Research from the University of Michigan-Dearborn may soon provide others with the ability to build robotic sensors that identify and translate information based on scent.
“The (e-nose) is not completely new, but we’re not building a specific sensor,” says Nevrus Kaja, who recently graduated from U-M Dearborn with a bachelor’s in computer engineering. “We’re concentrating on building the architecture.”
By building a standard algorithm, Kaja says his work could provide hospitals, universities, and companies with the foundation needed to build their own products for whatever applications they need. The military may use it to locate explosives, whereas a medical center or hospital could potentially detect diseases based on a person’s aroma, he says.
Kaja, who will continue his research while pursing a Ph.D. at the Dearborn campus, says it was his sister’s dog’s different reactions to a person without seeing them that first inspired him to pursue the project.
“(The dog is) in the basement, and when my sister gets home, he starts barking and jumping (without seeing her),” says Kaja, laughing. “But when I get home, since he doesn’t like me, he doesn’t do anything.”
Once complete, the architecture will be available to the public, says Kaja, who has presented on the topic at conferences including the International Symposium on Engineering and Natural Sciences in China and the Richard Tapia Conference in Washington, D.C.
“This will be open source,” Kaja says. “I’m not looking to get a patent or to make money off of this. It’s for the greater good.”
