A project by Ear-IT has seen the Spanish city of Santander be fitted with thousands of microphones to enable the real-time monitoring of mass flows of people and traffic. The project investigated how the sensors could enable systems to improve safety, surveillance, security and energy efficiency within the city. While the choice of acoustic sensors is less common in the Internet of Things, the project drivers say that monitoring audio can offer many advantages over traditional sensors.
As well as monitoring traffic density with the eventual aim of triggering pollution-reduction measures the technology can, for instance, detect sirens and change traffic lights in favour of an approaching ambulance.
Acoustic sensing was used as it offered two main advantages, according to the researchers. While a good visual feeds rely on CCTV cameras being positioned correctly, audio offers Non Line-of-Sight (NLOS) sensing offers a simpler, cheaper way of monitoring many different criteria including aspects which may or may not be seen.
They are also multipurpose, relying on software to flag up any items of interest based on unique acoustic signatures, such as a siren. This acoustic detection is a three part process, beginning with a sound being detected by the microphone, the background and target sound being separate, and the target sound being defined and finally categorised using computer modelling and manual labelling.
Testing for the two-year project, which was part-funded by the European Commission, ended in September and the project is due to be wrapped up by the end of the year.