A brother and sister at Albemarle High School left this year’s Piedmont Regional Science Fair as grand award winners — the first time siblings won the fair’s top prizes for separate projects.
Sam Rosner, a junior at Albemarle, was named grand winner in the Behavioral and Social Sciences category for his project that modeled how political polarization occurs. Meanwhile, his sister, Anna, a sophomore, designed and built a shoe to help prevent people from falling, which was the grand award winner in the Engineering – Electrical & Mechanical category.
The Piedmont Regional Science Fair was held virtually in March this year rather than at the John Paul Jones Arena. Sam has competed previously at the regional fair but this was his first year winning. He liked that he could present his project face-to-face, albeit virtually.
“Because it allows you to kind of explain your project and you get something from it that you don’t get from just looking at a presentation,”
For Anna, the virtual fair wasn’t as fun as the in-person event. She missed getting to see other students’ projects.
“I feel like the science fair is competition, but more of it is just like sharing the thing that you did,” she said.
Both students will head to the virtual International Science and Engineering Fair next month.
“My dream was just to go to the international fair, so I was absolutely thrilled just to get to do that, especially with this project that means so much to me,” Anna said. “Having him with me was just the cherry on top that we both get to do it together.”
Sam wanted to figure out what causes political polarization because it’s been a frequent topic in the news, especially following November’s presidential election and the insurrection at the U.S. Capitol on Jan. 6.
He had previous experience with agent-based modeling and wanted to apply that expertise to this topic. Starting in December, he got to work creating a computer program to see how political and social beliefs and behaviors relate to political polarization.
“My project is essentially using a computer model to simulate a system or an environment of people, and with this model you can change certain factors, which kind of correspond to variables in the real world,” he said. “What that all means is that you can use a computer program to tell what causes political polarization.”
His top takeaway from the model was that when social interactions and political beliefs are closely connected.
“Meaning that you choose your friends based off what they believe in,” he said. “If you strongly dislike socially people who disagree with you, that really increases polarization.”
Each agent in the model represents a person. The model simulated sustained social interactions, which he said is different from other models that have looked at political polarization.
“Basically, instead of just having random social interactions with everybody in the environment, you have a sustained social interaction with everybody that impacts the agents’ every step,” he said.
“Additionally, my model incorporates something called skepticism, which I haven’t seen in any other places. Basically what this is is that agents resist changing their opinions too much, and I found that it’s actually a really core mechanism in the model.”
Elections, he found, have varying effects on polarization. The strategy of the candidates and how personally connected people feel to who is running can affect polarization. Because of the many variables, he’s still studying the role of elections.
Sam doesn’t have immediate plans to move on to another project.
“I don’t have like another project lined up, so I’m just going to work on this one as long as it’s still interesting,” he said.
Anna wanted to build something that would help people.
Motivated by a family member who has Parkinson’s disease, she designed and built a shoe to help prevent people from falling. Using sensors, motors, micro-processors, neural networks and circuit boards, the shoe can tell when someone leans a little too far forward or backward and then it moves slightly — about 5 centimeters — in one direction to help a person balance.
“Your body’s last line of defense against the fall is widening your base of support to keep you from falling,” she said.
To turn a used Ked into a shoe prototype, Anna installed pressure sensors in the heel and toe and cut out part of the sole to make space for the motors.
“The stuff in there, it’s not meant to be cut, and I don’t have any super specialized tools, so I was just sort of hacking at it — precisely hacking at it,” she said.
To train the sensors and algorithms, she leaned forward and backward for about 30 minutes, she said. That data was fed to the neural network.
“What it does is it takes 10 data points over 50 milliseconds, and then from that, if it ever gives a return point of this of, uh oh, there’s about to be a fall, then that triggers the shoe to move.”
Anna also had to test the shoe to make sure the system wouldn’t trigger while a person was standing up or walking. The system didn’t detect a false positive, but her shoe isn’t ready for everyday use — it’s too fragile, she said.
“Currently testing it, I have big circuit boards strapped to my legs,” she said.
When testing it, she wore pads to protect herself, but nothing bad happened.
If a shoe company wanted to help make her prototype a reality, then that would be a dream come true, though she is skeptical that it will happen. Her dream job would be to have infinite funding and parts to make stuff to help people.
“This just made me think that if I conceptualize a thing and watch enough YouTube videos about how to do it, I can actually make it happen,” she said.