Enhancing Youth Soccer

Hi! My name is Aiden Guenther, I am a rising sophomore from Kailua, Hawaii. I am currently undecided on my major, but with that said, I’m strongly leaning towards a double major in Economics and Computer Science. Outside the classroom, I’m also a member of the Men’s Soccer program, which might help explain some of the inspiration behind my project. With the help of the Whitman Internship Grant, I spent this summer interning with a local soccer club on Oahu as a software developer. 

To preface my project, it’s important to know a little about youth sports in Hawaii. Youth in Hawaii face many challenges in gaining access to high-level coaching, partly due to the high cost of living, which poses a unique challenge for youth soccer clubs struggling to meet the demand from young players. I paired with FC Hawaii, a local youth soccer club, to address this challenge. Enter KickerPose, an AI-powered coaching tool that youth soccer players can access from a phone or laptop and get instant feedback on their kicking form.

The development of KickerPose took three stages. I spent the first two weeks of my internship in the research phase. Using excellent resources recommended to me by Professor Parteek Kumar of the Computer Science department at Whitman, I took several online courses in machine learning and computer vision. During this phase, I also spent time at the soccer field working with youth players collecting data, and working with players to identify the most important data points. Collaborating with youth players and coaches, I narrowed down the complex kinetics of kicking techniques to three key areas: the swing, the plant-foot, and body posture. These were identified as the crucial elements for generating inferences about an athlete’s kicking form.

The development stage took the next four weeks and I worked on transferring what I had learned in the research phase into functional models that we could test. This was the most challenging phase of the project and required lots of patience in reworking the approach to different problems I ran into. Finally, during the deployment phase which lasted the last two weeks, I worked on implementing the working models into an interface that was easy to use and provided impactful feedback. If you are interested in seeing the final product, you can check out KickerPose online at our website!

The Whitman Internship Grant gave me an amazing opportunity to collaborate with people across numerous disciplines on an impactful project that benefits real people. Reflecting on the project I was able to overcome many challenges, which made me a better learner. I can see this project enhancing my resume by giving me real-world experience and exposure to a career in computer science. I am excited to continue learning about machine learning and continuing to build off of this experience. I am very grateful for the incredible experience the Whitman Internship Grant afforded me this summer and I would highly encourage other students to take a look at the program.