WASN-Based Day–Night Characterization of Urban Anomalous Noise Events in Narrow and Wide Streets

Abstract

In addition to air pollution, environmental noise has become one of the major hazards for citizens, being Road Traffic Noise (RTN) as its main source in urban areas. Recently, low-cost Wireless Acoustic Sensor Networks (WASNs) have become an alternative to traditional strategic noise mapping in cities. In order to monitor RTN solely, WASN-based approaches should automatize the off-line removal of those events unrelated to regular road traffic (e.g., sirens, airplanes, trams, etc.). Within the LIFE DYNAMAP project, 15 urban Anomalous Noise Events (ANEs) were described through an expert-based recording campaign. However, that work only focused on the overall analysis of the events gathered during non-sequential diurnal periods. As a step forward to characterize the temporal and local particularities of urban ANEs in real acoustic environments, this work analyses their distribution between day (06:00–22:00) and night (22:00–06:00) in narrow (1 lane) and wide (more than 1 lane) streets. The study is developed on a manually-labelled 151-h acoustic database obtained from the 24-nodes WASN deployed across DYNAMAP’s Milan pilot area during a weekday and a weekend day. Results confirm the unbalanced nature of the problem (RTN represents 83.5% of the data), while identifying 26 ANE subcategories mainly derived from pedestrians, animals, transports and industry. Their presence depends more significantly on the time period than on the street type, as most events have been observed in the day-time during the weekday, despite being especially present in narrow streets. Moreover, although ANEs show quite similar median durations regardless of time and location in general terms, they usually present higher median signal-to-noise ratios at night, mainly on the weekend, which becomes especially relevant for the WASN-based computation of equivalent RTN levels.

Multilevel Annoyance Modelling of Short Environmental Sound Recordings

“…In the context of a smart-city framework, one could imagine a wireless acoustic sensor network (WASN) large enough to cover a whole urban area; having a noise annoyance prediction algorithm at the node position that can return live annoyance scores to a central server from sounds recorded locally by the sensor would make for a useful application for environmental protection officers and other stakeholders at community or local authority level [52]. A relevant issue to consider from the WASN perspective, is that previous studies conducted in both urban [21] and suburban [20] environments, there is a clear influence of the type of environment around the sensor location on the types of noise detected. Not all the urban or suburban locations for sensors have frequent sirens or horns, it depends on the more common activities (leisure, hospitals, etc.…”

“…In this section we detail the several methods applied our experiment from the perceptual test design based on an urban sound dataset [21] to the multilevel linear regression modelling applied to obtain the annoyance prediction described as contribution in this paper.…”

“…In order to obtain a proper representation of the acoustic environment in the design of the perceptual tests, a large quantity of recorded data is needed. The data gathered in this project belongs to different recording times and urban locations, using the Wireless Acoustic Sensor Network (WASN) deployed in Milan (Italy) in the framework of the LIFE DYNAMAP project [19,21].…”

“…The dataset used in this study has been obtained by homogeneously sampling several hours, in both weekday and weekend, with 24 sensors distributed along the urban District 9 of Milan [34]. After that, experts from the DYNAMAP developing team labelled the acoustic events of the recordings manually to obtain a 151-h dataset [21]. Due to the nature of the project, that consisted in removing events not related to traffic noise from the noise map computation, events were grouped in RTN (Road Traffic Noise) that belongs to the 83.7% of the total time of the dataset, and ANE (Anomalous Noise Event) with the 8.7% of the total time.…”

“…The most representative audio excerpts were selected, using a wide range of sound types (sirens, airplanes, people talking, dogs barking, etc.) [20,21], keeping the constants of location and sensor calibration. However, sound annoyance depends on the acoustic characterization of each sample, and it is possible to classify the acoustic excerpts depending on their characterization, which can be the basis to ask participants about their perceptions.…”