Noise Pollution Mapped using Crowdsourced Data Collected by Smartphone App

As reported by Wired UK: The main causes of noise pollution -- machines, transportation, poor urban planning, people -- vary considerably in terms of location and intensity throughout the year. Consequently, measuring these noise levels on a large scale is often a time consuming and sometimes inaccurate procedure.

However, Rajib Rana, Chun Tung Chou, Nirupama Bulusu, Salil Kanhere and Wen Hu at the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia's national science agency, have devised a novel way of combating many of the problems that have hindered previous methods of monitoring noise in the past. Rather than spending thousands on equipment, Rana and his team have crowdsourced the data from smartphones -- they call it the "Ear-Phone".

The average smartphone has enough sophisticated technology (on-board microphones, GPS, time stamping) to make it an extraordinary mobile monitoring device. However, the implementation of the Ear-Phone came with its own set of challenges -- there are many ambient noises picked up by a smartphone that would be of no use to the researchers, namely intimate conversations that sound loud to the phone, but wouldn't be disruptive to passersby, as well as the sound of rustling clothes, or keys or money jingling in pockets.

The team at CSIRO isolated these problems and fixed them accordingly, programming the phone to recognize a conversation, wait until it was over, and then start recording again. They also made sure the GPS was only collecting data when outdoors and while being held in the user's hand.  

This was made possible by taking advantage of the automatic proximity sensors and accelerometers that are standard in most smartphones. These sensors generate specific feedback when the phone is handled, which the software developed by Rana and his team can then use to assess whether or not it's an appropriate time to take a reading. Using this method, handheld usage can be detected with an accuracy of 84 percent.

Once all of the criteria have been met, the phone will take a sound recording, complete with exact location and time. The data is then sent to a central server the moment the phone is connected to Wi-Fi.

The only major downside with the crowdsourced method is battery life -- always-on GPS and Wi-Fi can drain a phone's power in a matter of hours, but the team claims it is working on a solution.

Noise pollution is thought to be a particularly destructive environmental hazard. Not only can it cause hearing-loss, stress and tinnitus in humans, but it's devastating to local Wildlife, reducing livable habitats, disrupting predator or prey detection and avoidance, and even causing infidelity in finches.

The Ear-Phone has been tested on various Android and Nokia phones in Australia.