겨울철 국내 도로 결빙으로 인한 교통사고가 증가하는 추세를 보이고 있으며 2018년~2022년까지 총 4,609건의 결빙 교통사고가 발 생하였다. 결빙 교통사고의 치사율은 2.3으로 일반적인 교통사고와 비교하여 높은 치사율을 보이며 최근 5년(2018~2022)동안 결빙 교 통사고로 인하여 107명이 사망자와 7,728명의 부상자가 발생하였다. 현재 국토교통부에서 제시한 결빙 취약구간 평가기준표에 따라 결 빙 위험 구간을 지정하고 있으나, 해당 기준은 결빙의 주요 요인으로 고려되는 기상조건을 충분히 반영하지 못하고 있다. 도로 결빙은 노면온도가 0℃ 이하이며 노면에 수분이 공급될 때 형성되며 기온, 구름량, 풍속, 풍향, 상대습도, 강수량 등의 기상인자들이 복합적으 로 작용하여 결빙이 발생한다는 점을 고려하였을 때, 기상 특성은 도로 결빙의 주요 인자로 판단된다. 따라서 국내 결빙 취약구간 평 가기준의 개선이 필요하며 본 연구의 목적은 국내 결빙 교통사고 데이터를 분석하고 결빙이 형성되는 기상 조건을 구체화하는 것이다. 분석을 위한 데이터로 2018년~2022년까지 5년동안 발생한 결빙사고 사례와 기상청 방재기상관측소(AWS)에서 제공하는 기상 데이터 를 적용하였다. 이후, 박스도표, 확률밀도함수 등의 통계분석을 적용하여 결빙 형성 기상 조건을 구체화하였다. 이를 통하여 기존 결빙 취약구간 평가기준의 기상학적 개선 방향성을 제시할 수 있으며 더 나아가 도로 결빙 예측 로직 개발을 기대할 수 있다.

PURPOSES : This study aims to analyze the causes of pedestrian traffic accidents on community roads. METHODS : This study collected variables affecting pedestrian traffic accidents on community roads based on field surveys and analyzed them using negative binomial regression and zero-inflated negative binomial regression models. RESULTS : Model analysis results showed that the negative binomial regression model is more suitable than the zero-inflation negative binomial regression model. Additionally, the segment length (m), pedestrian volume (persons/15 min), traffic volume (numbers/15 min.), the extent of illegal parking, pedestrian-vehicle conflict ratio, and one-way traffic (one: residential, two: commercial) were found to influence pedestrian traffic accidents on community roads. Model fitness indicators, comparing actual values with predicted values, showed an MPB of 1.54, MAD of 2.57, and RMSE of 7.03. CONCLUSIONS : This study quantified the factors contributing to pedestrian traffic accidents on community roads by considering both static and dynamic elements. Instead of uniformly implementing measures, such as pedestrian priority zones and facility improvements on community roads, developing diverse strategies that consider various dynamic factors should be considered.

PURPOSES : This study investigates the factors affecting extra-long tunnel accidents by integrating data on tunnel geometry, traffic flow, and traffic accidents and derives the underlying implications to mitigate the severity of accidents. METHODS : Two processes centered on three key data points (tunnel geometry, traffic flow, and traffic accidents) were used in this study. The first is to analyze the spatial characteristics of extra-long tunnel traffic accidents and categorize them from multiple perspectives. The other was to investigate the factors affecting extra-long tunnel traffic accidents using the equivalent property-damage-only (EPDO) of individual accidents and the aforementioned data as the dependent and independent variables, respectively, by employing an ordered logistic regression model. RESULTS : Gyeonggi-do, Gyeongsangnam-do, and Gangwon-do are three metropolitan municipalities that have a significant number of extra-long tunnel accidents; Busan and Seoul have the most extra-long tunnel accidents, accounting for 23.2% (422 accidents) and 18.6% (339 accidents) of the 1,821 accidents that occurred from 2007 to 2020, respectively. In addition, approximately 70% of extra-long tunnel traffic accidents occurred along tunnels with lengths of less than 2 km, and Seoul and Busan accounted for over 60% of the top 20 extra-long tunnels with accidents. Most importantly, the Hwangryeong (down) tunnel in Busan experienced the most extra-long tunnel traffic accidents, with 77 accidents occurring during the same period. As a result of the ordered logistic regression modeling with EPDO and multiple independent variables, the significant factors affecting the severity of extra-long tunnel traffic accidents were determined to be road type (freeway, local route, and metropolitan city road), traffic flow (speed), accident time (year, summer, weekend, and afternoon), accident type (rear end), traffic law violations (safe distance violation and center line violation), and offending vehicles (van, sedan, and truck). CONCLUSIONS : Based on these results, the following measures and implications for mitigating the severity of extra-long tunnel traffic accidents must be considered: upgrading the emergency response level of all road types to that of freeways and actively promoting techniques for regulating high-speed vehicles approaching and traversing within extra-long tunnels are necessary. In addition, the emergency response and preparation system should be reinforced, particularly when the damage from extra-long tunnel traffic accidents is more serious, such as during the summer, weekends, and afternoons. Finally, traffic law violations such as safe distance and centerline violations in extra-long tunnels should be prohibited.

