Optimizing Academic Scheduling: An Automated Timetabling System for the University of the Philippines Manila

Abstract

This study presents an automated timetabling system developed to address the inefficiencies of manual class scheduling at the University of the Philippines Manila (UPM), where diverse departmental constraints often result in conflicts and delays. The web-based system integrates a Genetic Algorithm (GA) with two local search optimization techniques—Great Deluge Algorithm (GDA) and Simulated Annealing (SA)—to construct and refine timetables that satisfy hard constraints (e.g., room capacity, faculty availability, conflict-free schedules) and optimize soft preferences (e.g., preferred rooms, accessibility). Users can upload CSV data, visualize schedules through dynamic calendar views, and export outputs, while a utilization scoring feature enables assessment of room efficiency. Built on a Django backend, the platform significantly reduces scheduling time and errors, demonstrating strong performance across complex scenarios and offering adaptability for broader academic use. By combining global and local search strategies, the system not only delivers highquality timetables tailored to institutional needs but also provides a scalable model for future integration in the university system.