Initial Design & Development of MEMS-based Sound Level Meters for Real-time Construction Noise Monitoring

Traditional sound level meters that comply with Class 1 standards are a requirement in some projects. However, they tend to be expensive, maintenance-heavy, bulky and inflexible. For many construction projects they are overkill. New noise monitoring options, built around MEMS-based sound level meters, provide a more cost-effective and flexible solution.

This research explores how SiteHive developed SiteHive Hexanode Noise devices, using MEMS-based sound level meters, that deliver accuracy in line with international standards, and are also cost-effective and easy to use. It demonstrates how MEMS-based devices can deliver key benefits including lower cost, scalability, and improved data insights for construction projects.

Authors

Ben Cooper-Woolley, Michael Manraj Darroch, Jin Cong, Xander Hendriks
SiteHive Pty Ltd, Sydney, Australia

Qiaoxi Zhu, Tong Xiao, Sipei Zhao, Benjamin Halkon
Centre for Audio, Acoustics and Vibration, University of Technology Sydney, Australia

Presented at Acoustics 2022, Wellington, New Zealand

Abstract

The paper describes the design, development and testing of SiteHive Hexanode real-time noise monitoring devices using MEMS microphones to capture noise. It includes how MEMS microphones were verified and calibrated to meet relevant international standards (IEC 61672), including testing to identify their suitability for use in real-time noise monitoring during construction.

Through collaboration with the National Measurement Institute and the University of Technology Sydney, the research demonstrates that MEMS microphones can achieve the required accuracy for noise monitoring during construction. The work highlights how this approach enables a new generation of cost-effective, real-time environmental monitoring devices that provide greater insight and value on construction sites.

What is a MEMS-based sound level meter?

At the core of this research is the use of MEMS microphones, which are compact, digital sensors that convert sound into electrical signals. Compared to conventional microphones, MEMS microphones offer a number of advantages including:

• Smaller size
• Lower cost
• Improved durability
• Reduced susceptibility to electrical interference.

As such, they provide great potential for use in construction noise monitoring as a more practical and cost-effective option to traditional sensors. However, they also introduce challenges. Existing international standards for sound level meters (IEC 61672) were designed for traditional analogue systems, meaning MEMS-based devices must be carefully tested and validated to ensure compliance.

Traditional monitoring solutions with Class 1 level sound meters can be expensive and labour-intensive. This creates a need for more scalable, automated, and cost-effective approaches, particularly as expectations for transparency and environmental performance continue to rise.

Aligning with relevant international noise-monitoring standards

Most project noise management plans in Australia and New Zealand refer to the international standard IEC 61672 Electroacoustics – Sound level meters, which is the global standard.
This standard defines performance requirements (Class 1 and Class 2); testing procedures for certification; and ongoing calibration and verification processes. The detail of these three parts covers:

Part 1: Specifications – this gives electroacoustical performance specifications for sound level meters. Two performance categories, Class 1 and Class 2, are specified. Tolerance limits for Class 2 specifications are greater than, or equal to, those for Class 1. Sound level meters conforming to the requirements of this standard have a specified frequency response for sound incident on the microphone from one principal direction in an acoustic free-field or from random directions.

Part 2: Pattern evaluation tests – provides details of the tests necessary to verify conformance to all mandatory specifications given in Part 1 when seeking pattern approval. The aim is to ensure that all testing laboratories use consistent methods to perform pattern evaluation tests.

Part 3: Periodic tests – describes procedures for periodic testing for sound level meters conforming to the Class 1 or Class 2 requirements of IEC 61672-1:2002. The extent of the tests in this part of IEC 61672 is deliberately restricted to the minimum considered necessary for periodic tests. Periodic tests described in this part of IEC 61672 apply to sound level meters for which the model has been, or has not been, pattern approved by an independent testing organization responsible for pattern approvals and in accordance with the test procedures of IEC 61672-2:2003.

The research and testing was conducted in line with these standards. Because MEMS microphones integrate digital processing at the chip level, some traditional test methods cannot be applied. As a result, the SiteHive Hexanode was designed to achieve Class 2 equivalent performance.

Initial prototype and testing

The first prototype used a commercially available microphone array to evaluate:

• Sound level measurement
• Direction of arrival (DOA) capabilities.

