SENSITIVITY ANALYSIS OF TRAFFIC ACCIDENTS CAUSES IN
SULTANATE OF OMAN
Ahmad M. Abu Abdo
Department of Civil and Infrastructure Engineering, American University of Ras Al Khaimah, Ras Al Khaimah, United Arab Emirates E-Mail: [email protected]
ABSTRACT
Traffic accidents numbers and fatalities and injuries resulting from these accidents are increasing in an alarming rate, especially in the Sultanate of Oman. This paper focuses on the verification of the factors affecting numbers of traffic accidents and fatalities and injuries resulting from traffic accidents in the Sultanate of Oman. To achieve the goal of this study, sensitivity analyses via Tornado Plots and Extreme Tail Analysis were deployed to identify main affecting factors, which would assist decision makers to focus their efforts on the reducing traffic accidents. Results have shown that Failure to Follow Traffic Rules, Speeding, and Reckless Driving were ranked the highest contribution factors to traffic accidents and fatalities and injuries from these accidents, while Weather Condition, Drunk Driving, Fatigue, and Sudden Stopping had very low effect on traffic accidents. Finally, recommendations were proposed to Omani decision makers for implementation in current and/or future policies.
Keywords: traffic accidents, sensitivity analysis, tornado plot, extreme tail analysis, sultanate of Oman.
INTRODUCTION
Traffic Accidents and fatalities and injuries resulting from these accidents are considered one of the main concerns worldwide, especially their direct and severe impacts on any society, socially and economically.Records show that fatalities resulting from traffic accidents in the developing countries are very high [1]. Many studies were conducted to evaluate the impact of these accidents and argued that proactive law enforcement strategies and polices, public awareness campaigns play an important role in reducing traffic accidents [2-9].
Numbers of fatalities due to traffic accidents are raising in an alarming rate. Predictions suggested that by the year 2030, Traffic accidents fatalities would be the 5th top cause for death globally and would be doubled in the Sultanate of Oman. Sultanate of Omanis located on south eastern coast of the Arabian Peninsula with an area of
309,501 km2 and low population of 3,632,000. Furthermore, studies argued that by 2020 traffic accidents fatalities would be the 3rd leading cause of death and disability in the Sultanate of Oman. Reports showed that among the Arabian Gulf Co-operation Council (GCC) countries, Sultanate of Oman has the highest death rate from road accidents and the 3rd highest in the Middle East [10-13].
Records showed that traffic violationsin the Sultanate of Oman have increased in recent years [14], especially in speeding violations, which is considered to be the big factor for the increase of reported traffic accidents. In 2013, there were 7829 vehicle accidents caused by speeding, 913 reported fatalities and 10802 injuries due to traffic accidents (Figure 2). Failing to buckle up offences increased from 54,429 to 85,708 since 2009. Furthermore, right of way, dangerous overtaking,
reckless driving, cutting a red light, and usage of mobile phones while driving violations increased by 155%, 295%, 174%, 72%, and 200%, respectively from 2009 to 2013[8]. This trend causes major concerns to all parties concerned. Therefore, an analysis of causes resulting in traffic accidents in the Sultanate of Oman is vital in order to focus on strategies and countermeasures to reduce the occurrence and severity of these traffic accidents.
Figure-2. Traffic accidents and fatalities and injuries in the Sultanate of Oman, 1999-2013 [14].
OBJECTIVES
Due to the severe impacts of traffic accidents and the fatalities and injuries caused bythese accidents, this study was conducted to examine and investigate causes leading to these traffic accidents and to provide decision maker’s insights to causes of traffic accidents. Thus, assist them in developing polices and proactive actions to prevent and reduce the large number of traffic accidents and fatalities and injuries caused by these accidents in the Sultanate of Oman.
SENSITIVITY ANALYSIS
Sensitivity analysis has been described as the best method to evaluate the effects of multiple factors on an end result. Thus, many studies have been conducted to evaluate the relationship between a result and factors affecting itby utilizing sensitivity analysis [15-22].Until recently, sensitivity analysis was carried out by altering one factor, while fixing others, therefore concentrating on one variable at a time, which might lead to loss of the effect of the interaction between variables collectively [17]. Therefore, vigorous sensitivity analyses are needed.
Tornado plots
Tornado plots are frequently used as illustrative methods of the effect of individual factors on the result, it is determined using the Spearman’s rank correlation coefficient (ρ) as described in Equation (1).
𝜌 = 1 − 6 ∑ 𝑖2
𝑛 𝑛2−1 (1)
where,
ρ: Spearman’s rank correlation coefficient,
di: difference in the ranks between factor and result values in the same data pair, and
n: number of simulations.