Ensuring the safe arrival of delivery cargo at its intended destination is of utmost importance. Truck drivers play a crucial role in guaranteeing the secure delivery of cargo without any mishaps. However, there are various factors that may lead to delayed arrival of trucks at their destination, such as late departures or prolonged loading operations. The timely departure of cargo transportation is contingent upon several variables, including the driver's experience, cargo volume, and loading time. If the transportation commencement is delayed, it may increase the risk of accidents due to an elevated operating speed. Consequently, we conducted a study to investigate the correlation between cargo loading time, cargo volume, driving experience, and the likelihood of accidents. Our findings indicate that both cargo volume and driver experience can impact the likelihood of vehicle accidents. Furthermore, all factors can have an interactive effect on the occurrence of accidents. However, extending the loading time may mitigate the impact on the likelihood of accidents.

PURPOSES : This study empirically analyzes the determinants of fatal accidents based on raw data on traffic accidents occurring in Chungnam in 2020. METHODS : Regression models based on theoretical arguments for fatal traffic accidents are estimated using a binomial logit model. RESULTS : The prediction model for fatal accidents is affected by the degree of urbanization of the region, month and day of the accident, type of accident, and type of law violation. In addition, speeding or illegal U-turns among law violations appear more likely to result in fatal accidents. The road surface conditions at the time of the accident do not show a significant difference in the probability of fatality among traffic accidents. However, the probability of a fatal accident is rather lower in case of a snowy road; this is plausible, as drivers tend to drive more carefully in bad weather conditions. CONCLUSIONS : Among traffic accidents, fatal accidents appear to be affected by the time and place of the accident, type of accident, and weather conditions at the time of the accident. These analysis results suggest policy implications for reducing fatal accidents and can be used as a basis for establishing related policies.

PURPOSES : In this study, the main factors affecting the severity of traffic accidents among elderly drivers were reviewed, and accident factors with a high accident risk were analyzed. This provided basic data for preparing a traffic safety system for elderly drivers and establishing policies. METHODS : Based on machine learning, the major factors influencing accident severity (from the analysis of traffic accident data for elderly drivers) were analyzed and compared with existing statistical analysis results. The machine learning algorithm used the Scikit-learn library and Python 3.8. A hyperparameter optimization process was performed to improve the safety and accuracy of the model. To establish the optimal state of the model, the hyperparameters were set (K = 5) using K-fold cross-validation. The hyperparameter search applied the most widely utilized grid search method, and the performance evaluation derived the optimal hyperparameter value using neutral squared error indicators. RESULTS : The traffic laws, road sections of traffic accidents, and time zones of accidents were analyzed for accidents involving elderly drivers in Daejeon Metropolitan City, and the importance of the variables was examined. For the analysis, a linear regression model, machine learning-based decision tree, and random forest model were used, and based on the root mean square error, the random forest accuracy performance was found to be the best. Ultimately, 18 variables were analyzed, including traffic violations, accident time zones, and road types. The variables influencing the accident severity were the speed, signal violation, intersection section, late-night driving, and pedestrian protection violation, with the relative importance of the variables in the order of speed (0.3490966), signal violation (0.285967), and late-night driving (0.173108). These can be seen as variables related to the expansion of life damage owing to physical aging and reduced judgment abilities arising from decreases in cognitive function. CONCLUSIONS : Restricting the driving of the elderly on the expressway and at night is reasonable, but specific standards for driving restrictions should be prepared based on individual driving capabilities.

