Bridging the gap- GIS a key in crash data display and analysisAarti Bhasin and M.Parida Civil Engg Deptt Punjab Engg College Chandigarh India E-mail: bhasinaarti@gmail.com ABSTRACT: - Traffic safety is a major concern in developing countries like India. An accident takes place every 3 minutes and a person is dead every 10 minutes on Indian roads. The Indian share in world vehicle population is only 1% while its share in road accidents is as high as 6%, contributing heavily to the suffocation of economic growth due to high cost related to them. Thus necessitating the need of an efficient and effective crash data analysis system so as to study collisions in depth and to determine if they could be prevented by geometrical field improvement methods. However the applicability and reliability of accident analysis depends on the ability to integrate relevant inputs from disparate databases in a seamless and automated manner. With powerful functionality in spatial referencing, data management and visualization GIS provides a natural platform for such a work. The paper describes the development of a GIS based application for the display and analysis of crash data. The study area includes a segment of NH-58 stretching from Delhi to Rishikesh. Crash data related to the stretch was collected along with the geometric characteristics of the road. Using various tools like dynamic segmentation and spatial analyst, display and analysis of the data was carried out. Queries were generated so as to establish every combination possible between high accidents and various geometrical features. The entire work aims at providing a user-friendly query based system, which can be used by various departments to ascertain the probable causes of accidents INTRODUCTION: The alarming growth rate of causalities and damage to vehicles in road accidents since the recent past, coupled with gigantic vehicle growth all over the world have created complex problems of ensuring traffic safety on roads. Thus the aspect of traffic safety on roads has been attracting a lot of attention in all countries of the world. The importance of preventing road accidents will rapidly grow in near future. Around 5,00,000 persons are killed on the roads annually throughout the world and those injured are many fold. Developing countries are having huge fatality rate in comparison to the developed countries. Many countries are experiencing an annual vehicle growth rate of around 17%. Examples are India (17%), China (18%), Vietnam (18%) and Malaysia (15%). The number of deaths and injuries has been reducing steadily in the developed countries where as the number of deaths in Asian countries is rising at alarming rates. During 1981-1993, the number of road accident deaths has increased in Asia Pacific region by 95% where as total population has increased only by 24%. Even on Indian National Highways the road accidents are increasing rapidly. The fatalities of road accidents have increased from 64,000 in year 1994 to 85,000 in year 2001.The National Highways though contribute only 2% of the total road network, 25% of road accidents occur on them, out of which 34% lead to fatalities and 28% to injuries. Road Traffic Safety The safety of road traffic is improved by conducting complex measures. A considerable portion of them is connected with improvement in the design of motor vehicles, beginning with the introduction of better braking systems and ending with the provision of such special contrivances as safety belts for drivers and passengers, a soft instrument panel and collapsing steering column. Improving road traffic safety among drivers, the population and at schools plays an appreciable role. Traffic safety is also achieved, however, by improving the methods of designing streets, roads and highway and technology of performing separate road construction and maintenance jobs, in particular the method of imparting a smooth and non-skid surface to pavements. Road traffic safety can be ensured only by simultaneously complying with the number of requirements namely:
Thus traffic safety can be ensured only by the joint efforts of the specialists in the branch of science and engineering and the road users. Road Vehicle Driver System Traffic accidents are caused by the complex interaction of "Road Vehicle Driver System". So for any accident, either all the above three factors are collectively responsible or the combination of any of the sub-systems i.e. road vehicles, vehicle driver and road driver is responsible. Accident may also occur due to any individual factor including the environmental factor. Traffic safety is affected by
Factors Influencing Accident Rates The research done in various countries have shown that almost 80% of the cases, the main cause of road accidents is bad road conditions. The accident rate is a function of large number of factors, which include
Need of the Study As can be easily inferred from the above stated statistics that there is an urgent need of an efficient and effective crash data analysis system so as to study collisions in depth and to determine if they could be prevented by geometrical field improvement methods. However the applicability and reliability of accident analysis depends on the ability to integrate relevant inputs from disparate databases in a seamless and automated manner. With powerful functionality in spatial referencing, data management and visualization GIS provides a natural platform for such a work. The objective of the study is development of a GIS based system for the display and analysis of crash data as well as "determination of accident-prone locations", which is user-friendly, efficient and effective for use in various departments. GIS- A KEY IN TRAFFIC ACCIDENT ANALYSIS: Traffic