Overview of Classroom Assignment and Optimization Process

The following is an overview of the classroom assignment and optimization process used by the Registrar’s Office to assign central classrooms each semester. Our goal is to be as transparent as possible about how we assign classrooms so that instructors and administrators have a better understanding of why we assign the rooms we do.

Which kinds of classes are included in the optimization process?

Historically, we have conducted the optimization process under the assumption that only lecture, seminar, and discussion sections require central classrooms. Other types of classes, such as laboratories or practicums, are generally scheduled in department-controlled space or do not need classrooms at all. While those assumptions are still true in most cases, we will now use the Room field in CLSS to determine which classes to include. To indicate that a class requires a central classroom, set the Room field to “General Assignment Room.” In addition, it is important to finalize room assignments or choose “Pending Department Room Assignment” for any classes that will be held in department-controlled space.

Step 1: Manual classroom assignments for all ADA accommodations

There is no way to automate assignments for ADA accommodations while also protecting instructors’ privacy, so these assignments are made manually by the Scheduling staff. Again, please have your faculty reach out to Ryan Bangham in HR to make any necessary ADA accommodations.

Step 2: Configuring the optimizer

The EMS Classroom Optimizer is a tool that uses an algorithm to quicky and efficiently assign classrooms to thousands of classes while respecting certain weighted criteria. Before we run the optimizer, we update any sections with room requirements specified in CLSS and make any updates to department building preferences. We then tell the optimizer how to weight different criteria when making room assignments using the following weights listed in the order in which they are ranked:

  1. Enrollment Capacity: Our highest priority is ensuring that we find classrooms for every class that needs one. We try to closely match enrollment capacity to room capacity to maximize seat utilization and to ensure that undersized classes in oversized classrooms do not displace larger classes.
  2. Room Requirements: These are instructional requirements reported to us by departments via CLSS. It is important that only requirements be listed in CLSS rather than preferences. Setting too many requirements can be risky; the optimizer will only assign classrooms that meet ALL requirements.
  3. Back-to-Back Instructors: This ensures that instructors are kept in the same classroom (or at least building) to the extent possible. If instructors teach back-to-back at peak times, this is not always feasible, particularly if one or more of the classes is large.
  4. Adherence to Standard Meeting Pattern: While deans may approve exceptions to the university’s standard meeting pattern policy, those classes that conform to the standard meeting patterns will be prioritized over non-standard classes in order to maximize classroom utilization.
  5. Departmental Preferences: Once the above criteria are met, the optimizer will try to assign classes to rooms only in the buildings or areas preferred by the department.

Step 3: Optimizing large classrooms

Because large classrooms – those with capacities of 100 or more – are so scarce, we run the optimizer for all large classes in a single batch. We do try to honor departmental preferences, but this is not possible for many large classes. For some departments, the buildings they identify as their top choices simply do not have classrooms large enough to accommodate their largest classes, so honoring those preferences is simply impossible.

Step 4: Optimizing in batches to honor departmental preferences

Rather than attempting to assign all classrooms in a single optimizer run, we run the optimizer in batches so that we can better match each department to its first, second, and third preferences. For example, if three departments have indicated that MCHU is their first choice for room assignments, we will do an optimizer run with only those departments and MCHU as the only building option. We will lock in any assignments the optimizer was able to make, and any sections that remain unassigned will be moved back into the pool of classes to be assigned.

Step 4a: Optimizing discussion sections

Because discussions almost always meet just once per week, we don’t include them in the optimization process until after the other class types have been optimized. We’ve found that including the discussions with the classes that meet more frequently has led to inefficient classroom assignments.

Step 5: Optimize all remaining unassigned classes

Once we’ve gone through the preferences building by building and department by department, any unassigned classes will be included in one final run of the optimizer to assign them to any available central classrooms. If there are still unassigned classes after that, we will attempt to manually assign them and, if there are no rooms available, we will notify departments.

Step 6: Publish classroom assignments, activate EMS Room Seek

Once the room assignment process is complete, we will publish the results so that room assignments are visible in EMS, CLSS, and Student Admin. EMS Room Seek will also become available for departments who wish to make any changes to their classroom assignments.