PURPOSES : A highway operates in a continuous flow and has restricted access. When an accident occurs on a highway, the impact on the traffic flow is large. In particular, an accident that occurs in a tunnel has a more significant impact than an accident that occurs in a general section. Accordingly, the management agency classifies the tunnel as a dangerous section and manages a tunnel of more than 1000 m using the Tunnel Transportation Management System. The purpose of this study was to select dangerous tunnels that require intensive management for the efficient management of highway tunnels. METHODS : In this study, for the selection of dangerous tunnels for expressways, all highway tunnels were classified into five clusters by characteristics. The traffic accident severity — equivalent property damage only (EPDO) — for each tunnel cluster was derived through a traffic accident analysis. Based on the severity analysis results, the safety performance function (SPF) for each cluster was established, and the accident risk tunnel was selected based on the potential safety improvement (PSI) value of each tunnel calculated using the empirical Bayes (EB) method for each tunnel cluster. RESULTS : As a result of the analysis, accident risk tunnels were selected based on the PSI values of the tunnels for each highway tunnel group. Finally, 55 hazardous tunnels were identified as hazardous tunnels: 13 tunnels in Cluster 1, 3 tunnels in Cluster 2, 15 tunnels in Cluster 3, 18 tunnels in Cluster 4, and 6 tunnels in Cluster 5. CONCLUSIONS : After classifying all 1232 tunnels on the highway into five clusters according to tunnel characteristics, EPDO analysis was performed for each tunnel cluster. To this end, the SPF for each cluster was constructed, and accident risk tunnels were selected based on the PSI value of each tunnel calculated using the EB method for each tunnel cluster. The tunnel cluster was classified as a typical tunnel type. As a result, most of the first and second values were calculated from cluster E (long tunnel cluster).

PURPOSES : In this study, we quantitatively prove the rubber necking phenomenon for highway traffic accidents and develop a calculation model based on the influencing factors. METHODS : Vehicle detector speed data in the opposite direction to the accident point were used based on the accident data on highways over the past three years, and a comparative verification was performed between nearby vehicle detector data to verify the reliability of the data. Accordingly, a binomial logistic model, ordinal probit regression model, and multilinear regression model were developed to compare the orientation. RESULTS : There was a difference in the influencing factors based on the dependent variable, and the day of the week, vehicle type, weather, longitudinal slope, and median height had an effect. Through a regression analysis, an influence coefficient was derived to calculate the driving speed deceleration value by rubbernecking. The results of the model analysis proved that the speed reduction caused by rubbernecking was more evident during the daytime than at night, during weekends compared to weekdays, and the speed reduction was more obvious for heavy vehicles compared to other types of vehicles. It can also be concluded that longer clearance time, higher accident severity, and higher traffic volume affect traffic delay. To verify the data and model equation, the mean prediction bias (MPB) and mean absolute deviation (MAD) were calculated for hundred cases randomly extracted from the collected accident data. These results were excellent. CONCLUSIONS : It can be developed into a human-engineered model that reflects various road/facility conditions, such as highways, other lanes, general roads, and roads without a median strip. This study is meaningful as a basic study on the quantitative effect of rubber necking.

PURPOSES : This study focuses on advance traffic information to prevent secondary traffic accidents on express highways. The purpose of this study is to analysis the optimal scenario by evaluating the effect of each advance traffic information scenarios using virtual driving simulation. METHODS : By designing traffic information scenarios and services with a combination of VMS and mobile PUSH notifications, driver behavior in the event of a traffic accident was analyzed. For this, a simulation environment was designed through engineering analysis. Through virtual driving simulation, the driver's deceleration point and the perception-reaction time are analyzed. RESULTS : Four scenarios were designed and reviewed so that VMS and mobile PUSH notification can be provided simultaneously after the driver drove for 5 km. As a result of driving with 30 drivers, the drivers reacted fastest when VMS was installed, followed by mobile PUSH notification, VMS+mobile PUSH notification, and NOTHING. CONCLUSIONS : When designing traffic information service, it was observed that providing information through VMS alone is more efficient than providing two services of traffic information. Therefore, it can be used as basic data for preventing secondary accidents on express highway.

