University course timetabling is a vital part of any university, directly impacting the engagement and overall learning experience of students (Castineiras et al.). The basic problem involves allocating activities to rooms and timeslots such that no student/lecturer/room is assigned to more than one activity in a given timeslot. The problem is known to be NP-complete which, in simple terms, means it is practically infeasible to solve by brute-force search of all possible allocations.

The problem is further complicated by individual timetabling constraints that vary across institutions, and even across departments within a single institution. In Munster Technological University (MTU) for example, there are various definitions of the problem being solved due to differences in student bodies and resources (e.g. School of Music versus Faculty of Engineering), requiring different software tools due to their specific needs. Many institutions also have room ownership constraints, and therefore cannot be solved in a centralised manner. These restrictions result in rolling over the previous timetable year-to-year, reactively adapting it due to changes in academic staff, student cohort sizes and room infrastructure. In a nutshell, this requires tackling the modelling of the problem as a minimal perturbation problem.

This minimal perturbation approach is also often driven by the short timeframe between all inputs being known and the start of academic year. In particular, the issue is that final cohort sizes are unknown until after Autumn repeat and leaving cert results are released. The latter not only affect allocation for 1st-year cohorts but also have a knock-on impact on the overall solution. However, minimising perturbation year-to-year may also have the untoward effect that, each year, the solution may move further from the initial “good” solution when the timetable was first built from scratch.

Finally, the requirements and preferences of the modern student have changed considerably over the past number of years, with digital enhancements such as hybrid learning being accelerated due to the COVID-19 pandemic, cost-of-living, etc. A separate but equally important concern is sustainability, and the impact of timetabling choices and assumptions as universities aim to move towards green campuses.

Given all these changes, now is an opportune time to revisit the university timetabling problem with fresh eyes and reflect on how these changes could be incorporated. With this in mind, we first sought the perspective of the students on their experiences and preferences with respect to timetabling via a survey with both closed and open -ended questions, with a focus on factors influencing engagement and attendance. Our findings showed that while there are certain commonalities in terms of obstacles to students such as commuting; there were also differences in preferences between science/engineering students and those of the business/humanities offerings in MTU. Further work with Cork School of Music (CSM) illustrated the challenge of generating university-wide timetabling policy. While students of the Science/Engineering/Business departments were mainly in favour of compact schedules (due to commuting, work, etc), students of Arts programmes such as those offered by CSM and those with specific equipment for students, prefer gaps between activities as they get to avail of dedicated on-site facilities.

Together our findings show that the decentralised approach (where departments timetabling somewhat independently), while suboptimal from a room-allocation perspective, is often necessary due to the significant variations in requirements of students across different departments. We discuss the impact this has on the AI tools available for generating solutions to the problem.

Castineiras, I., Grimes, D., & Ozturk, C. (2024, September). Developing a University Timetabling Assistant Tool to Address the Voice of Students. In International Conference on Lean Six Sigma for Higher Education Institutions (pp. 258-266). Cham: Springer Nature Switzerland.