Skip to main content

Research sees smartphones collect data to map noise pollution

Researchers at the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia's national science agency, have discovered a novel way to map the effects of noise pollution on our everyday lives.

The research is part of the growing trend in "citizen science" brought about by the advancing levels of technology being carried around in the average person's pocket.

The team, led by post-doctoral fellow Rajib Rana, have been using smartphones installed with a dedicated app called Ear-Phone to build complex maps of the urban environment, and determine which areas of cities are most effected by noise pollution.

Ear-Phone takes short audio recordings, and includes a detailed report of the time and exact GPS location of the noise reading. It also uses context-aware sensing to only take a reading if the situation is right.

For instance, it can detect whether or not the phone is being kept in a user's pocket, in which case the rustling of clothing would produce interference. It will only take a reading and log GPS co-ordinates if it's being held in the user's hand, and if no conversation is taking place nearby.

To perform this complex decision, the software can recognise the audio form of a conversation, stop recording audio, and then resume recording again once the conversation is over.

The teams claims that the accuracy of context detection using these methods is 84 per cent.

Noise pollution isn't a trivial problem. It's been shown to raise stress, and cause hearing loss and tinnitus in humans. It's also damaging to wildlife, reducing liveable habitats for many bird and animal species.

The team has criticised traditional methods of gathering noise pollution data, arguing that even "state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated, as they rely on population and traffic models rather than on real data."

The app is available for Nokia N95, N97 and HP iPAQ, HTC One mobile devices.

Image: Flickr (jovike)