The value of ρranges from -1.0 to 1.0 and the closer it gets to the ends of the range, the effect of the factor on the end result is the highest, and when it is closer to mid-range (i.e. zero), the effect of this factor is the least. In addition, a positive ρ value shows that the end resultis proportionally related to the variable (i.e., when the factor 0
2000 4000 6000 8000 10000 12000 14000
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Extreme tail analysis
Extreme tail analysis is utilized as a statistical method that is applied to methodicallyidentify uncertain critical variables in a model, which depends on the tail of the end result distribution. It is used to determine factors with the least effect on the end result to ensure better results. To recognize those factors, the normalized variable
α should be determined as shown in Eq.2 (more details are described in [24]). Factors in a group with |α| ≥ 0.5 are normally considered to be significant and contribute to the extremes of the end results. A positive αfor a factor indicates that extreme values of this factorresulted in positive extreme end result values and vice versa [17,19,20].
𝛼 = 𝑀 𝑖𝑎𝑛𝐺𝑟 −𝑀 𝑖𝑎𝑛𝑇 𝑎𝑙
𝜎𝑇 𝑎𝑙 (2)
where,
MedianGroup: median of a factor in the group,
MedianTotal: median of the factor in the total simulation, and
σTotal: standard deviation of the factor in the total simulations.
DATA ANALYSIS AND RESULTS
For the purpose of examining the sensitivity of factors affecting traffic accidents and number of fatalities and injuries due to traffic accidents, statistics (2003 -
2013) of traffic accidents and their causes were obtained from traffic statistics published by the General Administration of Oman Royal Police, Sultanate of Oman. To develop the tornado plots for factors causing traffic accidents, and based on their means and standard deviations, 50000 data points where generated using Monte Carlo Simulation, then the Spearman’s rank coefficients were determined using Equation (1).
When it comes to traffic accidents, results showed that among the studied factors, Failure to Follow Traffic Rules was ranked the most affecting factor, then Speeding and Reckless Driving with ρ values equal to 0.831, 0.431, and 0.272, respectively as shown in Figure-3. Furthermore, it was found that the Weather Condition, Drunk Driving, Fatigue, and Sudden Stopping had the lowest ρ values, thus their effect on Traffic accidents was low. As for fatalities caused by traffic accidents, Figure-4 illustrates the results of the sensitivity analysis. Speeding was determined the main factor in causing fatalities in traffic accidents with the highest ρ values equals to 0.803 and similar to traffic accidents, Weather Condition, Drunk Driving, Fatigue, and Sudden Stopping were ranked the least affecting factors. Furthermore, injuries due to traffic accidents followed the same trend (Figure-5), with Weather Condition, Drunk Driving, Fatigue, and Sudden Stopping were found to be the least affecting causes and Speeding, Reckless Driving and Failure to Follow Traffic Rules have the highest effects on injuries with ρ values equal to 0.661, 0.51, and 0.463, respectively.
Figure-3. Sensitivity analysis of factors causing traffic accidents.
0.007 0.007 0.009 0.010 0.017
0.031 0.044
0.056
0.272
0.431
0.831
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Weather Condition Drunk Driving Fatigue Sudden Stopping Road Defects Vehicle Defects Overtaking Safety Distance Reckless Driving Speeding Failure to Follow Traffic Rules
Figure-4. Sensitivity analysis of factors causing fatalities in traffic accidents.
Figure-5. Sensitivity analysis of factors causing injuries due to traffic accidents.
Using the same set of data points generated earlier, the extreme tail analysis was carried out. The reported traffic accidents, fatalities, and injures due to these accidentsvalues were arranged in descending order, for a reliable output, left and right tail of the distribution were evaluated. The lower and higher 5% of the reported
traffic accidents, fatalities, and injures due to these accidents values were taken as the left and right tails, respectively. Then α for all affecting factorswere determined using Eq.2 and the results are shown in Table-1.
0.002 0.023 0.024 0.042 0.046 0.068
0.105
0.281 0.307
0.331
0.803
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Sudden Stop Drunk Driving Weather Condition Road Defects Safety Distance Fatigue Vehicle Defects Overtaking Reckless Driving Failure to Follow Traffic Rules Speeding
ρ
0.009 0.011 0.019 0.022 0.022 0.040
0.121 0.131
0.463 0.510
0.661
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Drunk Driving Fatigue Sudden Stopping Weather Condition Road Defects Vehicle Defects Overtaking Safety Distance Failure to Follow Traffic Rules Reckless Driving Speeding
Table-1. Extreme tail analysis results.
