“I helped build up a database of images and write code that interprets the data from the robot into concrete instructions on how it should move.”
—Kayla Beckham ’19

Kayla Beckham ’19
Memphis, Tennessee
Computer Science

I spent my summer at Mac teaching a robot to travel. When we began, our squat little TurtleBot could already roll between posted signs marking the location with QR codes. Over the course of 10 weeks we made gradual improvements until it could navigate independently based on what it sees and how it had been moving.

Working in a small group with computer science professor Susan Fox and Xinyu Yang ’19 (Shanghai, China), another student researcher, I helped build up a database of images and write code that interprets the data from the robot into concrete instructions on how it should move. The robot compares its camera input with the image database and tracks its own movement. It uses a combination of all of its data to navigate our science building. It’s a little shocking to look back now and realize how much the robot has changed. With the accumulation of all our small improvements, we laid a whole new foundation for the TurtleBot’s navigation.

We learned a lot while we were working, from the proper terminal commands to communicate directly with the robot, to stories about competitions at artificial intelligence conferences where robots serve hors d’oeuvres using some of the same navigational procedures that we implemented. Most importantly, we got hands-on experience tackling a large, complicated program that is bigger than what we could accomplish in just 10 weeks. This project is the product of several years’ work by Professor Fox and other Mac students, and I couldn’t be happier to contribute.

Kayla’s research was supported by the Clare Boothe Luce Scholar Award.

December 13 2017

Back to top