UrbanRest Guardian - Project Introduction

This is my first blog for the Experimenting with Extreme Environments challenge in which we have to design experiments to test a Hammond Manufacturing enclosure under various extreme environments. I have not received the kit yet, it is being held due to customs problems, but I am going to present my project. For my participation I am going to show the UrbanRest Guardian, a project for a problem that affects me personally, the monitoring of nighttime noise in the streets that prevents citizens from resting properly.

Monitoring street noise poses several challenges, with the primary one being the harsh environment where the monitoring device must operate. Additionally, there's the task of accurately identifying the sources of noise while also respecting the privacy of pedestrians.

Addressing Nighttime Noise: Promoting Healthier Environments

I live in Spain. Spaniards are known for enjoying "fiestas" and being very noisy, especially at night. Fortunately, this is changing, and there is a greater awareness of the health problems associated with nighttime noise.

The UrbanRest Guardian project aims to serve as a monitoring aid tool for authorities to regulate noisy nighttime activities in the city, such as nightclubs, garbage collection, and nighttime street work and construction, which can affect citizens' rest.

Current campaign against night noise in my neighborhood

Research from the European Environment Agency highlights how high levels of nocturnal noise can raise stress hormone levels and blood pressure, increasing the risk of cardiovascular diseases. Sleep disturbances caused by noise pollution can result in daytime fatigue, irritability, and cognitive impairment. Chronic exposure to elevated noise levels at night is associated with a higher incidence of cardiovascular diseases and may worsen mental health conditions like anxiety and depression. Moreover, noise-induced sleep disruptions can impair cognitive function and memory consolidation.

It is estimated that in the EU alone, prolonged exposure to such noise leads to thousands of preventable deaths and new cases of ischemic heart disease each year. Thus, implementing measures to reduce nighttime environmental noise is essential for mitigating the risk of cardiovascular problems and promoting overall well-being.

Resources:

• World Health Organization. Regional Office for Europe. (‎2009)‎. Night noise guidelines for Europe. World Health Organization. Regional Office for Europe.
  
• European Environment Agency (EEA) - Beating cardiovascular disease — the role of Europe’s environment - Noise 

The Vision: Building a Network of Nighttime Rest Guardians Across the City

The ultimate goal of this project would be to have a network of night rest guardians strategically distributed throughout the city. These guardians, who monitor nighttime noises through on-the-edge classification, have the task of providing real-time information about the noise sources that disturb citizens' sleep.

Each guardian device will be equipped with sensors capable of detecting and classifying various types of nighttime noise, such as garbage collection, traffic, construction, street music and loud gatherings.

These devices would be strategically located in key locations in the city, including residential areas, commercial districts and recreational spaces, to ensure comprehensive coverage.

The data collected by these tutors will be transmitted to a central monitoring system, where it will be processed and analyzed in real time. Citizens and city officials will have access to this information through an easy-to-use interface, allowing them to visualize noise levels, identify noise pollution hotspots, and understand the impact of different noise sources on the well-being of the city community.

City officials would be able to make informed decisions about noise mitigation measures, such as implementing noise ordinances, adjusting zoning regulations, and improving urban planning practices. Additionally, citizens can use this information to advocate for quieter neighborhoods and take proactive steps to protect their own sleep health.

The system will be able to take advantage of the infrastructure of existing citizen WiFi networks in some European cities. In the image one of the WiFi routers that are throughout my city.

Ultimately, the vision of a network of nightly rest guardians aims to create a more peaceful and calm urban environment for all residents, promoting better quality of sleep.

Addressing Urban Street Challenges with an IP68-rated Enclosure

The UrbanRest Guardian device will meet the challenges of harsh urban environments using an IP68 rated Hammond enclosure. The IP68 classification guarantees solid protection against humidity, dust, impacts and temperature variations. The IP code or ingress protection code indicates how well a device is protected against water and dust.

The kit contains a 1554VA2GYCL Hammond enclosure. It is a watertight clear lid polycarbonate enclosure suited for mounting printed circuit boards or DIN rail mounted components.

