|
By Dan Rahn
drahn@uga.edu
 From the beginning, Takoi Hamrita (pronounced TAKwa hamREEta) has taken Robert Frost to the extreme. Whenever two roads have diverged anywhere, she has taken the one less traveled.
It’s in her nature.
That’s why she earned a Ph.D. in electrical engineering from Georgia Tech. “I wanted to get into something that was not traditionally a field for women,” she says. “I wanted to conquer something.”
That’s also why she ended up in UGA’s College of Agricultural and Environmental Sciences. “Most researchers in electrical engineering have a narrow focus. I didn’t want to spend my whole life developing a chip,” Hamrita says.
“I wanted to work in an interdisciplinary area, in something that could potentially affect the whole world,” she says. “So when I got the chance to work here in biological and agricultural engineering, I jumped at it.”
In the classroom, Hamrita again took the most challenging route, tackling teaching the way a talented researcher goes about solving a problem. “Electrical/electronic systems are very important for agriculture as it becomes increasingly automated,” she explains. “I wanted to find out how to teach EES topics to ag engineering students in a way that’s suitable for them.”
In the limited time available, it’s virtually impossible to teach topics like Hamrita’s class in advanced microcontrollers from the bottom up, the way such classes are almost always taught. The temptation is to simply water down the subject matter. Naturally, that option didn’t appeal to Hamrita.
“I worked with local industries to incorporate real-world engineering problems,” she says. “This approach eliminates the parts that aren’t relevant, and the students still get concentrated learning in the areas they do need.” The National Science Foundation awarded her a $100,000 grant to develop a national teaching model using her novel approach.
Hamrita came to the United States on a scholarship she earned as one of the top handful of students in her native Tunisia. In high school, she had wanted to be a medical doctor. Now, she “operates” on chickens.
To help Georgia’s huge poultry industry better control the climate in the state’s broiler houses, Hamrita uses an ingenious combination of sensors and artificial intelligence to get chickens to “talk.”
She implants tiny sensors inside three-week-old chicks, enabling them to “tell” a complex computer controller their deep-body temperature and, eventually, their respiration and heart rate.
With neural networks--digital simulations of human neurons--the computer is trained to “think.” “The neural networks,” Hamrita says, “can juggle the different variables in the bird, the weather, economic factors--all of the things that go into making the best climate-control decisions.”
As fascinating as her work is, Hamrita finds it easy to keep things in perspective with the help of her six-year-old son. He watched recently as she tended her research. “You shouldn’t be so excited about your work,” he told her. “You’re only a farmer, a farmer who works with chickens!”
|
|
|
