Why is Higher Education Classroom Scheduling So Constrained?

The use of higher education scheduling software deals with an extraordinary amount of resource planning constraints in any university campus. 

Academic resource planning and scheduling for higher education is virtually juggling several balls in the air, academically, financially and logistically.

At the beginning of each semester, faculties and administrators must deal with the angry face of both students and faculty members with these complaints:

• “I couldn’t enroll in this compulsory class.
• “This isn’t the classroom or time slot I requested.”
• “I’ve got two consecutive classes thirty minutes away from each other.”

These complaints ricochet from your latest academic planning session. Yes, you did everything in your power to make all ends meet. But apparently, it’s never enough…

Why is academic scheduling so hard?

Psychologist, Ann K. Newman reflects on the Chronicle of Higher Education that:

“Stretching the course schedule more evenly throughout the day and week, and balancing the classroom pool to class enrollments, can often result in found space that can be repurposed to meet other pressing needs (…). Many institutions, however, do not know how much space they have or how they are using it.”

Karen E. Hinton, from the Society for College and University Planning, has advised in academic planning for several higher education institutions in the United States. She has pointed out particular cases where an institution...

 “...could not decide about the number of classrooms it would need over the period of its new master plan. Instead, it relied on the estimates of its academic department chairs, who based their requests on current experience with classroom scheduling problems.”

After they had finished their assessment, they found out that the process had “overestimated the need by a significant amount based on realistic data analysis.”

What constraints our academic schedule?

M.A. Saleh Elmohamed, Geoffrey Fox and Paul Coddington, from Syracuse University, have compared techniques for academic course scheduling, and divided time constraints into three, depending on the costs for the institution:

As those “who physically cannot be violated, such as events that cannot overlap in time”:

• - Classes taught by the same professor.
• - Classes held in the same room.
• - A class and a recitation or a lab of the same class.
• - Space or room constraints. A class that cannot be assigned to a particular room unless its capacity is greater or equal than the class enrollment, while particular types of laboratories require specific types of rooms.

They consider time and space conflicts that cannot be physically violated, but can be avoided by making adjustments to “the specification of the problem,” such as the student preferences. Universities need, for instance, to address:

• - Time conflicts for classes with students in common.
• - Eligibility criteria for the class.

They are “preferences that do not deal with time conflicts” and have a lower cost associated.

• - A balanced spread schedules over the week.
• - Contiguous time slots.
• - Balance enrollment in multi-section classes.
• - Lunch and break times.
• - Teachers’ timing requirements.
• - The distance between a room the class is assigned and the building housing the department.

Michael Carter and Gilbert Laporte, from the University of Toronto, add an additional layer of constraints, as schedules are intertwined into:

• - Courses, as subjects “taught one or more times a week during part of the year. The same course may be taught in one or multiple sections, meaning that it can be repeated by different teachers during the week”, the authors define.
• - Classes, as “a group of students taking an identical set of courses and typically remaining together throughout the week”.
• Programs, as a “set of required courses and a set of elective courses to be taken by a student wishing to obtain a given degree.

Given these constraints, it’s no wonder that many universities (and a substantial number of researchers) have searched for a solution in mathematics.

Back in 1986, Suleiman K. Kassicieh, Donald K. Burleson, and Rodrigo J. Lievano, from the University of New Mexico, wrote about the complicated scenario of classroom allocation, looking for a decision support system. They found conflicting interests:

• - University administrators looked for more effective and efficient resource utilization.
• - Heads of departments had a varied set of goals, and had to balance “often-conflicting administrative, faculty, and student perspectives.”

They concluded that this was such a time constrained task that it was difficult to perform without the aid of a computer, as “even with a modest number of objects and destinations, the search space for feasible solutions is very large.”

Academics John J. Dinkel, John Mote, and M. A. Venkataramanan recall that in 1985 manual scheduling at the College of Business Administration of Texas A&M University was carried out in such a way that, in the end, faculties had to “bargain” unscheduled class sections, rooms and time slots.

They used a computerized network based decision support system to deal with scheduling, to allow decision makers “to maintain control of the process," and improved schedules by changing priorities and preferences, easily representing alternative solutions. They improved room utilization, reduced unassigned courses and streamlined solution models to time period shifts. At the same time, they reduced the amount of time required to produce the schedule. 

If this looked amazing, imagine what we’re doing with scheduling software over 30 years later, at the age of artificial intelligence and the Internet of things.

We can apply machine learning, where software is used to learn patterns and knowledge and apply solutions to decision-making problems.

• - What are the most efficient classroom combinations?

• - Can I prevent scheduling conflicts from both students and teachers?

• - How can my institution make the most of their classroom and free time?

• - Could I possibly plan these before the semester is over?

What are your main scheduling problems? How do you expect new technologies can help you address them?