Online MADS students guide University’s energy management
Their work in a Master of Applied Data Science capstone course helps managers make data-driven decisions to ensure efficiency.
Though these online master’s degree students never set foot on Carolina’s campus for their classes, their work in a capstone project will have a major impact on the energy and water use in University buildings.
Four students completed their capstone project in the Master of Applied Data Science program with UNC energy management. This Facilities Services unit provides dashboards to help energy managers make data-driven decisions to ensure electricity, chilled water and steam are used efficiently in the campus buildings.
Alexandria Cyr ’26 (MADS), Ross Matheny ’26 (MADS), Andrew Plattel ’26 (MADS) and William Reed ’26 (MADS) worked together over one semester to develop a full Medallion Architecture data pipeline in Microsoft Fabric. They produced forecast models for each utility type and displayed both the billing and forecasting data in a Power Business Intelligence dashboard.
“To me, it’s a win-win” said Vee Veerabhadrappa, interim executive director of engineering services and director of energy management. “We are serving the core mission of the University, which is to educate students and to give them exposure to real data and an opportunity to work with professionals in Facilities, while developing tools that can help our energy managers.”
“It was also really nice working with real data that isn’t just from the classroom, or what we find in a dataset online,” said Plattel, implementation associate at Heliocampus. “This is data that they’ve gathered over the last 10 years.”
These four students weren’t just working with real data but also a real problem. “A lot of times in school, you have these kind of textbook problems that are nice and wrapped up,” said Rick Marks, professor of the practice and creator of the MADS capstone course. “That’s not anything like the real world, so we want to give them a little bit of ability to use their knowledge from the classwork but on something more real and discover that things are messier in the real world.”
Building on the work of a capstone project from the previous semester, students received the historical billing data in a collection of spreadsheets and performed exploratory data analysis to understand utility coverage and seasonal patterns as well as potential gaps and anomalies in the data. They then formulated a reliable data processing framework to test the accuracy of multiple forecasting models for each utility. By leveraging Microsoft Fabric, the data processing, modeling, evaluation and data visualizations provide the energy management unit access to relevant insights from the analysis.
“Bringing awareness to what we were determining as anomalies is tremendously beneficial,” said Reed, senior financial systems analyst at Luck Companies. “Where being able to look remarkably quickly at a building and being able to say, ‘That’s an anomaly,’ I think brings tremendous value.”
It’s not just about the value that the students bring to clients but also the value that they derive from this co-curricular experience. “This class really allows them to take ownership on a full project, from the communication side, from the project manager side,” said Rei Sanchez-Arias, associate professor and MADS faculty director. “I think that’s a good experience.”
That experience is often balanced with other coursework and full-time jobs. “I understand that the students often have several other courses they have to work on, in addition to personal lives that they need to balance while dedicating time and efforts to the capstone project,” Veerabhadrappa said. “But they really invested their time and effort on the project, came back with really good questions, ideas and, at the end of it all, gave us a nice dashboard.”
