Turning reef lovers into data contributors
The ReefOS Community Science App by Coral Gardeners is a platform designed to engage a global community of ocean enthusiasts in coral reef conservation. Through live camera streams, species labeling, and citizen science participation, volunteers help build a data set used to train AI for large-scale biodiversity monitoring. As the UX/UI designer, I focused on making the experience accessible, intuitive, and engaging for non-experts.
Organization
Coral Gardeners
Project Type
Desktop App
Role
UX/UI Designer
Tools
Figma
The Problem
CG Labs needed a high volume of structured, labeled biodiversity images to train AI models capable of monitoring coral reef health. While early reef enthusiasts were already identifying marine species from livestreams and logging their observations into spreadsheets, the process was unstructured and the data largely went unused.
The Goal
Create a tool that makes it simple for volunteers to label more reef species. By streamlining the data collection process, volunteers can easily tag species and contribute to a larger dataset that will help train AI models for better reef health analysis.
User Interviews
The initiative began with a Facebook group of reef enthusiasts who enjoyed Coral Gardeners’ live streams. They started identifying fish and logging sightings into spreadsheets, but without a clear purpose for the data. To improve the process, we interviewed three members with marine life experience to better understand their needs and how to make the labeling more effective.
Key Insights
Volunteers are motivated but unsure of their accuracy
Lack of confidence in species ID
Slow data transfer process
Unclear how data is used or who sees it
Data quality is critical for AI training
Varying skill levels create inconsistencies
Outcome (So Far)
The app hasn’t launched yet. However, early prototypes helped clarify the concept and improve the observation flow based on feedback from the community.
Lessons Learned
Involving real users early helped uncover key pain points, especially around data submission and clarity in species identification. Iterating based on their feedback significantly improved the app’s usability and relevance
