Pervasive wireless sensors: A new monitoring tool for road traffic noise evaluation

Marouf, Shadman, Bell, Margaret Carol, Goodman, Paul, Namdeo, Anil and Neasham, Jeffery (2018) Pervasive wireless sensors: A new monitoring tool for road traffic noise evaluation. Applied Acoustics, 135. pp. 48-59. ISSN 0003-682X

Preview

Text FF3E233C-7A52-4BED-B9E2-435D09BC793C.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives 4.0. Download (3MB) | Preview

Abstract

Noise pollution is estimated to affect 170 million citizens in Europe, causing serious public health problems [1]. The World Health Organisation claims that at least one million healthy life years are lost per year in Europe due to road traffic noise [2]. Effective management of noise requires an understanding of its causes. This understanding is limited by traditional monitoring methods, which employ expensive equipment and are labour intensive. This paper presents the results of a comprehensive programme of correction and validation of a low-cost device referred to as an eMote for pervasive monitoring and is the first to quantify the accuracy of inexpensive noise systems that use microphones typically costing about one Euro. Pervasive wireless noise sensors (eMotes) were validated by co-location with precision sound level meters in controlled indoor, and at roadside outdoor environments. Strong linear relationships between the eMotes and the precision systems, across a noise range between 55 dBA and 94 dBA were observed and exhibited consistent bias compared to the precision measurement. Therefore, a generic, corrective relationship was derived and validated in three contrasting outdoor traffic noise environments, employing both short-term attended, and long-term unattended measurements, which were carried out during day and/or evening and/or night periods. The eMotes were shown to respond consistently to white and pink generated noise during the evaluation of the accuracy process, and the generic correction algorithm for white noise delivered better than 3 dBA accuracy in comparison to precision data at a one-minute averaging resolution. The correction algorithm improved the concordance correlation coefficient (ccc) and coefficient of determination (R2) of the eMote measurements against those of the precision instrument. Removal of short-duration, excessively loud events (e.g. sirens), which represented 2% of the total data, improved the ccc and R2 values further typically to 0.74 and 0.60 respectively, which is considered good, given the limitations of the experimental procedure. The research provides scientific evidence that whilst not acceptable for compliance monitoring to standards for noise exposure, the eMote is a valuable system to screen for excessive exposure; to understand the causes of traffic related noise in urban areas; to provide an indication of the spatial and temporal variation in noise levels and the knowledge to design appropriate solutions, in turn this will lead to more effective abatement. The continued monitoring allows the impact to be quantified giving confidence that intervention measures are worthwhile, delivering added value compared to current measurement methods.

Actions (login required)

View Item

Downloads