Dynamic Traffic Light Control: A Novel Approach for Congestion Mitigation and Traffic Optimization

Abstract

In cities, traffic congestion is a major
problem that negatively affects both the quality of life
and the effectiveness of transportation. To tackle this
difficulty, we introduce a unique method for dynamic
traffic signal regulation that takes waiting time and
vehicle density into account. Our approach makes
use of cutting-edge technology including Raspberry
Pi for hardware integration, YOLOv8 and CUDA for
vehicle detection, and SUMO for traffic simulation.
Using SUMO, we first create a 2D map of a traffic
scenario as input for further analysis. Next, we utilize
YOLOv8 and CUDA to identify automobiles inside
the simulated setting, allowing us to compute vehicle
density by taking into account the number of vehicles
and the length of the road. Congestion status is
evaluated using a predetermined threshold value that
was established by trial and error. To efficiently
regulate traffic flow, traffic lights are dynamically
turned green when the density is over the threshold,
signaling congestion. However, there is a 30-second
waiting period. Our algorithm’s output is smoothly
connected with hardware through the use of a
Raspberry Pi, which manages traffic light modules to
instantly make modifications. This integration makes
it easier to regulate traffic signals effectively, which
guarantees optimal traffic flow and reduces
congestion.Our test findings show how well our
method works in different lanes to optimize traffic
flow and lessen congestion. Our technology presents a
prospective option for enhancing urban traffic
management and transportation efficiency by
dynamically modifying traffic lights based on
real-time data