Detecting HVAC Underperformance in Residential Buildings
The SOURCE at Syracuse University · Sustainable South Campus — Spring 2026
Featured in Syracuse University Magazine, Fall 2025 See the feature ↓
8.6%
Heating issue rate in Unit A-8
3.6%
Unit A-7, same building and heating system
2.4×
Performance gap between the two units
3
Condition flagging system built in Python / pandas
This study examines whether existing building sensor data can detect poor heating performance before residents file complaints. Using indoor temperature, occupancy, and outdoor weather data from two Syracuse University dormitory units, a three-condition flagging system identified heating issues — revealing that Unit A-8 had a heating issue rate of 8.6% compared to 3.6% in Unit A-7, despite sharing the same building and heating system.
In the Field
In the Press
Retrofitting for the Future
“[The project] connects data, sensors and real building systems in a way that matters.”
— Diamone Cromer ’28, quoted in the article
1 / 2
Pages courtesy of Syracuse University Magazine, Fall 2025. Article by Jay Cox.
Full Research Paper
1 / 6
Click any page to open it full size.
Download full paper ↗