Sonos 2019-2020

Sonos Product Manufacturing Test Optimization

Sonos is reinventing home audio for today and tomorrow through its variety of home audio systems. As part of the manufacturing process, a large suite of tests is performed on each device in order to ensure the highest product quality. Audio files are generated at multiple testing steps for each device, leading to potential inefficiencies through redundancy. Further, some tests require a human to listen to audio test files in an enclosed audio booth to detect any suboptimal audio performance, leading to worker fatigue. This year, the Sonos SCOPE team employed tools from data science and machine learning to explore data from the speaker testing process, with the goal of providing insights into optimizing testing to increase manufacturing efficiency and reduce the burden on human speaker testers. 

Faculty Advisor

Scott Hersey

Team Members

Nicholas Sherman

Connor Novak

Sophie Schaffer

Nathan Sampo 

Pravallika Dhulipalla

Project Poster

Summit Breakout Session