Application of Genetic Algorithm and Or-Tools for Cloud-Based Course Scheduling Optimization

Course scheduling in higher education institutions is a complex combinatorial optimization problem involving numerous constraints such as lecturer availability, room capacity, time slots, and course distribution across semesters. Manual scheduling practices often result in conflicts, inefficient resource utilization, and prolonged preparation time. This study proposes a hybrid course scheduling system that integrates a genetic algorithm (GA) and constraint programming using the CP-SAT solver from OR-Tools. The GA is employed in the first phase to generate optimal course sections based on student enrollment, lecturer workload, and capacity constraints. The best solution produced by the GA is then refined using CP-SAT to generate a conflict-free timetable that satisfies all hard constraints, including lecturer, room, and time conflicts, while also optimizing selected soft constraints. The proposed system is implemented as a web-based application deployed on Microsoft Azure, enabling scalability and accessibility. Experimental results using real academic data demonstrate that the hybrid approach successfully produces feasible schedules with zero conflicts and significantly reduces scheduling time compared to manual methods. The results confirm that the integration of GA and CP-SAT provides an effective and flexible solution for university course scheduling problems.