snuzz

Snuzz capstone interface for a study of office noise pollution using audio classification

about

snuzz is our capstone project, an android-based noise detection application designed to help monitor excessive noise inside the it faculty office during class hours. the application detects surrounding noise levels and provides alerts when the environment becomes louder than the preferred threshold.

the project also explores the use of machine learning and audio classification to better understand and identify different sources of noise within the office.

my role in the project was full-stack mobile developer and ui/ux designer, where i worked on both the technical development of the application and the overall user experience and interface design.

case study

the idea for snuzz came from a problem we regularly observed inside the it faculty office at davao del sur state college.

the office receives many students throughout the day, especially those consulting instructors and having their capstone papers checked. with students frequently coming in and out, conversations happening at the same time, and instructors occasionally speaking loudly with one another, the office can become noisy during class hours.

we saw an opportunity to explore how technology could help the office become more aware of its noise levels without requiring someone to constantly monitor the environment.

this led us to propose snuzz: android-based office's noise detection application, a mobile application that monitors surrounding sound levels and alerts users when the noise exceeds a set threshold. the study also allowed us to explore machine learning and audio classification as a way of identifying and understanding different types of sounds rather than relying only on their volume.

as the full-stack mobile developer and ui/ux designer, i was responsible for turning the concept into a functional mobile application. i worked on the application's interface, user flow, frontend implementation, backend integration, data handling, and the core features used for monitoring and responding to noise levels.

one of the main challenges was making technical audio data understandable to the user. instead of showing raw values alone, the interface had to communicate the current noise level clearly and make alerts noticeable without making the experience complicated.

working on snuzz pushed me to think beyond simply building features. i had to balance usability, visual clarity, application logic, data handling, and the research requirements of the study while making sure the final product addressed the actual problem we observed in the office.

snuzz became an opportunity for me to combine mobile development, ui/ux design, research, and experimentation with machine learning in one complete project.

Role
Full-stack Mobile Developer, UI/UX Designer and Researcher
Project Team
Joshua L. Cambronero and Jessa H. Rubino
Capstone Adviser
Rhea Mae L. Perito, MSIS