Traffic Noise Assessment Using Intelligent Acoustic Sensors (Traffic Ear) and Vehicle Telematics Data.

Area of Science:

• Environmental acoustics
• Transportation engineering
• Machine learning applications

Background:

• Traffic noise pollution is a significant environmental concern.
• Accurate vehicle classification and fuel type identification are crucial for noise and emissions assessments.
• Existing traffic monitoring methods can be intrusive or lack detailed acoustic data.

Purpose of the Study:

• To introduce Traffic Ear, a novel acoustic sensor system for non-intrusive traffic monitoring.
• To evaluate the accuracy of Traffic Ear in classifying vehicle types and fuel using sound and image analysis.
• To develop and apply a new bottom-up approach for assessing traffic noise using Traffic Ear data and urban mobility maps.

Main Methods:

• Development of an acoustic sensor pack (Traffic Ear) integrating microphones and a computer vision camera.
• Application of sound wave analysis, image processing, and machine learning algorithms for vehicle class and speed estimation.
• Comparison of Traffic Ear data with Automatic Number Plate Recognition (ANPR) camera data for validation.
• Integration of Traffic Ear noise analysis with geospatial and temporal mapping of urban mobility (GeoSTMUM) data for noise assessment.

Main Results:

• High agreement (1-4% uncertainty) between Traffic Ear and ANPR for vehicle type and fuel determination.
• Demonstration of over 8% reduction in traffic engine noise during rush hours in the West Midlands, UK.
• Quantification of a significant weekday-weekend effect on traffic noise, nearly halving the benefit.
• Observation that traffic noise factors (dB/m) are consistently higher on motorways compared to other road types.

Conclusions:

• Traffic Ear provides an accurate and non-intrusive method for vehicle acoustic monitoring.
• The developed bottom-up assessment approach effectively quantifies traffic noise impacts using integrated data sources.
• Findings highlight the potential for traffic noise reduction strategies and the need to consider factors like road type and day of the week.