Design and Development of a Wireless Communication Network Based Ambient Air Quality Monitoring System for Metropolitan Cities

December 2015, marks the 32nd anniversary of the gas tragedy in Bhopal, India. Stagnant weather conditions caused a sharp increase in fatalities and for more than a year almost four people died each day leading an unexpected death toll over 4000 due to the release of 42 tons of toxic Methyl Isocyanate gas. This event is not an isolated incident that affected humans and the environment, as there have been many similar events such as Meuse valley, Belgium, London Smog event etc. Particulate air pollution has been conventionally measured by passing air through a particular filter paper and determining the density of black stain, known as “black smoke method”. Most metropolitan cities are seeing a significant increase in their vehicular registration count and have increased the black exhaust containing CO and other harmful gases along with particulate matter, thus rendering the black smoke method obsolete.


Introduction
December 2015, marks the 32 nd anniversary of the gas tragedy in Bhopal, India. Stagnant weather conditions caused a sharp increase in fatalities and for more than a year almost four people died each day leading an unexpected death toll over 4000 due to the release of 42 tons of toxic Methyl Isocyanate gas. This event is not an isolated incident that affected humans and the environment, as there have been many similar events such as Meuse valley, Belgium, London Smog event etc. Particulate air pollution has been conventionally measured by passing air through a particular filter paper and determining the density of black stain, known as "black smoke method". Most metropolitan cities are seeing a significant increase in their vehicular registration count and have increased the black exhaust containing CO and other harmful gases along with particulate matter, thus rendering the black smoke method obsolete.
The need for a simpler approach to detect and monitor these disease causing harmful pollutants in every polluted area with more concentration over metropolitan cities has never been so imperative and necessitated.

Literature Review
Ambient outdoor air pollution is renowned as a very thoughtful and top global public health issue, both in developed and developing countries. Most of the countries across Asia, Europe and America have failed, to a high degree, to maintain air pollution levels in urban or metro cities within the limit of WHO guidelines. The utmost populous country, China has shown a significant increase in the air pollution. The influence of air pollution on human and animal's health is abundant. Research over this has shown relations between air pollution and premature mortality because of cardio-respiratory disorders and cancers and heart attacks. Even the smaller behaviour of humans like anxiety and anger is linked to air pollution.
With all these diseases, research people came up with solutions and systems to measure air quality monitoring systems. A system designed with a network of tin oxide sensors for indoor and outdoor air quality monitoring [1]. In this paper, innovative processing based on multipleinput-multiple -output neural networks is applied at the network sensing node to obtain temperature and humidity compensated gas concentration values. Some researchers also concentrated on particular area/room air pollution like air quality control for health centres and schools etc., like a system designed for air quality control for health centres [2]. The designed system is based on a set of low cost sensors fortified with a communication module. A multi agent system is used to handle various features of the monitoring system such as: data integration, data analysis, user cantered interfaces, etc... In the same manner, a system for ambient air quality measurement for School area based off the ZigBee wireless network. The system designed measures CO, NO 2 and dust particles [3]. Some researchers designed and proposed systems to measure air quality check by monitoring vehicular emission in traffic signals. The quality of emission of each vehicle is analyzed and recorded individually. If the emission is found to be below the ideals then the number plate will be captured using the installed camera and the information is sent to the nearby traffic police. The complete simulation is carried out on OMRON PLCK [4]. Some researchers have only concentrated on indoor environment air quality check measures [5]. There are multiple papers where authors have concentrated on particular disease causing pollutants like asthma caused due to entry of dust particles in the human body through respiration [6].

Motivation
In order to monitor particular gases that are hazardous to humans *Corresponding author: Renuka Naik, Department of Electrical Engineering, MS Ramaiah University of Applied Sciences, Bengaluru, 560054, India, Tel: 088924 50187; E-mail: renukanaik41@gmail.com respective sensors. The control unit then checks the humidity to see if it lies between 30% RH and 45% RH. This is done because the values of the carbon monoxide, carbon dioxide and methane to determine their concentration in parts per million (ppm) depends on humidity. In this case, if the humidity is between 35% RH and 45% RH, the concentration in ppm is determined by formula set 1.
The controller then checks the real time clock continuously to determine the current time. At predetermined regular intervals determined by the real time clock, the controller sends an SMS to the base station with the values of the pollutants. However, if the humidity is not between 35% RH and 45% RH in the first conditional check, the humidity is checked to see if it lays between 46% RH and 65% RH. If like ozone, carbon monoxide, sulphur dioxide, nitrogen dioxide, methane, lead, particulate matter and other gases, it is crucial that a stand-alone system is to be designed, implemented and deployed. Currently in India, air pollution is monitored by independent third party environmental agencies, academic institutions or the government. It is often seen that severe injuries or fatalities occur even before these agencies and institutions have performed tests and released public reports warning about these hazards.
This work proposes a system aimed at designing an ideal low cost ambient air quality monitoring module which will communicate based on wireless communication networks. The designed system will be able to notify residents and pedestrians in that particular area with the air quality level.

