Abstract
I played tennis growing up and now coach junior players. At some point in junior tennis, the
question stops being how do I get better and starts being what does my rating say. Both projects
in this portfolio came out of watching that shift happen. The capstone built a sensor to capture
new kinds of swing and impact data from athletes during play. The STS paper asked what
happens when a numeric ranking system gets layered onto junior development and starts
changing how players, coaches, and parents behave around it.
In Fall of 2025, my capstone team of five built Flying Birdies, a sensor that clips to a badminton
racquet and records swing force, speed, and impact data during play. It sits on one side of a
problem. My STS research paper sits on the other, where instead of building a data system for
athletes, it asks what UTR has done to junior players who spent their development years being
measured by it. Flying Birdies clips to the neck of a badminton racquet and measures swing
force, racquet speed, impact force, and acceleration in real time. The hardware is a custom
printed circuit board with an accelerometer, gyroscope, and microphone inside a 3D-printed
enclosure, and data streams over Bluetooth Low Energy to a Flutter iOS app. Five of us built it
it for $531.57, including two PCB iterations. The original goal was actionable
feedback. We decided to eliminate that feature because the sensor hardware was not reliable
enough to produce recommendations we could stand behind, and generating coaching advice
from a $531 device would have introduced its own distortions. What the app shows is raw
numbers: swing force, impact consistency, how this session compares to last week. The player
gets their session data back on their phone over Bluetooth and can look at it on their own. It does
not go to a shared system or leaderboard.
The STS paper asks what a system looks like when it does not step back. The research question
is how the introduction of a metrics-based actor reshapes junior players' experience of training
and competition. Actor-Network Theory is the primary framework, used to map how UTR
changes the relationships among coaches, parents, players, and recruiters. Self Determination
Theory supplements the analysis to track what happens to motivation over time. The rating runs
roughly from 1 to 16, with Division I recruits typically falling in the high single digits. The paper
treats UTR as a non-human actor inside a network of coaches, parents, athletes, and college
recruiters, and each group has reasons to manage the number. Coaches schedule opponents
around the rating. Parents pull kids from losing matches before a result registers, and athletes
quietly stop entering draws where the risk outweighs any gain. Nobody designed this outcome. It
is what a stable network reading the same number will produce on its own, and over time those
adjustments redefine what players understand as progress.
Debugging firmware thresholds and reading accounts of junior players who stopped enjoying
their sport are not the same problem. Doing both in the same semester made the difference feel
smaller than expected. There is a narrower question running underneath each: what does a
measurement do to the person it is measuring? The same dynamic shows up in the players I
coach now. UTR shapes how they approach matches in a way a scorecard never would, because
a scorecard records what happened and stops there. UTR decides what that means.Building
Flying Birdies made that question more concrete. When we decided to cut the actionable
feedback feature, part of the reason was technical. But I also kept thinking about what it would
mean to tell a player something definitive about their swing based on data I knew was rough.
That concern probably cost us some features we could have shipped. The UTR research had been
showing me all semester what happens when a number takes on more authority than it deserves.
People adjust their behavior around it. I did not want to build that. Calling them deeply
integrated would be an overstatement. They ran alongside each other, and most of what I got
from doing both came from the friction of running them at the same time. There is still a tension
I have not fully resolved because I spent a semester building another data system for athletes
while also arguing that the data systems athletes already have are reshaping them in ways
nobody designed. I do not have a clear answer for that, but working on both projects made it
harder to treat engineering decisions as just engineering, and that is probably the most honest
observation I can say about the year.
