Tae Hong Park is an Associate Professor of Music Composition and Technology at NYU with a passion for understanding noise and its effects on our health and lifestyle. His areas of research are acoustic music, machine learning, computer-aided music analysis, timbre, and audio digital signal processing. Park is actively involved in the New York 'sound scene', with his Citygram visualization of Acoustic Ecology winning a Google Research Award and recent participation in the 24th Annual International Noise Awareness Day workshop. Issues from airplane noise in New York and Chicago inspired Park and his team to develop NOISY, an AI powered noise reporting system. The group is running a Kickstarter campaign to fund their first units.
NOISY detects airplane noise over a user's house, measures the sound level, and logs the data attached to that flight along with a sound byte. The sound information is crossed against airline data to find which flight flew over your house and lodge a complaint with your local airport. NOISY is built from a Raspberry Pi 3 with an audio interfaced connected to the USB port. The audio collector is a microelectromechanical systems microphone, chosen for its light weight and flat thin profile. The microphone can be attached to a window with adhesive tape supplied with the system, or using a Velcro tape option designed for attaching to walls or other smooth surfaces. The microphone is intended to be placed on the outside of a window or wall and then closed in a window using a Flat CAT 5 cable. Data is recorded through the NOISY dashboard, still in prototype phase but videos of the operation and several screenshots are available on the campaign page. Curiously, everything is shown on computer screens with no mention of an app in development. Complaints are automatically filed to your local airport and if the user opts in to the service their data can be collected with NOISY home base to create a database of airplane noise complaints and a map of the loudest areas over time.
The NOISY campaign is another example of citizen scientists taking data to meet my favorite goal of engineering: making the world a better place. There's a lot of programming and machine learning built into the fairly simple maker skills and electronics in this project and it will be interesting to see how the data is used in the near future. The campaign is not yet successful and ends on May 16, 2019.