The key findings from early testing were:

• Performance was inconsistent at certain frequencies
• Accuracy fell outside acceptable tolerance ranges at low frequencies
• A hardware limitation caused signal loss above 8 kHz.

This initial testing highlighted that a custom-designed solution would be required to meet the necessary performance standards.

Development of a viable product solution

To overcome the limitations of the prototype, the team developed a custom PCB (printed circuit board) and evaluated multiple microphone configurations. Key design improvements included:

• Increasing the sampling rate to capture a wider frequency range
• Selecting a more suitable MEMS microphone
• Optimising signal processing using digital filters.

A PDM (pulse density modulation) microphone with a sinc3 filter was ultimately selected as the best-performing configuration.

Calibration, validation and results

Extensive testing was conducted in collaboration with the National Measurement Institute (NMI). The key steps were:

• Calibration using a reference sound pressure level (94 dB at 1 kHz)
• Measurement of frequency response in controlled environments
• Development of a digital “flattener” algorithm to correct microphone response.

The results showed:

• Consistent performance across multiple devices
• Frequency response within IEC 61672 Class 2 tolerances (after correction)
• Successful validation under accredited testing conditions.

This demonstrated that MEMS microphones, when properly calibrated, can meet industry standards for construction noise monitoring.

Beyond calibration: advanced testing

After achieving compliant frequency responses, the next phase focused on broader performance testing, including:

Directional response

Assessing how accurately the device measures sound from different angles, as this is critical for identifying noise sources and the ability to map the direction of arrival of sound.

Level linearity

Ensuring the device responds consistently across a wide range of sound levels.

Windscreen performance

The SiteHive Hexanode devices must operate in harsh construction environments, which introduces additional design challenges such as:

• Protecting sensors from weather and physical damage
• Minimising the acoustic impact of protective materials
• Maintaining measurement accuracy in all conditions.

To evaluate how protective coverings affect acoustic accuracy, multiple windscreen designs were developed and tested to balance structural durability, weather resistance and acoustic transparency.

These tests were conducted in collaboration with the Centre for Audio, Acoustics and Vibration, at the University of Technology Sydney, Australia. The Centre’s advanced technology facilities, including anechoic chambers, provide a highly controlled environment with precise equipment to ensure reliable results.

Key insights from the research

This research demonstrates several important findings:

1. MEMS microphones can achieve standards-compliant performance with appropriate design and calibration.
2. Digital signal processing plays a critical role in improving accuracy.
3. Testing MEMS-based systems requires adapting traditional methodologies.

What this means for construction monitoring

The development of MEMS-based sound level meters represents a significant shift in how environmental monitoring can be delivered. The benefits include:

• Lower costs per device
• Easier deployment of devices
• Enabling of greater deployment density across sites
• Faster identification of noise issues.

Ultimately, this enables more proactive and data-driven management of construction impacts.

Next steps

Future work will focus on:

• Completing full IEC 61672 testing where applicable
• Refining directional and linearity performance
• Optimising windscreen and housing design
• Improving test repeatability and validation methods.

These developments will continue to enhance the reliability and capability of MEMS-based monitoring systems.

Conclusion

This research marks an important step in the evolution of construction noise monitoring. By combining MEMS technology, advanced signal processing, and multi-sensor integration, the SiteHive Hexanode demonstrates how modern digital systems can deliver accurate, scalable, and cost-effective environmental monitoring solutions.

As urban construction continues to grow, innovations like this will play a key role in helping projects meet regulatory requirements while minimising impact on surrounding communities.

Acknowledgments

SiteHive would like to acknowledge the innovators in industry who have contributed to testing and providing feedback and direction to the SiteHive team in the development of the SiteHive Hexanode. Technical specialist consultants from Pulse White Noise Acoustics, Acoustic Logic, Talis Consultants, Acoustic Studio, EMM and SLR Consulting all contributed valuable technical feedback and input through development. Support from construction contractors in field testing the device was invaluable, thanks to Laing O’Rourke, McConnell Dowell, CPB Contractors, John Holland, Gamuda, Built, and early projects Western Sydney Airport Bulk Earthworks and the Level Crossing Removals in Melbourne. Finally special thanks to Chris Thomas and Laurence Dickinson at the National Measurements Institute for the ongoing guidance, support and collaboration throughout the project.

Read or download the full Research Paper

FAQs