Artificial intelligence — the topic of one of this year’s January Term courses — played an important role in science projects presented by two University High School students last month.
Sophomore Uses AI to Maximize the Efficiency of Solar Cells
Working with Dr. Brandon Hogan, sophomore Caine Ardayfio used an artificial intelligence algorithm to discover a way to make solar panels more efficient. His project, titled “Development of optimal microstructure in organic solar cell active layer through genetic algorithm” advanced to the 31st annual Hoosier Science and Engineering Fair on March 30, 2019.
There, he won the Award for Excellence in Environmental Science Senior Division and a cash prize, second place overall in the 10th grade category, and a spot on the Indiana team that will compete in the Intel International Science & Engineering Fair in Phoenix next month.
“I used a machine learning (AI) algorithm to manipulate the shape of the active layer, which absorbs the light,” explained Ardayfio. “By changing the shape of it, you can make it more efficient. The shape that the machine learning came out with was a branched-off shape, and it made the solar panel about 30 percent more efficient.”
Freshman Uses AI to Identify Bone Fractures
Freshman Aisha Kokan also participated in the Central Indiana Regional Science & Engineering Fair. Her project, “Using artificial intelligence to assist radiologists in bone fracture diagnosis” used a neural network algorithm to distinguish between images of healthy bones and fractured bones. She was able to achieve about 80 percent accuracy, and she hopes to fine-tune the algorithm and use a larger data set to increase the accuracy next year.
Kokan’s research won the Mu Alpha Theta Award for the most challenging, original, thorough, and creative investigation of a problem involving mathematics accessible to a high school student at the fair.