TRAFFIC OFFICER PLACEMENT RECOMMENDATIONS BASED ON ROAD CONGESTION LEVEL CLUSTERING IN BANDUNG CITY

• Main Author: Munggaran, Akmal
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/84971
• Institution: Institut Teknologi Bandung

Traffic congestion and accidents are significant problems in Bandung City, which has the highest population density in West Java Province. The high population density triggers an increase in the number of vehicles, which results in severe congestion on various roads especially at intersections. One of the efforts to reduce this congestion is to conduct surveillance. Police posts are provided at each intersection to assist with surveillance efforts. However, the large number of intersections in Bandung City cannot be allocated by all security officers. As a result, some police posts become unoccupied and cannot help regulate the road when congestion occurs. This study aims to address the problem by designing recommendations for the placement of traffic security officers using clustering and data mining approaches. This research identifies and maps road groups based on congestion characteristics, designs recommendations for officer placement at the right police post point, and determines the distribution of officers to roads that will or are experiencing congestion. The method used in this research is the CRISP-DM method. This research resulted in 3 groups of roads, namely light traffic, medium traffic, and high traffic. The results of cluster evaluation using the Davies-Bouldin Index, Calinski Harabasz Index, and Silhouette Coefficient show fairly accurate results with values of 0.9727, 156.9300, and 0.4339, respectively. In addition, this study produced 18 recommendations for new police station points on roads that have high traffic labels. Thus, there are a total of 53 police station points that can be utilized for distribution to roads labeled as light traffic and medium traffic. This distribution is based on measurements made by tracing each connected road until finding the nearest police station point.