Causing factors
Traffic accidents Fatalities Injures
α, Left tail α,
Righttail α, Left tail
α,
Right tail α, Left tail
α, Right tail
Speeding 0.852 -1.244 1.679 -1.679 1.359 -1.481
Failure to Follow
Traffic Rules 1.757 -0.934 0.71 -0.752 0.967 -0.83
Fatigue 0.023 -0.015 0.134 -0.117 0.021 -0.026
Drunk Driving -0.008 -0.015 0.075 -0.039 0.064 -0.036
Overtaking 0.052 -0.146 0.621 -0.59 0.272 -0.3
Weather Condition 0.007 -0.031 0.053 -0.069 0.025 -0.041
Sudden Stop 0.002 -0.024 -0.003 0.002 0.046 -0.015 Safety Distance 0.066 -0.194 0.089 -0.071 0.309 -0.287
Reckless Driving 0.52 -0.767 0.671 -0.651 1.137 -0.792
Vehicle Defects 0.038 -0.097 0.221 -0.227 0.095 -0.11 Road Defects 0.052 -0.047 0.078 -0.046 0.036 -0.061
Similar to the tornado plots (Figures 3-5) results, it was found that Failure to Follow Traffic Rules, Speeding, and Reckless Driving had the highest effects on traffic accidents, fatalities, and injures due to these accidents extreme values. However, Speeding could be considered as the primary contributor. Thus, an effective control of Speeding would lead to a reduction if traffic accidents and fatalities and injuries resulting from traffic accidents.On the other hand, the conducted Extreme Tail Analysis showed that Overtaking, Vehicle Defects, and Safety Distance had a significant effect on the numbers of traffic accidents and resulting fatalities, and injures.
SUMMARY AND CONCLUSIONS
The large number of fatalities and injuries caused by traffic accidents in the developing countries has been on the rise in an alarming rate. This large number has a devastating impact on all societies. In the Sultanate of Oman, with its low population, the numbers of these accidents are very high, especially when it is compared to neighboring countries. Thus, this study was conducted to investigate the leading causes to these accidents to assist policy makers in reducing traffic accidents in the Sultanate of Oman, by identifying the leading causes to traffic accidents.
To determine the role of causes and their contribution to numbers of fatalities and injuries resulting from traffic accidents in the Sultanate of Oman, sensitivity analyses were conducted to identify key factors, which would enable decision makers to focus their efforts on the reducing traffic accidents. Until recently, changing a one factor while maintaining other factors constant was the norm when conducting sensitivity analysis,which might lead to loss of evaluating the interaction between the different factors and their collective effect on the end
main advantage of these tools is that they can capture the combined effect of different factors and quantify the sensitivity of a specific factor on anend result.
The means and standard deviations of the traffic accidents and fatalities and injuries resulting from those accidents were utilized in a Monte Carlo simulation to generate 50000 data points. These data points were used to generate tornado plots and to conduct the extreme tail analysis. Results have shown that Failure to Follow Traffic Rules, Speeding and Reckless Driving were ranked the highest contribution factors to traffic accidents and fatalities and injuries from those accidents, while Weather Condition, Drunk Driving, Fatigue, and Sudden Stopping had very low effect on traffic accidents. It may come as a surprise that drunk driving had low effect;however, it is speculated that since the Sultanate of Oman have strict regulations when it comes to drugs and alcohol consumption and culturally considered as unacceptable behavior, the number of traffic accidents caused by drunk driving is low.
RECOMMENDATIONS
a) Ideal use of road safety measures: construction of road dividers and speed humps, installation of speed control devices in critical locations.
b) Addressing drivers’ wrong behavior by developing traffic safety awareness programs and campaigns from early stages (school level), and strengthen promoting compliance with traffic rules and regulations.
c) Similar toneighboring countries (e.g. United Arab Emirates), a point system that is based on traffic violations and offences should be introduced to assist in the reduction of wrong drivers’ behaviors.
d) Developing a better and effective public transportation system to reduce number of private vehicles on the road.
e) Establish and develop outlet venues (e.g. race tracks), similar to the European countries experience, where drivers can practice speeding and drifting in a safe environment, without affecting other road users. f) Authorities should intensify the efforts to enforce
traffic regulations especially when it comes to speeding and the usage of mobile phones while driving.
ACKNOWLEDGEMENT
The author would like to extend their gratitude to the General Administration of Oman Royal Police in Salalah, Sultanate of Oman for their kind cooperation and for providing needed data and statistics to complete this study. Furthermore, the author would like to would like to extend his gratitude to Eng. Yousuf A. Al-Ibrahim for his efforts and assistance in conducting this study.