PURPOSES : In this study, the installation of drowsy rest areas and accidents are analyzed. The factors that affected the accidents caused by drowsy drivers in rest areas are analyzed to improve the safety of rest areas. METHODS : By comparing and analyzing the installation status of the rest areas for drowsy drivers, the accident status were analyzed. The logistic regression model was used to analyze the factors that affect accidents in the drowsy rest area. RESULTS : Most rest areas were installed below the installation criteria. Several accidents occurred when the vehicle entered the drowsy rest area. These rest areas had a short entry ramp, and no safety facilities were installed. The logistic regression model showed that the risk of an accident is lowered when the deceleration lane is longer than 215 m. Additionally, the risk of an accident is lowered when the rest area is installed in the straight section or the curve section, wherein the curve radius is greater than 2 km. CONCLUSIONS : In this study, we evaluated the installation status of the rest areas for drowsy drivers by comparing installation elements. Most rest areas for drowsy drivers were installed at different lengths of the ramp. Some of these were installed on the slope or curved sections of the road. We analyzed the accident status and developed an accident modal using the logistic regression model to identify the factors that affect accidents. It will be necessary to analyze accidents in drowsy rest areas continuously to improve safety for drowsy drivers.

PURPOSES: The purpose of this study is to develop a traffic accident prediction model using statistical data, to analyze child traffic accidents in school zones. Furthermore, we analyze the factors affecting child traffic accidents, as obtained from the results of the developed model. METHODS : From the literature review, we obtained data for child traffic accidents and various variables relating to road geometry and traffic safety facilities in school zones. We used these variables and data to develop a child traffic accident analysis model. The model was then developed into three types using the Limdep 9.0 statistical tool. RESULTS: As a result of the overdispersion test, the Poisson regression model was applied to all types of models with an overdispersion coefficient of close to zero. The results of the model development are as follows. The main model (all scope of analysis) was for a kindergarten, considering a local roadway, the accessibility of the roadway, the number of unsignalized intersections, and the school zones in commercial area as variables that increase traffic accidents. Sub-model typeⅠ(only the roadway connected to the main entrance) was for a kindergarten, considering a local roadway, skid resistant pavement, no-parking signs, the number of unsignalized intersections, and the number of commercial facilities as variables that increase traffic accidents. The main model and sub-model type Ⅰ showed a reduction in accidents when using forward-type traffic signals. Sub-model typeⅡ(only the roadway not connected to the main entrance) shows that the local roadway is the variable that most increases the probability of traffic accidents. However, when the roadway and walkway are separated, the probability of traffic accidents decreases significantly, by up to 90%. CONCLUSIONS: The results of this study demonstrate the need to restructure the method used to improve school zones. Moreover, the effect of various traffic safety facilities was quantitatively analyzed.

In recent year, marine safety has been one of top concerns in Korea. In this paper, general statistics of ship in 159 waterways in Korea from 2007 to 2017 were considered. Main objective of this research is to investigate the relationship between the number of marine accidents and traffic conditions in narrow waterways by multiple regression analysis method. The result shows that the number of vessels, the width, the length and the depth of narrow waterway have an influence on the number of maritime accidents in corresponding area. Additionally, the number of vessels sailing has the most significant impact on the number of accidents in narrow waterway area.

PURPOSES : This study was conducted to develop a traffic accident prediction model using traffic accident data and management and service evaluation data on bus companies in Busan, and to determine the possibility of establishing customized traffic accident prevention measures for each company. METHODS: First, we collected basic data on the characteristics of urban bus traffic accidents and conducted basic statistical analysis. Then, we developed traffic accident prediction models using Poisson regression and negative binomial regression to examine the characteristics of major items of management and service evaluation affecting traffic accidents. RESULTS : The Poisson regression model showed overdispersion; hence, the negative binomial regression model was selected. The results of the traffic accident prediction model developed using negative binomial regression are acceptable at 95% confidence level (a = 0.05). CONCLUSIONS : The traffic accident prediction model indicates that the management of the traffic record system and internal and external management items in service evaluation have a significant effect on the reduction of traffic accidents. In particular, because human factors are the main cause of traffic accidents, bus traffic accidents are expected to greatly decrease if drivers' dangerous driving behaviors are effectively controlled by bus companies.