The clear polycarbonate lid visibility of the protected internal components and provides protection against access of oil, dust and water. The lid is also secured with self-captivating - M4 stainless steel machine screws, threaded into integral stainless steel bushings for repetitive assembly and disassembly to avoid corrosion caused by dissimilar metals.

Main characteristics:

• UL94V-0/UL94-HB Flammability rating
• NEMA 4, 4X, 12, 13 Protection rating
• Gasket is factory poured using high temperature with silicone rubber gasket material
• Lid screws are self-captivating. This prevents the screws from getting lost or misplaced during maintenance or servicing of the device.

Image source: https://www.hammfg.com/electronics/small-case/plastic/1554

What does UL94V-0/UL94-HB Flammability rating mean?

UL 94, the Standard for Safety of Flammability of Plastic Materials for Parts in Devices and Appliances testing, is a plastics flammability standard released by Underwriters Laboratories of the United States.

• UL94V-0: burning stops within 10 seconds on a vertical specimen; drips of particles allowed as long as they are not inflamed.
• UL94-HB:  slow burning on a horizontal specimen; burning rate < 76 mm/min for thickness < 3 mm or burning stops before 100 mm

IP68, what do those numbers mean?

IP68 is a standardized rating used to indicate the level of protection provided by an enclosure against solids and liquids.

The first digit, "6," signifies the degree of protection against solid particles, with "6" indicating that the enclosure is dust-tight, meaning it prevents any ingress of dust and offers complete protection against contact with solid particles.

The second digit, "8," denotes the level of protection against liquid ingress, particularly immersion in water. A rating of "8" indicates that the equipment is suitable for continuous immersion in water at a depth of 1 meter (3 feet 3 inches) or more, as specified by the manufacturer.

This enclosure will allow our guardians to resist water ingress, prevent dust infiltration, and withstand physical impacts and vandalism, ensuring durability even in high-traffic areas.

What about NEMA 4, 4X, 6, 6P, 12, 13?

NEMA stands for the National Electrical Manufacturers Association. NEMA ratings are standards used to classify the degree of protection provided by electrical enclosures against environmental factors such as moisture, dust, corrosion, and other environmental hazards.

NEMA ratings are designated by a number, which indicates the level of protection offered by the enclosure, and sometimes followed by a letter indicating additional protection features. For example:

NEMA 4, 4X: Watertight. Must exclude at least 65 GPM of water from a 1 in nozzle delivered from a distance not less than 10 ft for 5 min. Used outdoors on ship docks, in dairies, in wastewater treatment plants and breweries. X (as 4X) indicates additional corrosion resistance.

NEMA 6, 6P:  Submersible. Design depends on specified conditions of pressure and time; submersible in water or oil; used in quarries, mines, and manholes.

NEMA 12: General-purpose. Intended for indoor use, provides some protection against dust, falling dirt, and dripping non-corrosive liquids. Meets drip, dust, and rust resistance tests.

NEMA 13: General-purpose. Primarily used to provide protection against dust, spraying of water and non-corrosive coolants. Meets oil exclusion and rust resistance design tests.

Balancing Noise Monitoring with Privacy. Edge Processing Approach

The UrbanRest Guardians will operate by employing an strategy to address the challenge of monitoring nighttime noise without compromising privacy. Instead of recording all the night noise in a given area, which could understandably raise concerns about privacy invasion, UrbanRest Guardian will take a different approach.

The system will process sound at the edge, meaning it will analyze the audio locally rather than recording it all. Through offline classification, the system will identify and will categorize different types of noises in real-time. Once a noise is classified, only the classification of the detected noises will be transmitted, rather than the raw audio data. This will ensure that only relevant information about the types of noises present is shared, minimizing the risk of privacy infringement.

By adopting this approach the system will be able to monitor nighttime noise levels while respecting privacy concerns, providing authorities with valuable insights to regulate noisy nighttime activities without compromising the privacy of individuals in the community.

