Developing an Innovative Urban Traffic Noise Solution

  Read in   |   Print Print

"LabVIEW played a vital role in our proposed traffic noise solution prototype design. The entire process— from design, to implementation, to testing—was seamless."

- Apoorv Agha, Nanyang Technological University

The Challenge:
Developing an automated traffic noise monitoring and surveillance system to aid government authorities in efficiently enforcing against illegally modified vehicles that produce excess noise on roads.

The Solution:
Using a high-performance dual core processor and FPGAs with CompactRIO hardware and LabVIEW system design software, we developed a rugged and reliable traffic noise monitoring system for overhead bridge deployment that continuously picks up sound from vehicles in a monitored lane, performs real-time analysis to compute the sound pressure level (SPL) of acquired audio, and accordingly triggers the on-board high-speed camera to capture the vehicle plate number.

Apoorv Agha - Nanyang Technological University
Prof Woon Seng Gan - Nanyang Technological University

A Severe Yet Underestimated Threat

According to a World Health Organization (WHO) study, noise is an underestimated threat that can cause many short- and long-term health problems, such as sleep disturbance, cardiovascular effects, poor work and school performance, and hearing impairments. The WHO guidelines for community noise1 are less than 30 dBA in bedrooms during the night for high-quality sleep, and less than 35 dBA in classrooms for optimal teaching and learning. According to European Union statistics, nearly 40 percent of the population in its countries is exposed to road traffic noise at levels exceeding 55 dBA; 25 percent is exposed to levels exceeding 65 dBA during the daytime; and more than 30 percent is exposed to levels exceeding 55 dBA at night. The National Environment Agency (NEA) of Singapore identified construction and traffic noise as major contributors to noise pollution. NEA statistics reveal that the problem has worsened in recent years (Figure 1).

Existing Technology and Its Shortcomings

To continuously monitor vehicle noise emission levels and maintain a quality living environment, the NEA defined motor vehicle exhaust noise level standards for each category of vehicle on the road. Vehicles found noncompliant with the stipulated standards can be fined by the authorities. The NEA provides a hotline number, complaint email ID, and smartphone app called myENV for people to use to report noisy vehicles in their areas. Depending on complaint volumes from different areas, the NEA, in collaboration with the Land Transport Authority (LTA) and traffic police, conduct roadblocks near complaint areas and manually check the tailpipe SPL of each vehicle to identify offending vehicles. This method of enforcement is manual, random, and conducted fewer than 10 times a year because of the costs incurred per roadblock.

Our Motivation for Innovation

The increasing concern towards urban noise pollution and present law-enforcement technique shortcomings highlight the need for an automated, cost-effective, and efficient solution. We built automated,  stand-alone system prototype to deploy on overhead bridges for single-lane monitoring on expressways. Every vehicle in the monitored lane is scrutinized for tailpipe emission noise, and if a violation is found, the system triggers an onboard camera to capture the vehicle license plate number. The gathered evidence includes audio and video, a picture of the vehicle’s license plate, and log files containing SPL values with date and time stamps. Authorities use the gathered-evidence report to issue an inspection notice to the suspected vehicle for possible illegal modification.

Hardware Setup

We designed the current system prototype for deployment in overhead locations mostly on expressways (Figure 2). It consists of two highly directional shotgun microphones hanging vertically from a foldable structure that is welded to a rugged aluminum casing. The horizontal bar houses a GigE high-speed camera for capturing license plate numbers and a wide-angled view camera for recording continuous video. The cRIO-9082 controller sits snugly inside the top compartment of aluminum casing behind the screen (Figure 3). The preamplifiers and power circuitry components are housed in the lower compartment of the casing. All externally exposed parts of the system are waterproof to withstand varied weather conditions.

Hardware and Software Integration

The system is deployed on an overhead bridge with two microphones hanging just above the center of the monitored lane, and a high-speed camera focuses on a spot a few meters away from the microphone. The microphones are connected to preamplifiers with outputs fed to the CompactRIO controller via an NI 9234 DAQ device. The two cameras are connected to the controller via Gigabyte Ethernet ports. The wide-angled camera continuously records video while the high-speed camera is triggered by the controller if a certain threshold is exceeded. With the prewritten DLLs for GigE protocol in LabVIEW, interfacing cameras with the controller was simple. LabVIEW provided a unified platform to interface all hardware components with the controller, reducing low-level coding and saving development time.