accident analysis means to investigate the causes of accidents, to determine hazardous locations (Accident Prone Locations) and to determine to enhance road features, to evaluate traffic safety and enhancement. GIS Based Traffic Accident Analysis A geographical information system (GIS) can be simply defined as a collection of hardware and software that is used to edit, analyze, and display geographical information stored in a spatial database. In recent years, many transportation departments and other related organizations have examined the feasibility of using GIS for transportation planning, systems management, and engineering applications. In some Countries, GIS is being used to plan transportation routes, manage pavement and bridge maintenance, and perform a variety of other traditional transportation-related functions. One area where GIS has yet to be extensively applied is in the analysis of crash data. Computerized crash analysis systems in which crash data, roadway inventory data, and traffic operations data can be merged are used in many countries to identify problem locations and assess the effectiveness of implemented countermeasures but unfortunately it is not applied in India. By integrating this traditional system with a GIS, which offers spatial referencing capabilities and graphical displays, a more effective crash analysis program can be realized. Benefits of using GIS in Traffic Accident Analysis GIS can easily represent accident and road accident based results using various tools like linear referencing, dynamic segmentation, and spatial analyst. Moreover query can be easily performed, enhanced by graphical representation. Road characteristics, demographic and socio-economic data to enhance highway safety analysis can also be integrated into the analysis. Other benefits of GIS technology are: -
METHODOLOGY The aim of the study was to delineate the multi-functionality and applicability of GIS to various Non-GIS users especially in the field of transportation by providing an efficient, effective and comprehensive system for accident analysis. The various points that were taken into consideration while designing the system were
Study Area The study area selected for implementing the methodology involved is a segment of NH-58 stretching from Delhi to Rishikesh covering a distance of 213 kms. The road is highly congested leading to a high rate of accidents. Delhi being the National Capital is visited by different genre of people. Moreover the road links different Tourist spots like Haridwar and Rishikesh, which are very popular amongst Indians as well as people coming from abroad are prone to a lot of accidents. Due to high amount of traffic and faulty geometric characteristics of the road, large amount of accidents take place on this Highway. Besides this, during festivals like Maha Kumbh and Ardh Kumbh unmanageable numbers of pilgrims are attracted to this place leading to huge traffic chaos. During rainy season, devotees of Lord Shiva collect water from Haridwar and walk down to Pura Mahedeo near Meerut and almost for a period of one month the highway is hardly available for traffic. Thus attracting our study on this stretch of NH-58.  Fig showing Study area of NH-58  METHODOLOGY FLOW CHART System Design
Data Collection Geometrical Characteristics Keeping in view the objectives of the study, necessary information about the stretch was collected for every 200 meters. Parameters involved were
This information was collected by field survey. Accident data The source of accident data was Traffic Police, Insurance agencies, hospitals, etc. The data for the last ten years was collected including the following parameters taking Delhi-0 as the Bench Mark.
The data so obtained was compiled in excel sheets which were then converted to DBF format to make them compatible with the GIS environment. Registration of the site map Registration is a process of locating all the spatial data with respect to a common frame of reference. This registration was carried out using ERDAS 8.5 software. In registration the projection parameters are also defined, so that the image gets converted into a Geo-referenced image. In this method at least four ground control points were required for registration. Digitization After registration on screen digitization in ARC-GIS 8.2 was carried out for the road feature layer. Theme Generation The different themes used in the project were
The road layer consisted of small segments 200m in length which were plotted using the split and divide option in ARC-GIS. The attributes related to these segments were attached to the layer using Join option after creating a common field so as to provide the system with flexibility for changes in the future as well. Moreover the lengths of the different segments were checked in ERDAS. However the accident layer was created as a route event after making the road layer as the route using the Create Route wizard in Arc GIS Toolbox. After doing so the route events were added to the route using the add route events tool. Adding a route identifier field in the EXCEL.dbf file created, which matched which the route identifier field of the route, did this. Thus making the system dynamic in nature as any amendment made at any point of time could be easily displayed and incorporated easily later on. This way of attaching data made the study dynamic that is changes could be incorporated at any point automatically in the system as well. ACCIDENT DATA ANALYSIS This section explains the spatial analyses performed in GIS on crash data, the results were displayed on maps. Multi-year crash data for the stretch was mapped and then spatially compared to yield the following results. The analysis was done in the following manner.
