Air Quality Monitoring

Project Title

Project Area of Specialization

Information & Communication Technology

Project Summary

The effects of air pollution are becoming evident in Pakistan due to the rapid increase in industrialization, urbanization, deforestation and ever-growing number of vehicles. Since, Pakistan is majorly dependent on primary non-renewable energy sources, these activities massively contribute to air pollution hence having destabilize the emission trends across the country. According to the World Health Organization (WHO) around 200 people per 100,000 population die due to environmental factors/changes in Pakistan. (Health Organization, 2018) Specifically, the WHO Global Health Observatory estimated that the morbidity rate due to outdoor air pollution accounts for 25 deaths per 100,000 population and indoor air pollution accounts for 30 deaths per 100,000 population. (Organization, 2018) Therefore, it is highly imperative to have a network of air quality monitoring systems which can be used to study the air quality in Pakistan. Currently, there are only 6 air quality monitoring systems in Pakistan in major cities such as Karachi, Lahore, Islamabad and Peshawar. (IQAir, 2017) Air quality monitoring integrated with IOT allows us to carry out analysis and prediction on big data. LoRaWAN technology is well suited to develop a network architecture for the air quality nodes deployment across cities. LoRa and LoRaWAN is playing an integral part in digitization of the cities around the world due to its ultra-low energy consumption, wide coverage
area from 5km to 15km for urban to rural setting respectively, and cost effective due to cheap infrastructure, license free frequency operation, low maintenance cost and low
cost base stations to make the application more viable. Hence, the proposed solution for air quality comprises of a single channel LoRaWAN gateway. The end nodes are
integrated with sensors to measure Smoke, Carbon Monoxide, Hydrogen, Ozone, Hydrogen Sulfide, Sulfur Dioxide, Nitrogen Oxides, Particulate Matter, temperature and humidity. The measured sensor data is transmitted to the LoRa network server, The Things Network, through the LoRa gateway to enable access to data on the internet. The data available is then projected on the application server, All Things Talk, which is compatible with the network server, hence allowing to use the data for analysis andprediction through tables, figures and plots. Machine learning is applied to the gathered data to predict the emission trends and how to control the ongoing environmental hazard. As a result, more informed decisions can be made in daily life and policies can be implemented to make air breathable again by cutting down and/or restricting the harmful human activities.

KEYWORDS: Air Quality Monitoring, LoRa, LoRaWAN, Cloud, Network Server, Application Server, Big Data, Data Analytics, Data Prediction, Machine Learning.

Project Objectives

The aim of this project is to develop an air quality monitoring system which is able to determine the ambient air quality. Furthermore, to give a tangible form to such a concept LoRaWAN is being used due to its IoT related significant advantages over other IoT standards.In addition to this, future air quality prediction will be predicted based on the measured air quality data through machine learning so that the trends of the air pollution and the health of the air can be understood; this in turn will allow to improve the health benefits not only for the millions of the inhabitants of Earth but the Earth itself as well.

Project Implementation Method

The methodolgy used to implement the project is as follows:

From the communication perspective, IoT applications require long range, low data rate, low energy consumption and cost effectiveness. These requirements inevitably lead to the emergence of low power wide area networks (LPWAN), which are ideal due to its low power, long range, and low-cost
communication i.e. 10km to 40 km in rural areas and 1km to 5km in urban areas. Moreover, it has an efficient energy consumption i.e. 10+ years of battery lifetime (Mekki, 2018).LoRaWAN is based on LoRa Technology which is designed to connect devices wirelessly to the internet. Its frequency of operation is unlicensed which makes it very
useful and it uses 868 MHz in Europe, 915 MHz in North America and 433 MHz in Asia, from available ISM band. It has a bandwidth of 125 kHz and 250 kHz along with data rates of up to 50kbps (kilobits per second). Bidirectional but half duplex communication is possible along with the AES 128b encryption (SEMTECH, 2020). The bidirectional communication is essentially provided by the chirp spread spectrum (CSS), a modulation technique which spreads a narrow band signal over a wider channel bandwidth. The resulting signal has a low noise level along with high resilience towards interference. In comparison to other modulating techniques it has an increased link budget and is much robust towards the interference since it uses its entire bandwidth to transmit a signal. LoRa uses six different spread factors (SF7 to SF 12) which enable to adapt the data rate and range.

Once the LoRa infrastructure is implemented the PM2.5,PM10 and Carbon Monoxide sensors (since these are the major pollutants in the atmosphere) will be ntegrated with the loRa infrastructure so that the sensors can measure the concentration of the air pollutants in the air. The gathered dataset is then used to measure the air quality index by using the known concentrations of the air pollutants. For calculating the air quality index an algorithm/calculator will be designed which based on the available standards for calucations. In my project i will be using US EPA AQI calculation standards as reference. Once the AQI is known, the data is sent to the cloud such as Amazon AWS or The Things Network, so that the data log can be accessed remotely. The data log can be downloaded as well for data analytics. Furthermore, for data analytics, machine learning algorithm for predictive air quality is used to predict future air quality of the region of interest. The air quality will be accesible by the population so that they can keep track of air quality and take appropriate actions for their safety and stay updated on the environmental health.

