Novel wireless pervasive sensor network to improve the understanding of noise in street canyons

Abstract

This paper presents the first results of the application of an array of up to 50 novel wireless pervasive sensors, referred to as motes, developed at Newcastle University in the project, MESSAGE (Mobile Environmental Sensing System Across Grid Environments). The motes measure noise, carbon monoxide, nitrogen dioxide, temperature, humidity and traffic occupancy/flow. The system has wireless communication (using ZigBee standard) to allow daisy-chaining of data from mote to mote (either static or dynamic) to router technology which forwards the data into a gateway. This passes the data through GPRS (or hard wired web link) to be captured in real-time into an Urban Traffic Management and Control, compliant database. Following a brief discussion of the health impacts associated with noise exposure and the policy context, the validation and proof of concept of the use of a pervasive sensor array to measure noise in an urban area is demonstrated through separate deployments in Leicester (UK) and Palermo (Italy). The motes provide real-time minute by minute, noise data [Leq in dB(A)] and their low cost and small size means that they can be pervasively deployed. From the two applications the results clearly show the potential of these motes to measure noise levels and to improve the understanding of the effect of traffic flow characteristics on the minute by minute variation in noise levels in busy urban streets. The distribution of one minute LAeq noise levels measured by the standard deviation were shown to exhibit consistent but different characteristics depending which of three measurement clusters, primary congested road and well ventilated secondary roads with busy traffic, tertiary roads without acoustic screening from busy roads and finally residential streets. Predictive relationships which allow the magnitude of the standard deviation to be expected depending on the noise level are presented. The results demonstrated a measurement system and predictive method that offers potential to explore more suitable noise parameters that correlate the human response to traffic related noise.