· Accident growth trends according to yearly variations Analysis of accident data can be easily made through the attribute table itself. The accidents from 1998 to 2003 were recorded and graphs were plotted between different years as well their fatalities. It becomes very clear that the number of accident per year have increased even though the fatalities have remained more or less the same, showing the severity of problem.  · Accident growth trends according to monthly variations The following figures show the method to calculate the monthly variations in the accidents. Simply a query as MON= Jan is to be given in a particular year and the results are displayed. A sample query is generated in the year 2003 layer  Thus indicating that the months of May, June, September and October have highest accident in rates in the year 2003. · Accident growth trends according to daily variations  The analysis shows the increased number of accidents on Mondays followed by Friday and Sunday. · Accident growth trends according to time variations An exhaustive analysis about the time of accidents can also be performed using the system that is the number of accidents in the morning and in the evening so as to ascertain the whether the accident took place due to poor lighting conditions, visibility conditions or due to faulty geometric design. Not only does the number of accident in each type important but also the number of fatalities of each type is also important. For day time accidents query like Time >=6 and Time <=18 was performed For night time accidents query Time >=18 or Time<=6 was performed on the Selection by Attributes toolbox as shown. More over the number of accident in the different time slots can also be seen using statistics tool in ARC-GIS.  Graph showing the number of fatal accidents in morning hours.  Graph showing the variation of fatal accidents in year 2003 in different time slots From the above graph it becomes clear that maximum fatalities occurred in the time slot of 12.5 to 16.5 that is in the afternoon and then at night between 22.00 to 24.00. Moreover analysis yielded information like the toll of accidents during daytime was higher as compared to the toll at night. But however, the number of fatalities at night were more (15 out of the 25 accidents) that took place in comparison with the 19 out of 46 that took place in the morning. · Accident growth trends according to comparative vehicle type Further the accidents were analyzed according to the vehicle type that is heavy vehicle light vehicle. Selections were made and the graphs were plotted between the heavy and light vehicle accidents along with their time slot.  Graph showing time variation in the number of heavy vehicle accidents The number of heavy vehicle accidents exceeds the number of light vehicle accidents (37-26). But however there is a large variation in the timings of these accidents like the heavy vehicle accidents mainly took place in the time slot of 7.00 to 9.00 A.M and 13.00 to 15.00. Whereas the timings of light vehicle accidents vary from 7.00 to 11.00 A.M. and 17.00 to 19.00 thus conforming to the real scenario that accidents are more of lighter vehicles during the starting and ending of working hours.  Graph showing the number of dead in the case of heavy vehicle accidents However the number of dead that is fatalities involving one dead were more in the case of light vehicle accidents (15) while the total fatalities were more in the case of heavy vehicle accidents (18). · Accident growth trends according to the number of vehicles involved in the , accident Depending on whether the accident was one vehicle or two vehicle query can be generated like one vehicle= one vehicle/ two vehicle in the Selection by Attributes Table. However the number of people dead in the each type of accidents can also be ascertained by again using the Selection By Attribute tool of ARC-GIS.  Graph showing number of dead in one-vehicle accidents The analysis showed that two vehicle accidents were more, i.e. 47 in comparison to one-vehicle accidents, which were 21. Moreover, the number of fatalities in two vehicle accidents where 19+3 while one vehicle accidents where 12. · Identification of Accident Prone Location (i) Statistical Method Accident-prone locations can be easily identified using ARC-GIS by the use of Raster calculator. Firstly the important parameters involved are identified as in our case it was realized that shoulder width played a crucial role in accident analysis. Second field was spot speed. So after identification of these parameters these fields are converted to raster images using Spatial Analyst tool, so that calculations could be performed. Based on the sample model Fatality rates = .00014*spot speed-0.328*shoulder width + constants Accident-prone locations were identified as shown. Following are the details of the accident-prone locations The accident-prone locations identified by the study were
All these location were very congested along with a low value of shoulder width and high spot speeds. (ii) Other methods An accident occurs from a complex interaction of vehicle, road, road user and/or environment factors and hence considered a multi-factor event whereas Accident Black Spot is a location, where accident frequency is high. The following four criteria have been adopted for ranking based on accident data:
Here "Accident Rate" and "Accident Fatality Rate" have usual meaning and denote the average accident number and fatalities respectively per km per year. "Accident Severity Index" denotes average accident weightage points per km per year, based on following weightage points assigned to the severity of each accident (as recommended by TRRL, UK in Interim Guide on "Identifying, Prioritising and Treating hazardous Locations on Roads in Malaysia", 1995). ,br> In police records, only fatal and serious injury accidents are reported whereas minor injury and non-injury accidents are hardly reported. To account for such a non-availability of reliable data on minor injury and non-injury accidents, their weightage has been taken as 1.0. The critically of each section for various configurations was determined based on threshold values for AR, AFR, ASI and Number of accidents per 1,00,000 veh. Km. These threshold values represent critical accident rate at 95% confidence level and have been taken as average accident rate (based on last 10 years accident data) plus 1.5 times of standard deviation. (X + 1.5 s) The accident locations where any one more than one of four selected parameters (AR,AFR,ASI and Number of accidents per 1,00,000 veh. km) exceed their threshold values have been considered as Accident Black Spots.  Map showing the black spots CONCLUSION NH-58 serves as a lifeline to the hilly area of Uttaranchal. The accident trends on the stretches show a significant impact of traffic as well as geometric parameters. The study was carried out on the Delhi Rishikesh road section and the analysis yielded the following conclusions:
Page 1 of 1 |