Low-Cost Sensors for Urban Noise Monitoring Networks - A Literature Review

Judicaël Picaut, Arnaud Can, Nicolas Fortin, Jeremy Ardouin, Mathieu LagrangeSensors 2020 · DOI

TL;DR

This paper reviews the burgeoning field of low-cost sensor networks for urban noise monitoring, highlighting their potential to overcome the limitations of traditional noise mapping and professional observatories. It synthesizes existing literature to identify essential technical characteristics for these sensors and outlines the challenges for their widespread deployment.

What Problem Does This Paper Address?

Urban noise pollution is a significant environmental and health concern, leading to policies and strategies aimed at its reduction. Traditional methods for assessing noise, such as noise mapping software, rely on numerous assumptions and often only consider transportation and industrial noise, failing to capture temporal dynamics or the full complexity of urban sound environments. Professional measurement observatories, while more accurate, are prohibitively expensive for dense, widespread deployment. This gap necessitates the development of affordable, scalable solutions for comprehensive noise monitoring.

Methods

This paper is a literature review, synthesizing existing research on low-cost noise sensors and networks. It categorizes sensor network architectures (fixed, mobile participatory, mobile) and delves into the technical aspects of sensor development. Key areas examined include:

  • Sensor Platforms: Analysis of microcontroller (MCU) based platforms, custom-designed boards, and mini-PCs, evaluating their trade-offs in terms of cost, computing power, and energy consumption.
  • Microphones: Comparison of electret condenser microphones (ECMs) with analog and digital MEMS microphones, highlighting the advantages of MEMS for cost, size, reliability, and performance.
  • Signal Processing: Discussion of essential pre-processing steps like frequency weighting (A-weighting) and frequency equalization, and the trade-offs between analog and digital implementations.
  • Calibration: Review of methods for calibrating low-cost sensors to ensure accuracy, including the use of acoustic calibrators and comparison with reference sound level meters.
  • Data Transmission Protocols: Evaluation of various wireless protocols (Bluetooth, Wi-Fi, Zigbee, LoRaWAN, Sigfox) for their suitability in noise sensor networks, considering range, data rate, and power consumption.
  • Additional Sensors: Exploration of integrating sensors for meteorological conditions (temperature, humidity) and other environmental factors to enhance data interpretation.
  • Sensor Life: Assessment of factors affecting sensor longevity, particularly battery life and component durability.

Datasets

Not discussed in the paper. The paper is a literature review and does not present new experimental data or datasets.

Key Insights and Ideas

  • The "Democratization" of Noise Monitoring: Low-cost sensors, driven by advancements in miniaturized electronics and affordable processors, are making detailed urban noise monitoring accessible beyond traditional professional networks.
  • MEMS Microphones as the Future: Digital MEMS microphones are emerging as the preferred choice for low-cost noise sensors due to their cost-effectiveness, small size, reliability, and integrated analog-to-digital conversion, which reduces susceptibility to electrical interference.
  • Trade-offs are Paramount: Designing effective low-cost noise sensors involves balancing performance (accuracy, frequency range, dynamic range) with cost, power consumption, and sensor lifetime. There isn't a one-size-fits-all solution; the optimal design depends on the specific application.
  • Autonomy is a Major Hurdle: Achieving long-term autonomous operation for wireless noise sensors remains a significant challenge, often requiring direct connection to the power grid or sophisticated energy harvesting solutions.
  • Beyond Basic Leq: While equivalent sound levels (Leq) are standard, the paper hints at the growing need for more sophisticated acoustic indicators and event detection to capture nuanced aspects of urban soundscapes.

Key Results

The paper synthesizes findings from numerous studies, indicating that low-cost sensors can achieve performance levels that are "good enough" for many urban noise monitoring applications, particularly for strategic noise mapping and public information. For instance, some systems demonstrate deviations of less than 2 dB compared to Class-1 sound level meters. The review highlights the successful deployment of networks with dozens of nodes and the potential for these systems to provide rich acoustic data for urban planning and citizen engagement.

Limitations and Future Work

The authors acknowledge several limitations in the current literature and research:

  • Lack of Standardized Performance Metrics: There's a consistent lack of detailed description of acoustic performance metrics like residual noise, sound level dynamic, and precise frequency range in many studies. Often, reported characteristics are based on component datasheets rather than integrated sensor performance.
  • Maturity Levels Vary: Developments range from simple proofs-of-concept to operational deployments, indicating that the field is still evolving.
  • Autonomy Not Extensively Studied: The operational autonomy of wireless noise sensors is not always thoroughly investigated, with many recent advancements relying on wired power.
  • Environmental Robustness: While some studies touch upon the effects of temperature, humidity, and wind, more comprehensive testing and protection strategies for sensors in outdoor environments are needed.

Future work should focus on:

  • Standardizing Performance Evaluation: Developing consistent methods for evaluating and reporting the acoustic performance of low-cost sensors.
  • Improving Sensor Autonomy: Further research into energy harvesting, battery management, and power-efficient processing is crucial for truly autonomous operation.
  • Addressing Network Challenges: Developing robust algorithms for fault detection, optimizing sensor placement, and managing data from large-scale hybrid networks.
  • Enhancing Data Interpretation: Exploring advanced acoustic indicators and machine learning for sound source identification and event detection.
  • Developing Hybrid Networks: Integrating low-cost sensors with professional networks and other environmental sensors for a more holistic understanding of urban environments.

Summary updated

BibTeX
@article{picaut2020low,
  author = {Judicaël Picaut and Arnaud Can and Nicolas Fortin and Jeremy Ardouin and Mathieu Lagrange},
  title = {{Low-Cost Sensors for Urban Noise Monitoring Networks - A Literature Review}},
  journal = {Sensors},
  year = {2020},
  doi = {10.3390/s20082256},
}
Figure