해양사고에 관한 많은 연구와 분석에 따르면 약 80%가 인적 오류에 의하여 발생되고 있는 것으로 파악되고 있다. 해양사고의 예방대책 수립을 수립하기 위하여 사고를 일으킨 배후 인적 요인을 파악하는 연구가 반드시 필요하다. 따라서 본 연구의 주목적은 m-SHEL 모델을 이용하여 해상교통 관련 사고의 배후 인적 요인을 파악하고 분석하는 것이다. 다른 분야에서 사용되는 m-SHEL 모델은 일반적인 인적 요인의 개념을 기반으로 되어 있기 때문에 본 연구에서는 선박운항시스템에 수용하기 위하여 이 모델을 확장하여 인적 요인 을 정의하였다. 또한, 이 확장된 모델의 타당성을 SPSSWIN의 신뢰성 분석을 통하여 검증하였다. 그리고 이 확장된 m-SHEL 모델의 분류표 사용하여 해양안전심판원의 재결서에서 추출한 자료로부터 해상교통 관련 사고의 배후 인적 요인을 분석하였다. 해상교통 관련 사고의 배후 인적 요인을 분석한 결과 조선자 자신에 관한 요인 L이 가장 많았으며 다음으로 L-E, L-m, L-H, L-S 및 L-L 순으로 나타났다. 이 연구는 해상교통 관련 사고의 예방 및 해상안전관리시스템 구축을 위한 유용한 분석 결과를 제시함으로써 인적 요인에 의한 해상교통 관련 사고 방지에 기여할 것으로 판단된다.

PURPOSES: The effects on traffic accidents change with the changing environment. Accordingly, this study analyzes the characteristics of traffic accidents based on the personal characteristics (gender and age) of drivers, and those of 25 autonomous districts in Seoul, and suggests improvements. METHODS: Based on data pertaining to traffic accidents in Seoul, the analysis of accident characteristics was conducted by categorizing the types of traffic accidents according to the drivers' gender and age, and characteristics of 25 autonomous districts in Seoul. Further, for statistical verification, the SPSS software was used to derive influence variables through a multinomial logistic regression analysis, and a method for reducing traffic accidents was proposed. RESULTS : Analysis results show that males tend to be more involved in speed-related accidents and females in low-experience drivingrelated accidents such as those during parking and alleyway driving. In addition, variables such as age, automobile type, district, and day of the week are found to influence accident types. CONCLUSIONS : This study analyzed the accident characteristics based on personal and city characteristics to reflect the sociological characteristics that influence traffic accidents. The number of traffic accidents in Korea could be decreased drastically by implementing the results of this study in customized safety education and traffic maps.

PURPOSES : The objective of this study is to identify the characteristics affecting traffic accidents that have occurred in 564 industrial complexes nationwide from 2011 to 2015.METHODS : The traffic accidents were specified using various factors such as industrial complex type (national VS. general), industrial complex degradation (old VS. non-old), location of complex (capital VS. non-capital), and traffic law violation (speeding, signal violation, and median invasion). The average number of crashes and accident ratio (fatal, severe, and both) in terms of characteristics of industrial complexes were calculated. With a sample of crashes of the industrial complexes for 5 years, statistical significances were tested to analyze and compare the differences based on industrial complex and traffic law characteristics using parametric and non-parametric methods.RESULTS: From statistical results, it is observed that the crash frequency occurring in old industrial complexes is three times higher than that in non-old industrial complexes. Old industrial complexes located in a capital area, old national industrial complexes, and old general industrial complexes are considerably related to higher crash frequency, but the fatal accident ratio appeared to have no statistical difference across industrial complex characteristics. Severe crashes are more likely to occur in non-old industrial complexes on an average.CONCLUSIONS : It is necessary to eliminate potential threats to roads and traffic in the same manner as illegal parking in industrial complexes through the restoration of old industrial complexes. To improve the efficiency of road infrastructure, efforts should be made to improve traffic safety in accordance with industrial characteristics such as planning and operation of relevant local government programs.