REFERENCES
[1] Ali GA. 2010. Traffic Accidents and Road Safety Management; A Comparative Analysis and Evaluation in Industrial, Developing and Rich-Developing Countries. Proceedings of 29th Annual Southern African Transport Conference, South Africa: 530-540.
[2] Tamimi TM, Daly M, Bhatty M, and Lutt A. 1980. Causes and Types of Road Injuries in Asir Province, Saudi Arabia, 1975-1977: Preliminary Study. Saudi Medical Journal. 1(5): 249-255.
[3] Badawi IA, Alakija W, and Aziz MA. 1995. Road Traffic Accidents in Asir Region, Saudi Arabia: Pattern and prevention. Saudi Medical Journal. 16(3): 257-260.
[4] Al-Ghamdi AS. 2003. Analysis of Traffic Accidents
[5] Bendak S. 2005. Seat belt utilization in Saudi Arabia and its Impact on Road Accident Injuries. Accident Analysis and Prevention. 37: 367-371.
[6] Gharaibeh ES and Abu Abdo AM. 2011. Assessment of Traffic Safety and Awareness among Youth in Al-Ahsa Region, Saudi Arabia. Journal of Emerging Trends in Engineering and Applied Sciences (JETEAS). 2(2): 210-215.
[7] Abu Abdo AM and Al-Ojaily AA. 2015. Assessment of Awareness of Livestock-Vehicle Traffic Accidents in Dhofar Region, Oman. International Journal of Applied Engineering Research (IJAER). 10(18): 38955-38959.
[8] Abu Abdo AM and Al-Ibrahim YA. 2015. Assessment of Traffic Safety and Awareness among Road Users in Salalah, Sultanate of Oman. Research Journal of Applied Sciences, Engineering and Technology 12(5): 574-581.
[9] Abu Abdo AM, Akash BA, Mohsen MS and Akash OB. 2016. Scopus Based Analysis of Related Literature to Traffic Safety and Awareness in the Arab Countries. International Journal of Applied Engineering Research (IJAER). 11(6): 4076-4080.
[10]Mazharul MI and Al Hadhrami YS. 2012. Increased Motorization and Road Traffic Accidents in Oman. Journal of Emerging Trends in Economics and Management Sciences. 3(6): 907-914.
[11]WHO. 2013. Global Status Report on Road Safety: Supporting a Decade of Action, Department of Violence & Injury Prevention & Disability. World Health Organization, Geneva, Switzerland.
[12]WHO. 2013. Road Traffic Injuries, Fact sheet N°358, World Health Organization, Geneva, Switzerland.
[13]Farag SG, Hashim IH and El-Hamrawy SA. 2014. Analysis and Assessment of Accident Characteristics: Case Study of Dhofar Governorate, Sultanate of Oman. International Journal of Traffic and Transportation Engineering. 3(4): 189-198.
[16] Sues RH, Lua YJ, Dass SM, and Murfee J. 1993. Reliability-Based Analysis and Design of Flexible Airfield Pavements. Airport Pavement Innovations Theory to Practice. pp. 76-89.
[17]Thyagarajan S. 2009. Improvements to Strain Computation and Reliability Analysis of Flexible Pavements in the Mechanical-Empirical Pavement Design Guide. Thesis (PhD), Washington State University, USA.
[18]Killingsworth BM and Zollinger DG. 1995. Sensitivity Analysis of Input Parameters for Pavement Design and Reliability. Transportation Research Record, 1482: 111-122.
[19]Abu Abdo AM. 2011. Comparison of Two Fracture Parameters for Asphalt Mixes Using Sensitivity Analyses. Journal of Emerging Trends in Engineering and Applied Sciences (JETEAS), 2 (4): 706-710.
[20]Abu Abdo AM. 2012. Sensitivity Analysis of a New Dynamic Modulus (| E*|) Model for Asphalt Mixtures. Road Materials and Pavement Design, 13(3): 549-555.
[21]Haider SW, Salama HK, Buch N and Chatti K. 2007. Significant M-EPDG Design Inputs for Jointed Plain Concrete Pavements in Michigan. MAIREPAV5 - Proceedings of the Fifth International Conference on Maintenance and Rehabilitation of Pavements and Technological Control, Utah, USA.
[22]Saltelli A, Chan K, and Scott EM. 2000. Sensitivity Analysis. USA: John Wiley & Sons.
[23]Eschenbach TG. 2006. Technical Note: Constructing Tornado Diagrams with Spreadsheets. The Engineering Economist. 51: 195-204.
