What I Learned Creating A Self-Balancing Robot
Anna Garverick talks about her experience creating a robot that remains centered on a moving ball and the most important lessons she took away from the experience.
The night before Anna Garverick was to present her independent project for Northwestern Engineering's Master of Science in Robotics (MSR) program, she learned a valuable lesson: Don’t turn your back on an unstable autonomous robot.
Garverick challenged herself to build a self-balancing robot that would stay centered on a moving ball for her standalone project. She was nearly ready to showcase her work when she was treated to an unwelcome surprise.
“I turned around for a second and my robot drove itself off the table,” she said. “It totally shattered, but I managed to hold it together with enough hot glue to get in a few test runs that night.”
The autonomous stage dive was just one of many hurdles Garverick overcame during the development of her ball-balancing robot, and it helped her hone the resilience she said is necessary to succeed in the robotics field.
“It is easy to get discouraged and feel like everything is going wrong, when in reality you are one simple change from everything working," she said. "Anything in robotics is rarely a straightforward, linear process, and I have learned it is important to continue iterating and push yourself to try new approaches.”
Garverick takes pride in pushing herself. It’s what led her to take on the challenge of building the self-balancing robot.
“Much of my project focused on concepts that I had not taken courses in yet," she said, "so I used the resourcefulness I developed in my first quarter to help seek out good information to guide me through the process."
During that process, she drew on inspiration and help from her MSR classmates and professors.
“Building my own system completely from the ground up meant that any problem I ran into could have many potential causes and many potential solutions,” she said. “Fortunately, if I explained where I was stuck to my classmates working on similar projects or Professor Matt Elwin, they had great suggestions to push me in the right direction.
“There were many algorithm re-designs and hardware modifications along the way, but even when it seemed like I was going in circles, I was learning a lot through the process."
That learning helped her during the overnight work it took to get her battle-scarred robot ready to present the next morning.
“Fortunately, I designed the chassis with modular 3D printed components to allow for quick iterations," she said, "so I was able to print a second chassis overnight and assemble a new robot the next morning,”
With that project behind her, Garverick turned her attention to her final project, which will feature working on controls for an exoskeleton. While the two systems will be different in nature, Garverick expects to apply many of the lessons she learned from working on her independent project.
Not all of those lessons were technical, she joked.
“Now I know to never leave omnidirectional robots unsupervised on elevated surfaces,” Garverick said.