For every watt of electricity generated in the United States, 74 percent is consumed by buildings throughout the country. Philip Odonkor, assistant professor at the School of Systems and Enterprises, is researching how automated tools can help buildings use energy more efficiently and address climate change in the process.
In his “SSE Dean’s Virtual Seminar Series” presentation on “Making Dumb Homes Smarter,” he underscored the intriguing title by pointing out that whether it’s a luxury apartment in Hoboken, a cottage in the outskirts of London or a shanty town in Soweto, most homes simply don’t know how to use electricity. View the complete lecture here.
“If you were to leave your house and forgot to turn off the lights, would your house turn the lights off for you? Or would you have to wait until you came home at the end of the day to turn them off?” he asked, adding that the same question applies to air conditioning and heating systems.
The key to applying systems knowledge to create homes that use less electricity, Odonkor stated, requires understanding the behaviors of each person who lives in the house and who uses energy on a daily basis.
Can Systems Really Understand People?
“Understanding the human being is not an easy task,” said Odonkor, whose research has been featured in the Journal of Mechanical Design. “What we are trying to do in our lab is to create a home that understands people.”
Automation has been used for smart home devices, but Odonkor’s research probes deeply into decision-making – specifically, into the decisions that either save or waste energy within a home. Odonkor uses artificial intelligence to simulate the experiences and preferences of the human mind to make good decisions to save energy.
Decisions on Generating, Storing and Selling Electricity
Specifically, the research has led to an automated energy management system that can make decisions regarding when to generate electricity, store it or sell it back to the grid. He walked the online audience through a trial-and-error situation that used deep reinforcement learning to teach an automated system to make decisions that save as much energy as possible.
“This leads to a truly smart home,” said Odonkor, as he discussed the automated energy management tools that can help residential buildings optimize electricity. “Our system was able to save on the energy costs for a home. We’re making exciting progress.”
Odonkor, a native of Ghana, joined the Stevens faculty in 2019, after earning a doctorate in mechanical engineering at the University at Buffalo (SUNY), where his research focused on energy optimization in net-zero energy building clusters. Today he runs the Design Informatics Lab at Stevens, focusing his efforts on smart systems to combat climate change.
About the Seminar Series
The SSE Dean’s Virtual Seminar Series is created to help keep students informed on the advances of faculty research and the mission of the School of Systems and Enterprises as a leader in systems science and engineering. Lectures will be recorded and made available on Kaltura for those who are unable to attend the live Zoom sessions.
For a list of forthcoming presentations, visit the "SSE Dean's Virtual Seminar Series: The Future of Systems" event page.