The system is implemented in real time, with audio signals being picked up at high sampling rates and processed in blocks. Each block of data is appropriately filtered and processed to compute the SPL. An algorithm compares the computed SPL with the threshold SPL for successive blocks, and based on the decision, the controller generates triggering signals for the high-speed camera to capture the vehicle’s license plate number (Figure 4). With the CompactRIO facilitating high-speed data acquisition using the FPGA, and the high-performance multicore processor asynchronously running multiple threads in parallel, we built a reliable real-time application.

LabVIEW Logic Implementation

We implemented algorithms using the prebuilt LabVIEW function blocks and subVIs. We developed a versatile algorithm to distinguish and suppress interferences from neighboring lanes with nearly 80 percent accuracy. The modular programming approach helped us independently develop several modules and laterally integrate each one into the main program. We implemented the overall logic in two programs online and offline.


LabVIEW played a vital role in our proposed traffic noise solution prototype design. The entire process— from design, to implementation, to testing—was seamless. Using the LabVIEW system design approach, we developed an attractive, user-friendly GUI, as shown in Figure 7.

We deployed the prototype at the PIE overhead bridge near NTU Yunnan Garden in Singapore at different times during the day, with varying traffic volumes and weather conditions, throughout a three-month span. We acquired more than 100 GB of data for testing. We also conducted a demonstration to showcase the system to the NEA of Singapore, where an NEA officer drove a noisy motorbike in the monitored lane to test system functionality. The system successfully tracked the noisy motorbike. Gathered evidence includes the log file with the SPL values and time stamps (Figure 9), a photo of the vehicle’s license plate (Figure 8), audio and video of the noisy vehicle’s operation, and an SPL-versus-time plot (Figure 10).

Benefits of Using an NI Platform

Prewritten LabVIEW DLLs eliminated the tedious task of interfacing different hardware components to work seamlessly on a unified platform. Because the simplified system model, consisting of an NI 9234 module with a cDAQ-9171 chassis and a laptop, scaled to the more rugged CompactRIO platform, we could reuse our hardware. LabVIEW examples, functional blocks, and subVIs were instrumental in quickly implementing algorithms with a user-friendly GUI. LabVIEW forums and prompt online support helped us quickly resolve issues during development. With low-level coding complexity abstraction, we could prototype faster, which helped us reduce development time and devote more time to testing in a real environment, thus gaining confidence in the system.

Future Work

The current system prototype acted as a proof of concept in an operational environment; however, the system is bulky and currently limited to monitoring a single lane. Further improvements to the system will include scaling up for multilane monitoring by building a modular, compact, and easy–to-deploy setup. We also plan to add smart features, such as interfacing with a 3G module for remote monitoring and data transfer to a server, and implementing an automated license plate number identification algorithm and self-calibration algorithm that takes several inputs from a user and calibrates per ISO 362 pass-by noise standards2.

In addition to the current traffic noise surveillance application, this technology can also remotely monitor and generate heavy machinery noise maps on industry floors, automate sound measurements in vehicle workshops/manufacturing plants, and track enemy movements in military operations from a distance. To facilitate further development and transform the system into a commercially feasible product, we applied for the NRF Proof-of-Concept Grant in Singapore. The NEA, ST Kinetics, and NI are confident in our work, and we are optimistic about the possibility to collaborate on system commercialization.


This work is part of the ongoing project at the Digital Signal Processing Laboratory, School of Electrical and Electronic Engineering, Nanyang Technological University, and is funded by the NEA of Singapore.

Author Information:
Apoorv Agha
Nanyang Technological University
S2-B4a-03, 50 Nanyang Avenue,

Bookmark and Share

Explore the NI Developer Community

Discover and collaborate on the latest example code and tutorials with a worldwide community of engineers and scientists.

‌Check‌ out‌ the‌ NI‌ Community

Who is National Instruments?

National Instruments provides a graphical system design platform for test, control, and embedded design applications that is transforming the way engineers and scientists design, prototype, and deploy systems.

‌Learn‌ more‌ about‌ NI