Pengembangan Sistem Kendali Cerdas Alat Pemberi Isyarat Lalu Lintas (APILL) Berbasis Machine Learning

Abstract

In modern cities, population growth directly contributes to an increase in the number of vehicles, leading to significant traffic problems and a decline in road service quality and capacity. Conventional traffic light control systems (APILL) that rely on fixed-time scheduling often fail to adapt to the dynamic nature of traffic conditions, potentially exacerbating congestion. This study proposes an innovative approach to traffic management by utilizing the YOLO (You Only Look Once) object detection algorithm. By analyzing CCTV streaming data at intersections, the system dynamically assesses traffic density, identifies vehicle types, and adjusts signal timings in real-time. Leveraging YOLO's ability to perform fast and accurate object detection, the system can respond to traffic conditions in a timely manner. This approach integrates Artificial Intelligence (AI) and Machine Learning techniques to address the urgent need for adaptive traffic management strategies in urban areas. The primary goals of this solution are to reduce congestion, improve traffic flow, and minimize environmental impact. Therefore, the integration of YOLO technology with adaptive traffic signal control algorithms represents a strategic step toward addressing the complex challenges of urban traffic congestion