Proposed System Overview
The design described here is a wireless communication network based ambient air quality monitoring system. It mainly consists of two major subsystems as shown in Figure 1, namely: 1. The air quality monitoring module

The base station
The design is derived based on the necessity of an autonomous and ubiquitous system to allow for large scale data collection, which can be studied and used to issue alert automatically in case of unusually high levels of pollutants. A wireless system is chosen in order to facilitate real time data collection without the need for human intervention. A wired system would increase infrastructural cost, especially towards physically wiring a large number of devices to a centralized server. It has no tangible benefits when compared to a wireless setup.

The air quality monitoring module
The air quality monitoring module is responsible for detecting the CO, CO 2 , and CH 4 gases present in the environment along with the temperature and humidity of the corresponding location and communicate with the air quality data monitoring base station as show in Figure 2.

The base station
The base station is responsible for receiving the distress messages from the air quality monitoring module.
The base station forms the back end of the system, receiving data from multiple measurement modules and taking an intelligent decision based on the collective data as shown in Figure 3.

Methodology
The system design simulation is carried out in Proteus ISIS Professional v 8.1 SPI for Arduino UNO development board. Proteus combines mixed mode SPICE circuit simulation, microprocessor models and animated components to simulate microcontroller based designs. Proteus helps simulate the design without the need to actually setup the physical components. This helps in prototyping and testing the design without physically implementing it, thus saving a lot of time in case there is a bug in the design. The system is initialized by powering it on. Upon power up, the peripherals are then prepped and initialized for use. The various sensors which are connected to the MCU are then read. The values from the Carbon Monoxide (CO), Carbon Di-Oxide (CO 2 ), and Methane (CH 4 ) and dust particles are then retrieved from the sensors along with temperature and humidity readings from their  it lies between 46% RH and 65% RH, then formula set 2 is used, else formula set 3 is used. The controller then checks the real time clock to determine whether to send an SMS to the base station. If it determines that not enough time has elapsed since the last transmission, the controller reads renewed values from the sensors and repeats the entire process as shown in Figure 4. If an SMS is sent, the controller again starts reading new values and continues the process explained above.

Calibration of MQ sensors
Where, Where, R S = Sensor Substituting R S in the Equation (1), Computing the R O values and plotting the S O R R against the obtained PPM, a sensor will be calibrated.

Results and Validation
The graph on the left side indicates values calculated from the data sheet for CO 2 , whereas the one on the right shows experimental values for CO 2 . As seen above, the values of R 2 are 0.9961 and 0.8876, which are approximately the same from Figures 5 and 6 respectively.     sensor also shows high congruity between the values of calculated data sheet and experimental values of R 2 at 0.9105 and 0.998 respectively. These readings indicate that the sensors are working within a reasonable range of error as shown in Figures 9 and 10 respectively.
The Table 1 shows the typical values of CO and CO 2 and the experimental results obtained. For CO, the typical value expected in a residential environment is 0-5 ppm, and the experimental detection in clean air is well within the limits. Similarly the experimental values of an incense stick and vehicle exhaust approximate the typical values for CO 2 shows the typical values found in the air outdoors and in residential complexes. The experimental values are found to be within reasonable limits as shown in the table above under detection in clean air, incense stick and vehicle exhaust. Thus, the system is found to be accurate enough to give reliable readings

Future Work
1. An application can be developed in PC to display the plot of the obtained data daily, weekly, monthly and yearly basis.
2. More number of sensors can be interfaced like to detect mercury, sulphur etc.

Calibration of MQ sensors can be done in two ways, one way
is being carried out here by calculating R s /R o can also be done by injecting known amount of gas into the vacuum chamber.

Conclusion
This paper suggest the design, development and implementation of a wireless communication network based ambient air quality      monitoring system for metropolitan cities, as an alternative to the existing traditional air quality monitoring systems facing issues in the area of portability and calibration part.
The system consists of two main blocks 1. The air pollution quality monitoring unit with all the measuring sensors, which continuously monitors the respective areas air quality and keeps logging it and send a message to the base station at the particular instant of time.

2.
A base station, which receives the messages from the air quality monitoring modules placed at different locations and keeps a record of each module on yearly, monthly, weekly and daily bases.