기후변화로 인한 결빙, 폭설, 집중호우 등 도로에서의 기상조건은 해마다 수많은 인명 피해를 유발하고 있다. 그러나 급변하는 기상상태에 신속하게 대처할 수 있는 도로기상 대응체계가 미흡하여 교통사고로 인한 사회·경제적 손실이 가중되고 있다. 최근 교통사고 통계자료에 의하면 기상상태가 좋지 않을 때의 교통사고치사율은 맑은 날씨일 때 보다 2～3배 심각한 것으로 나타났다(교통안전공단 2016). 따라서 기상악화 시 교통사고예방에 필요한 구체적이고 상세한 기상정보를 운전자에게 제공함으로써 교통사고예방 및 도로 운영의 활용성을 높일 수 있는 도로기상 모델개발이 필요하다. 도로기상 모델과 관련해서 도로기상 선진국(유럽, 캐나다, 일본, 미국 등)은 1990년대 초반부터 도로 노면온도와 노면 결빙점 예측을 위한 다양한 모형을 개발하고 모델의 성능을 검증하기 위한 노력을 기울이고 있다(Meng, 2014; Sato, 2004). 이들 국가는 주요 도로에 대해서 도로기상 관리지원시스템을 구축‧운영하고 있으며, 개발된 모델로부터 추정된 노면상태 정보를 도로이용자 및 도로관리자에게 제공함으로써 교통사고 및 도로파손 손실에 따른 사회적 비용을 줄여나가고 있다. 우리나라 도로기상 분야는 결빙 취약구간인 터널의 입·출입부에 수용액 형태의 제설제 자동 분사장치를 설치하여 운영하고 있는 단계로서 도로기상 선진국과 같이 도로구간 단위로 확장하여 서비스할 수 있는 기술은 미흡한 단계이다. 이들 국가와 같이 도로기상 관리지원시스템을 운영하여 주요 도로에 대해서 도로기상 정보를 실시간으로 서비스하기 위해서는 도로의 한 지점만을 대상으로 하는 기술대응 보다는 전체 도로구간의 노면상태를 예측할 수 있는 기상상황별 통합모델 개발이 우선시 되어야 한다. 이러한 기술개발의 필요성에 따라서 본 연구에서는 기상상황별 노면상태를 서비스할 수 있는 통합모델을 개발하였다. 첫 번째 모델은 겨울철 도로노면의 상태를 예측할 수 있는 모델이다. 노면상태 예측모델은 기계학습 방법 중에 하나인 의사결정나무(Decision Tree) 알고리즘 기반으로 개발되었으며, 이동형 차량(Probe vehicle)에서 수집되는 기상입력 자료를 학습하여 기상요소에 따른 4가지 노면상태(Dry, Moist, Wet, Ice)를 예측할 수 있다. 두 번째로 집중호우 기간에 도로에서 발생하는 도로의 수막정보를 예측할 수 있는 모형이다. 도로의 수막예측 모델은 기상청에서 제공하는 1시간 단위의 AWS(Automatic Weather System) 자료와 10m 단위의 도로기하구조 정보를 활용하였다. 도로에서의 강우량 추정은 kNN(k-Nearest Neighbors)기법을 활용하여 대상도로와 가장 가까운 AWS 관측소의 강우정보를 각 해당 도로구간에 할당하였다. 현재 본 도로기상 통합모델은 서울시 내부순환로와 올림픽대로에 대해서 실시간 서비스할 수 있도록 생산체계가 구축되었다. 도로기상 통합모델은 우리나라 주요 도로에 적용이 가능하여 하절기의 집중호우로 인한 물이 고인 도로지점과 동절기의 기온강하로 인한 도로의 결빙현상 등 도로노면상태 정보를 실시간으로 서비스할 수 있다. 또한 기상 악화 시 도로를 효율적으로 관리할 수 있는 의사결정 도구로 활용되어 교통재난 및 도로파손을 줄이는데 기여할 수 있을 것이다.

Even though cars have a good effect on modern society, traffic accidents do not. There are traffic laws that define the regulations and aim to reduce accidents from happening; nevertheless, it is hard to determine all accident causes such as road and traffic conditions, and human related factors. If a traffic accident occurs, the traffic law classifies it as ‘Negligence of Safe Driving’ for cases that are not defined by specific regulations. Meanwhile, as Korea is already growing rapidly elderly population with more than 65 years, so are the number of traffic accidents caused by this group. Therefore, we studied predictive and comparative analysis of the number of traffic accidents caused by ‘Negligence of Safe Driving’ by dividing it into two groups : All-ages and Elderly. In this paper, we used empirical monthly data from 2007 to 2015 collected by TAAS (Traffic Accident Analysis System), identified the most suitable ARIMA forecasting model by using the four steps of the Box-Jenkins method : Identification, Estimation, Diagnostics, Forecasting. The results of this study indicate that ARIMA (1, 1, 0)(0, 1, 1)12 is the most suitable forecasting model in the group of All-ages; and ARIMA (0, 1, 1)(0, 1, 1)12 is the most suitable in the group of Elderly. Then, with this fitted model, we forecasted the number of traffic accidents for 2 years of both groups. There is no large fluctuation in the group of All-ages, but the group of Elderly shows a gradual increase trend. Finally, we compared two groups in terms of the forecast, suggested a countermeasure plan to reduce traffic accidents for both groups