Scale
Efforts often focus on specific locations, limiting broader impact.
Student Team Project
Our capstone brief asked us how we could use design and AI to address a pressing societal issue. We chose coral reef decline, a problem with significant ecological and economic consequences. Coral Keepers was our response: a classroom-based system pairing a live reef tank with an AI-supported learning platform, giving students the opportunity to grow real coral that can eventually be returned to restoration organizations and the ocean.
I led the team’s research and usability testing, recruiting and conducting interviews, coordinating contextual inquiries, and synthesizing what we learned into design opportunities. On the product side, I led the mobile experience from concept through its initial design and contributed to its usability testing, while also contributing to the design and development of our web dashboard.
PROBLEM
Coral reefs are declining faster than restoration efforts can save them, driven by climate change, pollution, and overfishing. What were once vibrant ecosystems are turning into bleached, empty landscapes.
Restoration works, the science and tools already exist. But it depends on specialized scientists working with limited funding in demanding and dangerous field conditions. More accessible ways are needed to engage people in reef conservation and give them a meaningful role in restoration.

RESEARCH
Over ten weeks, we combined secondary research with firsthand research at three marine science organizations, including a site visit to the Skidaway Institute of Oceanography. More than 50 outreach attempts resulted in 12 interviews, so referrals and proxy participants became an important part of our research.



What we heard consistently was that restoration organizations were already operating at the limits of their capacity. Growing and maintaining coral is expensive and labor-intensive, while opportunities for citizen participation remain underused. Across our research, these challenges began to converge around four areas we wanted a solution to address.
Efforts often focus on specific locations, limiting broader impact.
Limited funding keeps restoration efforts small and localized.
Growing and maintaining coral requires a lot of time and resources.
Greater public involvement could expand the reach of restoration.
These findings shifted our focus from supporting existing restoration efforts to exploring how we could expand participation itself.
How might we?Make coral restoration accessible and educational so more people can actively participate?
DESIGN
We explored that question across three directions: tools for the general public to participate independently, support for transporting and deploying coral from labs back to reef sites, and a classroom-based system. The transportation direction came directly from a logistics gap raised by several interviewees.

A tank alone couldn’t carry the classroom experience. We needed a digital layer that worked for both students and teachers: students could use it to explore their coral and follow its growth, while teachers could monitor the tank and manage the classroom experience. The web platform brought both into the same system, connecting the physical tank to the classroom.
Connecting the tank to the platform didn’t solve the harder problem: raw sensor data means little without context. Early testing showed that students struggled to understand what the readings meant and why they mattered. We explored how AI could make the science more approachable without taking the learning away from students. That led us to three different applications of AI.
Analyzes tank data and video to surface daily health scores, vitals, and issues.
Interactive tools, like the coral simulation, help students explore coral concepts.
A friendly AI companion for questions about anything coral restoration.
Partway through, I started questioning whether the teacher’s workflow was supported outside the classroom. Their responsibility for the tank and students doesn’t stop when they leave the room, yet the platform was designed primarily for the desktop. I brought the question to our capstone professor, who pointed to her own use of Blackboard from her phone to stay connected with students.
That reinforced the gap for me, so I advocated for a mobile experience and led its initial design. Rather than recreate the dashboard, I focused the teacher-only app on what they would need most away from the classroom.
VALIDATION
We ran three rounds of usability testing, two through Maze with teachers, and a third in-person at Demo Day. Two issues stood out: participants confused announcements with messaging, and Cory's placement caused many to miss it entirely.
At Demo Day, 12 students completed hands-on tasks with the Coral Keepers prototype. We also collected 56 survey responses to understand how the experience affected participants’ engagement with coral conservation and their perception of public involvement.
TASK
Operated the virtual tank simulation without instruction.
Survey
Said Coral Keepers would motivate them to contribute to restoration.
Survey
Saw potential for public involvement in conservation
Survey
Reported a deeper understanding of coral restoration.
Testing was encouraging but limited — small samples, no live classroom, and no restoration partner actually in the loop. It showed that participants could operate the core interactions without help and that the concept resonated with students and outside reviewers. Whether Coral Keepers could meaningfully expand restoration capacity at scale remains untested. That would require real adoption.
SOLUTION
Coral Keepers offers a classroom-ready coral-growing tank connected to a web platform and teacher mobile app. The tank makes coral growth tangible, while the digital experience connects live tank data to learning and day-to-day management.
Our AI Cory is found throughout the entire product, helping students and teachers translate tank vitals and coral science into information they can understand and act on.

Our coral tank is designed for hands-on marine science and STEM learning, allowing students to cultivate real coral fragments that can be donated to reef restoration efforts.

An AI-supported learning dashboard combining coral care, tank monitoring, lessons, assignments, and grades; all in one place.

Monitor tank health, sensor data, and a live stream from anywhere, and stay connected with classrooms and experts on the go.
Coral Keepers was designed, prototyped, and usability tested, but not deployed. The tank interface and AI-driven health scoring were built as working prototypes. The workflow for returning classroom-grown coral to restoration organizations was modeled as part of the concept, but has not yet been piloted with a partner organization.
REFLECTION
This project was the culmination of 20 weeks of work and four years of design education, but the most important lesson came from outside the process itself; the value of a good team. Working with the right people made it easier to be honest, hold each other accountable, and take creative risks we wouldn't have taken with others. Because of that, we built something with far more meaning behind it.
With another 10 weeks, we'd focus on two areas: refining our digital UI, adding features and expanding on existing ideas, and building out more of the service side of Coral Keepers.