HowLoud - Mapping Noise Levels with Math Models

Brendan Farrell and his team at HowLoud want to make sound a part of big data. After building a map of the sound levels around Los Angeles and Orange County the company is running a Kickstarter campaign to fund a soundmap of the entire continent.

HowLoud takes an innovative approach to developing their noise studies by using data to generate more data. Traditional companies use tens of thousands of microphones to generate noise studies. HowLoud builds a three dimensional model of a city and calculates the sound profile generated during different times. Finding the angles of reflection for the sound waves and pushing the noise through the environment creates a visual map.

Human hearing and comfort takes more factors into effect. We dislike noise in the night more than noise during the day, and intense noises for short times irritate us more than moderate noises that run as constant background noise. Based on these factors HowLoud doesn't report decibel levels but rather a sound level.

Scores below 40 indicate that the location is loud but above 90 the location is quiet. Users are expected to compare data from areas they already know with sound scores from new areas.

HowLoud is taking information already available and turning it into information that has more value. Brendan was looking for a quiet place to live in Los Angeles. After a few failed, loud attempts he realized that a method to understand how much noise exists geographically would be useful. This is a great example of seeing a need and using applied mathematics and engineering to solve the problem.

Data is due to be released in September 2015. Developers who want to add sound information to their websites can download json APIs or an adaptive widget to display their sound scores.