Residential Heat Loss Identification and Analysis Through Convolutional Neural Networks Without Utilization of Thermal Imaging

Energy efficiency for homes has been of tremendous interests in recent years as the overall energy consumption of households takes up a significant portion of the total energy consumption. Energy efficiency is key for homeowners to save. Therefore, the ability to clearly identify heat losses in a household is vital in the creation of unique solutions to homeowners to curb the identified heat losses. This project aims to identify the heat losses of residential homes using computer algorithms so an informed decision can be made, by the utility or their customers, of energy saving upgrades to the investigated households. Furthermore, the project will be integrated into the Saint John Energy customer portal so that existing customers can check their energy efficiency scores of their houses.

Faculty Supervisor:

Bo Cao;Liuchen Chang

Student:

Partner:

Saint John Energy

Discipline:

Engineering

Sector:

Utilities

University:

University of New Brunswick

Program: