Robot Coder

“We tried to code a robot so that it could move autonomously around the Olin-Rice Science Center. This research surprised me because locomotion and navigation are so innate to the human experience.” —Malini Sharma

By Malini Sharma ’19
New York
Neuroscience, computer science

Last summer I got to conduct research with Professor Susan Fox on artificial intelligence; namely, we tried to code a robot so that it could move autonomously around the Olin-Rice Science Center’s first floor using image and location data.

We went about this by creating in the robot a convolutional neural network, which is a type of artificial neural network. Artificial neural networks are machine learning techniques modelled on the human brain. We chose to use a convolutional neural network because they’re best for processing image data, which we wanted to use.

Day to day, our work looked very different. Sometimes we’d spend the entire day driving the robot around Olin-Rice and collecting all the image data we could. Other days we focused on creating code that would better allow us to analyze our data. We also read papers about robotic navigation, trying to learn new techniques we could use with our robot.

This research surprised me because locomotion and navigation are so innate to the human experience. For example, once we’ve been in a building a handful of times, it begins to look familiar, and we can eventually navigate from point A to point B with ease. Our robot, though, could get lost in a hallway it had seen thousands of times before.

I felt very prepared for my research from my previous classes, the majority of which included an independent research final project. So, going into my final year at Macalester, I’m fairly well versed in how to go about research in a general sense. Also, whether I’m studying for an exam or investigating a problem, my professors have always encouraged me to think creatively and approach the material however worked best for me. This carried over into my research, too. In particular, my research mentor, Professor Fox, encouraged us to be creative by giving us free rein in how we tried to address problems. She was always there for consultation, but for the most part we could solve each problem we were assigned as we saw fit.

Being a science major at Mac has been hard but rewarding. Your professors expect a lot out of you, but you are also surrounded by hundreds of other passionate students who are more interested in collaboration than competition. Your professors want you to succeed, and if you go to them for advice, they always take the time to help.