High Level System Diagram

In the following diagram you can see the main components of the UrbanRest Guardians:

The system is composed of:

• Sound Sensor: Microphone
• Threshold Sound Detector Trigger
• Raspberry Pi 4 Compute Module + Raspberry Pi 4 Compute Module I/O Board
• WiFi Antenna
• MIDAS LCD Display, Alphanumeric, 20X4
• Battery
• IP68 Hammond Enclosure

Simplified Sound Detection Workflow

In order for the device to operate on batteries throughout the night, the system must have low consumption and be active only when necessary.

• The Threshold Sound Detector Trigger detects sounds over a prefixed sound level threshold and sleeps the Compute Module.
• The Compute Module start recording and processing sound.
• Sound is classified and labeled offline using a pretrained model.
• Label classification is logged and transmitted via WiFi to the central server.
• The Compute Module goes to sleep again.

Throughout the entire workflow, the Compute Module sends informative messages to the LCD display.

Sound Classification

Sound waves are transformed into digital format through a process called sampling, where the wave is captured at discrete intervals known as the sampling rate. Each sample represents the amplitude of the wave at a specific time interval. 

The raw audio data is converted in a Mel-frequency cepstral coefficients (MFCCs) representation. A representation of the short-term power spectrum of a sound, based on a linear cosine transform of a log power spectrum on a nonlinear Mel-scale of frequency.

MFCCs are a way of representing sound that tries to mimic how our ears perceive it. The Mel-scale is a way of measuring pitch that's based on how humans actually hear. MFCCs are calculated by taking a sound, breaking it up into tiny pieces, and then measuring how much energy there is at different frequencies within each piece using the Mel-scale.

Then the system extracts several features a pre-trained neural network. Features as temporal characteristics, or frequency components.

After processing the input data (in this case, MFCCs of a sound), the neural network produces an output. A set of probabilities, with each probability representing the likelihood that the input sound belongs to a particular sound class or category.

Location of components within the enclosure.

To test the placement of the components inside the box I will use several techniques as 3D design programs such as FreeCAD or cardboard prototypes.

In the image you can see how tight the enclosure is for the IO expansion module of the Raspberry PI compute module 4. The image is captured using FreeCAD.

Construction of an inner panel mounting

While I wait to receive the kit, I have been making progress in the construction of an inner panel mounting thanks to the detailed plans provided by the manufacturer Hammond.

For the construction of the panel I have used a 3mm thick transparent acrylic sheet. Recycling of a counter separator panel used in a bank during the pandemic.

I have done the machining of the panel with a small Dremel tool.

And I can test the distribution of the elements. For these first tests I use a raspberry Pi 4B.

Roadmap

Finally, this is the roadmap that I will try to follow to successfully reach the end of this new challenge.

1. Introduction
2. Device Construction
3. Sound Level Monitoring
4. Noise Recording Samples
5. Sound Classification Model Training & Field Tests
6. Summary and Final Report

About me

I am a Spanish self-taught programmer. I studied Industrial Electrical Engineering specialized in Robotics, Electronics and Automatic Control in the 80's, but developed my career focused on computer programming. I currently work as a software developer for payment solutions in the payment card industry. I have done projects as a maker, for platforms like Arduino, Raspberry Pi, TI MSP430, TI TIVA C MCUs, Nordic Thingy:91, AMD Spartan-7 and AMD Zynq FPGAs. 

Thanks for reading!

UrbanRest Guardian blog series

Want to know more about this project? Check out the full blog series here:

•  Blog 1 - UrbanRest Guardian - Project Introduction 
•  Blog 2 - UrbanRest Guardian - First contact with the kit components. 
•  Blog 3 - UrbanRest Guardian - Prototype Construction Journey 
•  Blog 4 - UrbanRest Guardian - Classifying Urban Sounds 
•  Blog 5 - UrbanRest Guardian - Remote Monitoring 
•  Final Blog - UrbanRest Guardian - Smart Street Noise Monitor