TWO EB VEH. INCORRECT CODING

42
UD-10 Traffic Crash Report
Instruction Manual
Revised 2018

43
44
45
Germany

46
New Zealand

47
48
Road Accident Data Management System
(RADMS)
 Tamil Nadu Road Sector Project
 TN Action plan for Road safety
 Sponsored by World Bank
 $500,000
 Improved data collection process & reliable safety
information database
 GIS/GPS based

49
50
52
53
 What is this?
 Figure?
 Diagram?
 Map?
• Collision diagram
– Graphical Representation
of crashes occurring at a
location
– Assist in Determining
Potential Causes of
crashes and
Countermeasures

Which info missing?

54
55
56
57
58
Safe Stopping Distance

If vehicle position is less than

xc = cannot stop (rear-end
conflict/collision otherwise)
59
 Minimum Safe clearing distance

If vehicle position is more

than xo = cannot clear
intersection (right-angle
collision/conflict otherwise)
60
61
TYPE-I Dilemma Zone (neither pass nor go)

When Xc > Xo i.e., clearing distance is less

Min. Clearing distance

62
TYPE-II DZ

63
Utility of collision diagrams

 Collision diagrams are a very illustrative tool for the

comparison of the accident frequency before and after the
implementation of a particular road safety measure.
 The minimum length of the “before-and-after” period
considered in any ‘before/ after’ analysis should be at least
3 years.
 WHY?
 Short: Ensure responsiveness to change in crash patterns
 Long: Ensure maximum reliability
 Thumb rule ~ previous 3 to 5 years data
 With 3 years as a working minimum

64
65
66
What is this diagram?

Condition Diagram
Road Safety Improvement System @Delhi, India

 Shantivan Marg, Delhi –A 4-legged signalized intersection

 Coordinate : 28°39'02.8"N 77°14'42.2"E

68
69
70
71