TRAFFIC SAFETY ANALYSIS PLATFORM

Due to the increasing population in cities, traffic congestion is being experienced and this increased congestion leads to emergencies and increased accident rates. Smart transportation solutions offered by smart cities can shorten travel times, ensure and protect life and provide increased safety in traffic. Projects that will contribute to different smart cities components, which are determined under 16 different topic in the "2020-2023 National Smart Cities Strategy and Action Plan", such as environment, security, infrastructure, and governance are implemented using the geographical information systems under the smart transportation component. In this direction, the "Traffic Safety Analysis Platform" project is implemented to increase safety in transportation and to create a decision support mechanism for traffic safety by effectively using traffic data within the scope of the "2020-2023 National Smart Cities Strategy and Action Plan" and "Road Traffic Safety Strategy and Action Plan".

Through this project, data that have an impact on road safety will be collected in a single center and this big data will be analyzed. Thanks to the artificial intelligence algorithm that will be created with the analyzed data, it will be possible to analyze the High Accident Intensity Zone and Accident Black Spot. The created platform will be web-based and suitable for the use of Municipal Corporation throughout the country. Platform will contribute to traffic safety and decision support system country-wide by successful completion. Within the scope of the 2020-2023 National Smart Cities Strategy and Action Plan, pilot implementation studies for smart cities were started in order to increase the smart city technical capacity of Municipal Corporation and to accelerate smart city transformations.

CONTENT

A web platform based on big data, which provides decision support for urban road network safety analysis and planning by using the information related to traffic accidents and its submechanisms is developed. Thus, time-spatial analysis of Accident Black Spots can be generated with geographical information systems based thematic maps and a flexible analysis structure that includes corridor and micro-level can be created through big data analysis.

METHODOLOGIES USED IN THE PLATFORM

Accident Black Spots: The distribution showing a section of the road where traffic accidents are dense is often referred as an accident-prone point or region, and these points are known as Accident Black Spots. The series of analysis made on the platform will cover different Accident Black Spot calculation methods and will be obtained with the appropriate method according to the available data.Accident Black Spots can be generated from the module on the platform for the selected method, region and time period.

The Accident Black Spot calculation methods used in the Traffic Safety Analysis Platform are as follows:

1. Number of Accident Method: In this method, which is preferred in regions where the traffic volume is not excessively heavy; Accident Black Spots are determined and ranked based on the number of accidents.
2. Accident Rate (Accident Repetition Rate) Method: In this method, which is preferred in regions where the traffic volume is variable; The number of accidents per unit vehicle passing through a certain road section during a certain time interval is calculated.
3. Accident Severity Method: While grading or ranking the regions with excessive number of accidents using the accident severity method, the vehicles and casualties involved in the accident are also evaluated. In order to determine the severity index of the method, the accident severity value is obtained by including the number of accidents with fatality, injury and material damage in different weights. (Accident severity = 9 * Accident with Fatality + 3 * Accident with Injury + Accident with Material Damage)
4. Number-Rate Method: The method takes into account the lengths of the zones where fatal accidents occur and the ratios with the annual average daily traffic in the variables of the user-determined time period, threshold value, and zone length
5. Ratio-Quality Control Method: The method takes into account the zone lengths, traffic volume, and types of accidents in the variables of user-determined time period, threshold value, and zone length.
6. Artificial Intelligence (AI) and Machine Learning: AI is used to analyse traffic data such as traffic accident, factors affecting the accident and accident frequency analysis with machine learning. In the development of the Traffic Safety Analysis Platform, the space-density based DBSCAN unsupervised machine learning clustering algorithm, which also includes a kind of screening method, was used. The DBSCAN method is preferred because the algorithm works, is fast and flexible.

DATA

The data used in the platform consists of the accidents with fatality, injury and material damage and obtained from the Turkish Ministry of Interior IVME Project, which are collected through official accident reports of the General Directorate of Security and the General Directorate of Gendarme. Statistical analysis of the data is carried out with data cleansing and editing methods. Thus; mean, standard deviation, skewness, kurtosis and total values of the data were found. As a result of these analyses, the most common data types in accidents with fatality and injury were obtained.

Data can be filtered by parameters such as Accident Severity, Accident Types, Weather Conditions, Causes of Accident, Location Information, Road Features, Time, and displayed on the map with options such as; heat map, point representation and clustering methods.

Through Accident Black Spots and High Accident Intensity Zones which are generated by choosing one of the traditional traffic engineering or artificial intelligence methods in a region and period, factors causing the accident, areas that need improvement and the possible situations that may encounter in certain conditions can be determined and analysed. In the Traffic Safety Analysis Platform, the statistics obtained from the data of the detected Accident Black Spots and High Accident Intensity Zones can be visualised and reported within the panels on the platform.

TARGET USER GROUP

It is expected that the relevant institutions use the platform in order to effectively identify problematic areas detected by Accident Black Spots and High Accident Intensity Zones within their responsibility areas and to take appropriate and effective action to contribute to cities. At the same time, analysis and reporting features on the platform can be used for periodic monitoring and preparation phase of transportation projects and plans which will contribute to the public interest.

Local Authorities

• Provincial Special Administrations 
• Metropolitan Municipalities
  
• Provincial Municipalities
• District Municipalities

Relevant Public Institutions Responsible for Traffic Safety

• Republic of Türkiye Ministry of Interior Department of Strategy Development
• General Directorate of Security
• General Directorate of Gendarme
• General Directorate of Highways

EXPECTED RESULTS

Expected results across Türkiye with the Traffic Safety Analysis Platform are as follows;

• Identification of accident black spots for urban transportation at the national scale
  
• Automatic generation of Accident Black Spots by different methods
• Monitoring of traffic accidents in comparison to previous years
• Taking precautions to prevent traffic accidents in a more detailed way
• Using artificial intelligence screening and clustering methods in accident analysis
• Monitoring the improvements made for accident points over the years
• Determining the parameters that are crucial in the accidents as a result of the analyses

FUTURE PROJECTION

With the effective and integrative use of the Traffic Safety Analysis Platform and the "Urban Mobility Analysis Software (SUMO Türkiye Adaptation)" developed by the Smart Cities Department of the General Directorate of Geographical Information Systems of our Ministry within the scope of the "2020-2023 National Smart Cities Strategy and Action Plan", a valuable contribution can be made to transportation planning and projects of different public institutions and organizations.

With the use of the platform by academicians working in the field of transportation, it is possible to contribute to the studies in the field of transportation in our country with the calculation methods, analyses and visualised reports available in the "Traffic Safety Analysis Platform".