IMPROVING KARACHI’S BRT SCHEDULE WITH AI AND MACHINE LEARNING: A DATA-DRIVEN APPROACH

Abstract

Karachi’s Green and Orange Line BRT systems play a crucial role in the city’s transportation network, yet challenges related to scheduling inefficiencies, punctuality, and uneven passenger flow persist. This study applies Artificial Intelligence (AI) and Machine Learning (ML) models to optimize bus scheduling dynamically based on real-time ridership trends and operational data. By integrating historical ridership data with simulated passenger distribution models, we enhance predictive accuracy in demand forecasting, leading to improved schedule adherence, passenger satisfaction, and operational cost-efficiency. The study further develops a supervised learning model for demand forecasting and reinforcement learning for adaptive scheduling, ensuring optimal fleet distribution, and mitigating peak-hour congestion [1], [2].