ACCIDENT BLACKSPOT ANALYSIS ALONG DIGOS – TORIL NATIONAL HIGHWAY

REVIEW OF LITERATURE

Black Spot Criteria

Road accident black spot is a stretch of approximately 500 m on the National

Highway that has recorded five road accidents in the past three years. However,

hazardous locations are evaluated using Accidents Severity index (ASI) method

according to National Highway Authority of India (NHAI). Hazardous spots with

Accidents Severity Index (ASI) more than Threshold value (Average Severity +

1.5*Standard Deviation) is called Accident Blackspot. Based on NHAI's criteria for

estimation of ASI, the weightage to fatal accident will be assigned as 7 and to grievous

injury accident as 3.

The selection of suitable criteria varies from country to another. Contrary to

Developed and emerging countries face financial challenges when it comes to

improving road safety. That means treating all identified black spot is impracticable for

developing countries. An approach that can promote not only the identification but also

the prioritization of black spot must therefore be established.A parameter with the task

of identifying and prioritizing black spots is required for such an approach (Nguyen et

al., 2015).
 The Department of Public Works and Highways on “Current effort by the Public

Sector that addresses the road safety problem” a blackspot is a place where three

major accidents occurred in the past three years.

In 2015, the WSI or Weighted Severity Index approach was usedto rate the

accident location. As per WSI value from the collected data, the top five spots were

selected as accident blackspot. Some potential alternative steps to strengthen the

assessment and treatment of accident blackspots has been found productive if

adequate data is available (Vivek & Saini, 2015).

In 2016, in Kerala state, three years of accident data, details of accidents such as

date, time, location, etc. The severity of the accident, type of accident and the reason

for the accident recorded in First Information Report (FIR) of major accident spots from

police department. Accident Density, Weighted Severity Index (WSI), Severity Index and

Method of Ranking is the methods used for identifying and prioritizing accident

blackspots are (Binu&Anusha, 2016).

In 2017, a study on accident blackspot identification was conducted employing

methods such as Accident Density and Weighted Severity Index in analyzing 2

complete years of crash data in India (Dhule et al., 2017). Similarly, the identification

and analysis of accident blackspots help to find stretches where there are more injuries

and these spots usually decrease road safety. The spot where traffic accident occurs

regularly is called accident blackspots (Mohan &Landge, 2017).

In 2018, a study situated in Baddi, India was conducted in a period of six years,

using the accident data from the nearest police station. There are various methods in

identifying accident blackspots; these are Accident severity index of the location,

Weighted Severity index of the location, and Critical Crash Rate (Chetna et al., 2018).
Accident Modelling

In 2010, using logistic regression technique, a study in Bali, Indonesia

investigates the influence of risk factors on road accident fatalities. Considering all

vehicle type, the development of Logistic Regression models were separated for fatal

accidents. Based on the accident data from State Police of Bali Province seven
predictor variables were employed in the developed models The results of the study

shows that odds of fatal accident due to male motorcyclist and motorist at fault were 0.3

and 0.4 respectively lower than females. Therefore the odds of male motorcyclist and

motorists contributing more fatal motorcycle and motor vehicle accidents were around

79 percent and 72 percent respectively. In addition, age was also significant that

influences all vehicle fatalities (Wedagama&Dissanayake, 2010).

In 2011, a paper presents thestudy of accident blackspot, road crash analysis

and the development of accident predictive models based on the available crash data at

rural roadway, Federal Route 50 (F050) Malaysia. To link the discrete crash data to the

road and traffic flow explanatory variable, amultiple non-linear regression approach was

used. The result shows that the potential contributor to the increase in accident rates on

multiple rural roadways are the number of major access points, without traffic lights, the

increase in speed, the growth on number of Annual Average Daily Traffic (AADT), the

growing number of motorcycle and motorcars and the reduction of time gap (Mustakim&

Fujita, 2011).

In 2012, a study conducted In the Philippines aims to define important personal

and environmental variables in predicting motorcycle accident, compare the findings

with the results in other countries, and suggests potential government intervention.

Using a survey at a licensing centre in the largest city in Metro Manila, data was
obtained from 177 participants. To estimate the probability of an accident from variables

considered in the model, logistic regression was used. Major predictors of motorcycle

accident were found to be three variables: age, driving activity, and form of junctions. In

collisions, younger motorists are more likely to involve. In particular, driving conduct,

committing violation predicts the probability of an accident. Motor accidents are also

expected by driving at T and Y junctions. A unique set of variables has been used in the

Philippines to predict motorcycle. While the influence of these variables on the risk of an

accident was identified by previous research, the combination was unexpected.

(Seva et al., 2012).

In 2014, in Spain, the ordered probit model is used in this study to analyze the

impact of the variety of variables on the degree of injury faced by the occupants of

motor vehicle involves in road accidents. Those travelling in light vehicle, on a two-way

road and on the surface of a dry road appear to sustain more serious injuries than those

travelling in heavy vehicle, on one-way road and on the surface of a wet road. Also, the

driver’s seat, by comparison is obviously the safest seat, and urban areas, despite

having the highest incidence of occurrence of injuries, area associated with a reduce

level of severity. Women also tend to be more likely than men to suffer severe or fatal

injuries (Garrido et al., 2014).

In 2016, Centered on two-year New Mexico crash reports, an analysis uses

hierarchical ordered logit model to analyze the important factors in predicting driver

injury in rural non-interstate crashes. Model show that in the model fit and parameter

estimation, the model used in this analysis outperform the ordinary logit model. In rural

non-interstate accidents, variables about accident characteristics, environmental

factors, and driver and vehicle characteristic have been found to have substantial

impact on the prediction of driver injury severity. Factors such as road segments far

from intersection, wet road surface condition, animal accident, heavy vehicle drivers,
male driver and driver seatbelt used tend to induce less serious driver injury outcomes

than factors such as multiple vehicle accidents, severe crash damage to cars,

motorcyclist, women, senior drivers, alcohol or drug impairment drivers, and major cold

impairment (Chen, C. et al., 2016).

In 2017, a research using real-time traffic and weather data obtain from urban

arterials in Athens, Greece, on the risk of accident and severity. For the purposes of

preliminary research, Random-Forest (RF) is used. In particular, the goal is to identify

the variables according to their relevant significance and to provide a first insight into
the possible significant variable. Next, Bayesian logistic regression and logit models for

finite mixture and mixed are used to analyse variables related to the probability of

accident and severity, respectively. The Bayesian logistic regression showed that traffic

disparities substantially lead to the frequency of incidents in relation to the probability of

an accident. Although international literature notes that the variation in traffic are rising

in magnitude the analysis of the severity of the accident reveals a generally mixed

contribution of the differences in traffic to the severity of the accident (Theofilatos,

2017).