Boston University

08/06/2026 | News release | Distributed by Public on 08/06/2026 12:03

Student Creates AI-Powered Tool to Analyze Classroom Space Usage on Charles River Campus

Student Creates AI-Powered Tool to Analyze Classroom Space Usage on Charles River Campus

Joey Russoniello (CDS'26) collaborated with IS&T as part of CDS' clinics program on his senior capstone project

CLASS SELECTION

Student Creates AI-Powered Tool to Analyze Classroom Space Usage on Charles River Campus

Joey Russoniello (CDS'26) collaborated with IS&T as part of CDS' clinics program on his senior capstone project

August 6, 2026
Twitter Facebook

Here's a problem almost every college student has experienced: a seminar with a dozen classmates meets in a huge lecture hall, while a full 100-level science course is squeezed into a room. Or the daily sprint required when back-to-back classes for the same major are assigned to opposite ends of the campus.

Every semester, Boston University faces the challenge of fitting nearly 4,500 classes into hundreds of classrooms each semester. Most students expect scheduling quirks on a large urban campus like BU. But for recent graduate Joey Russoniello, those quirks represented an intricate engineering puzzle waiting to be solved.

Russoniello (CDS'26) spent the 2025-2026 academic year helping to design a new optimization system that is already improving how BU assigns classrooms across the Charles River Campus.

The program, named the Classroom Utilization Project, was developed through one of the Faculty of Computing & Data Sciences' clinics, launched in collaboration with Information Services & Technology. The program engages undergraduates in an internal consulting "clinic" addressing BU's most pressing data-related technical needs. It's based on a "work-integrated learning" model, an educational approach that combines classroom study with practical workplace experience, enabling students to apply their knowledge and prepare for the job market.

Under the guidance of Lauren Wheelock, a clinical assistant professor in the Faculty of Computing & Data Sciences, Russoniello built a tool that analyzes enrollment data, classroom capacity, and campus geography and recommends smarter classroom assignments. The system automatically matches classes to rooms, maximizing the use of available space while minimizing the distance between departments' home buildings and the classrooms they use. It also prevents scheduling conflicts by ensuring rooms are never double-booked.

In the web application, administrators can upload course and room data and view proposed assignments. The new optimization tool could eventually work alongside 25Live, BU's existing room scheduling system.

Russoniello says he was all too familiar with the issue at hand. "I think this problem speaks to every single BU student," he says. "I was a CDS student, and I took maybe two or three courses in my home building during my time at BU. Sometimes students have classes that are stretched a mile apart. It just seemed random."

When Russoniello, who had been a teaching assistant for Wheelock, asked her about research opportunities, she pointed him to the CDS clinic.

Photo by Cydney Scott

Other projects to emerge from the clinic include an app that shows where open parking spots are available across campus and a database that flags publicly available SEC filings to help BU better track its alumni.

Pete Groustra, assistant vice president of projects, programs, and governance for IS&T, led his department's efforts to work with Russoniello throughout the project. He says IS&T has made a concerted effort over the past year-especially as AI tools have become more prevalent-to collaborate more closely with faculty and staff across BU. Students, he says, bring interesting ideas and a different perspective to a problem.

"We are talking to many faculty in CDS, Questrom, and the College of Engineering on how to better coordinate the student projects going forward," Groustra says. "We can help students with setting a defined scope, working with project sponsors, and establishing project outcomes. Is the project strictly an academic exercise to learn, or will we be able to apply some of this technology? We try to set that expectation up front."

Facing the Challenge

Russoniello's challenge was translating thousands of scheduling decisions into mathematics.

Instead of evaluating classroom assignments one course at a time, the team-which included front- and back-end support from Zach Gentile (CDS'26)-built an optimization model that simultaneously weighed every possible classroom assignment while accounting for hundreds of scheduling constraints.

"This solution aims to fix this scheduling issue by creating a giant math problem-minimizing the number of wasted seats, the distance from a preferred set of classrooms, all of these things," Russoniello explains. "Essentially, we just translate these semantic concepts into math. So a class can't be allocated into a room that's too small for it. Two classes can't be in the same space at the same time."

Once those constraints are satisfied, the model optimizes for "bonuses," like minimizing unused seats and keeping classes closer to their departments' home buildings.

"Just about everybody I talked to wants more classes in their school," Russoniello says. "I think it would build a better sense of community too. You'd see your classmates more often as you're walking from place to place, as opposed to being randomly sprawled out." What's more, when a department's courses are clustered near its home building, students can get to class on time, and office hours are easier to hold.

Because the optimization model is flexible, additional priorities can be incorporated over time, such as classrooms with amenities like air-conditioning and accommodating dynamic class schedules that change from week to week, Russoniello says.

A Big Success

Before Russoniello tackled the problem, he says, there was no way to optimize the schedule, and any optimizations or adjustments were made manually by the registrar. His first iteration of the program resulted in a nine-hour solve; his final version can complete the same calculations for roughly 4,500 courses in just 90 seconds.

His model also reduced the average number of wasted seats per room by approximately 92 percent. The average walking distance between a department's home building and its classroom location dropped from 0.15 miles to 0.05 miles.

Russoniello began the project as his CDS capstone, earning course credit before continuing the work as a paid CDS intern supported by the CDS Launchpad fund during his second semester. He believes this optimization system is the first of its kind and could one day be used at other universities. He is excited that the project is open source and that the code is public on his GitHub profile.

Russoniello is now a data engineer for a real estate company in California. But the Classroom Utilization Project will continue this fall with a new team of students under Wheelock's guidance.

"This issue affects every single BU student taking classes, all professors," Russoniello says. "It touches everyone."

Explore Related Topics:

  • Share this story

Share

Student Creates AI-Powered Tool to Analyze Classroom Space Usage on Charles River Campus

Copy URL: Copy

Latest from BU Today

Boston University published this content on August 06, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on August 06, 2026 at 18:04 UTC. If you believe the information included in the content is inaccurate or outdated and requires editing or removal, please contact us at [email protected]