Brown CS News

Brown CS PhD Student Eric Ewing Helps Multi-Robot Research Lift Off At Brown And Beyond

None

by Juan Siliezar (Communications Manager and Writer for the Physical Sciences)

For Brown University doctoral student Eric Ewing, teaching his first college-level class is a long-awaited dream that is finally taking flight.

Through a course being offered for the first time at Brown this semester — called Computer Science 1952-Z Robots as a Medium: Creating Art with Teams of Robots — Ewing is helping to offer undergraduates an introduction to using robots to create paintings, music and even choreographed movements.

For Ewing, the course is helping to advance his greatest passion: making robotic systems and robotics curricula more accessible.

“Part of the great appeal for me getting into computer science and CS education was that anyone with a laptop or access to a computer and the internet could learn how to program, but robotics is typically much harder to get into because you actually need a robot,” Ewing said. “So it’s increasingly important for me to make sure that robotics education becomes as accessible as computer science education has become — and the easiest way to do that practically is for the people with robots to provide those types of opportunities.”

“Brown allows these types of classes where the broad goal is to have students from a wide range of disciplines — including engineering, computer science, art and any other background — come together,” said Ewing, who transferred to the University in 2022 from the University of Southern California when Ayanian, his long-time Ph.D. adviser, left USC to come to Brown.

Now a graduate student in Brown’s Department of Computer Science, Ewing conducts research in Ayanian’s lab that involves studying how to control multiple drones at once. While drones are used with increasing frequency across a range of practical applications, building more advanced multi-robot systems could have benefits for everything from manufacturing to environmental monitoring to emergency response. Setting up the complex operating system for these drones, called Crazyflies, and the motion-capture systems to collect data is difficult, Ewing said, which often presents a barrier for those looking to enter the field — a barrier that he aims to lessen.

“A lot of my recent research has basically been trying to lower the bar to working with the types of drones we use,” Ewing said. “They fit in the palm of your hands and run on open-source code, making it so you can program a team of them to do just about anything. But it’s a hard and complicated system to get up and running. So many things can go wrong during setup and it can be hard to tell issues apart. When something in the pipeline isn’t working, the end result is almost always that the drones crash.”

At Brown, Ewing has helped to open Ayanian’s lab (the Automatic Coordination of Teams Lab), both figuratively and literally: to novices who have no experience with Crazyflie drones as well as professionals who could benefit from their expertise.

“I’ve broken our system so many times that I’ve gotten good at figuring out what parts are broken,” Ewing quipped.

Ewing has collaborated with several labs around the country to help them get started, guiding them on what type of camera systems they can use, kinks they’ll have to work through, common problems like the drones flying straight up into the ceiling, and even how to physically configure their space. Ewing has also spent the past year developing a web application to make the software on which the Crazyflie drones operate easier to use for a much wider array of researchers.

“There are not many individuals who’ve done what we have, so we try to help others get established,” Ewing said. “There aren’t enough researchers in multi-robot systems yet to solve the grand challenges that the field faces in the next 20 to 30 years, so the more labs that can get set up, the better the research we will produce as a field.”

At Brown, Ewing has expanded his work by hosting interactive demos in the lab with students from Girls Get Math, a program for high schoolers, and he taught a summer course for another group of high schoolers from across the U.S. and the world. He’s also been supervising a high school intern from Barrington, Rhode Island, and undergraduates at Brown, teaching them to build, operate and troubleshoot the Crazyflie drones.

“The fact that we can have undergrads working with us and adapting what they learn to their own projects has been a pretty good success so far,” Ewing said.

Now that CSCI 1952-Z launched this semester, Ewing is relishing the opportunity to co-teach a course that builds on curricula he’s created and introduces robotics to undergraduates, regardless of their concentration or prior experience.

“We’ve now got many students who come into the lab to work on their own personal projects or as part of the class, and I’m really excited about that,” Ewing said. “I really enjoy advising students and helping them achieve their goals.”