Benefits of the Project

When it comes to the air pollution crisis around the globe; the effort and the financial capital being invested in the purpose does take into account the key stakeholders that are affected by the progress made in this regard. The key stakeholders involved which will be directly affected by the outcome of this project are as follows:
i General Public: The general public will be affected since they will be able to make more informed decisions in their daily lives. They will become more aware of what is present in the air around them and what precautionary measures they should take to prevent themselves from it. Furthermore, being aware of the actual intensity of the crisis will also result in them being more conscious of their contribution to the air pollution.

ii Environmental protection Agencies and governmental bodies : The environmental protection agencies and governmental bodies will be able to mitigate the rising
crisis by carrying out a thorough analysis of the gathered dataset. This dataset will educate them of the air pollution sources and health impacts of it on the state citizens.
As a result this will allow them to draft and implement regulatory standards which will aid in reducing the air pollution. In case a precautionary measure needs to be taken and it is not available then the policies by the government will
create opportunities for manufacturing and innovation of relevant resources.

iii Earth’s ecosystem: The preventive measures will as a consequence allow the earth’s ecosystem to gradually normalize itself. Once normalized, the situation will be in control and a lot of resources which were solely being put into action to contain the threat can be pulled away again and used for some other crisis. As a result, the overall ratio of efficiency for the utilization of resources will increase.

The impact of the proposed solution is as follows:
i Society: The society will benefit by being more aware of their contribution to the air pollution. It will also benefit by breathing clean air as a result of improvement in air quality through the relevant interventions such as government and/or environmental protection agencies. The society will experience longer lives, better quality of life, lower medical expenses, fewer school absences, and better worker
productivity and improved standard of living since less respiratory and other related diseases will reduce as a result of clean air.

Economic: There are three different market impacts of air pollution: reduced labour productivity; increased health expenditures; and crop yield losses. Currently, all these factors project a GDP that is below expected levels due to the costs encountered. The economic benefit of the proposed solution will result in better work productivity, reduction in expenditure on health and healthy crops. This essentially
means that the world GDP will overall increase.

Technical Details of Final Deliverable

The design approach used to give a tangible form to this project is described in this section. In order to understand how LoRaWAN is working as the backbone for this project it is ideal to first understand the LoRaWAN architecture. The end nodes integrated with sensors are deployed at the air pollution sources and other desired target locations where the air quality needs to be determined. These end devices/nodes are able to wirelessly communicate to the LoRa gateway through LoRa modulation. Once the data is transmitted from the end devices to the gateway, it is forwarded to the network server which is essentially a cloud storage platform for IoT devices. Once the data is on the cloud it is then displayed onto the application server which gives more meaning to the dataset by forming appropriate graphs and finding appropriate trends. Python for machine learning is then used to predict future air quality trends based on the original graph and trends formed and a co-relation can be derived.

The end node design comprises of all the necessary sensors which are used to determine the concentration of the air pollutants present in the atmosphere. The sensors are interfaced with the Arduino MEGA 2560 board. Furthermore, a LoRa Dragino shield which comprises of the RF9x module (SX127x SEMTECH chip) is also interfaced with the Arduino board to carry out LoRa modulation. Once the end node is powered, it will be able to transmit the measure data to the LoRa gateway. The hardware components of the end node are as follows:

Arduino MEGA 2560.
LoRa Dragino Shield.
DHT-22: Temperature & Humidity Sensor.
MQ-7: Carbon Monoxide Sensor.
MQ-135: Ammonia, Benzene, Alcohol, Smoke.
MQ-131: Ozone Sensor.
Hydrogen Sulphide Spec Sensor.
GP2Y1010AU0F: Dust Particle Sensor.

The schematic diagram for the end node is shown in the Figure below, which comprises of all the aforementioned sensors.

Each sensor interfaced with the Arduino uses its corresponding library. The DHT22 uses the Adafruit Library, MQUnifiedsensor Arduino Library for MQ sensors, spec H2S Sensor library for Arduino, Sharp GP2Y Dust Sensor Arduino Library and LMIC Library for LoRa Shield. The sources for the libraries have been referenced in the appendix, along with the source codes. The LoRa Shield is responsible for transmitting the data from the sensors to the gateway. The LMIC Library is built for this specific purpose by IBM. A gateway is a very important device in communication and networking. In the context of IoT the gateway is responsible for forwarding packets back and forth between the LAN and WAN on the IP layer. The LoRa gateway is set up for carrying out the same function. It is responsible for forwarding the data coming in from the end nodes to the cloud service which is essentially a network server. The hardware components required for the gateway setup are as follows: