Advanced Air Quality Monitoring System Using

Raspberry Pi
Indranil Banerjee Rajarshee Naskar Kamelia Deb
Electronics and Communication Electronics and Communication Electronics and Communication
Engineering Engineering Engineering
University of Engineering and University of Engineering and University of Engineering and
Management,Kolkata Management,Kolkata Management,Kolkata
Kolkata,India Kolkata,India Kolkata,India
[email protected] [email protected] [email protected]

Debanjan Saha Sanchari Bhattacharjee Deb Kumar Roy

ElectronicsAnd Communication Electronics and Communication Electronics and
Engineering Engineering Communication Engineering
University of Engineering and University of Engineering and University of Engineering and
Management,Kolkata Management,Kolkata) Management, Kolkata
Kolkata,India Kolkata,india Kolkata, India
[email protected] [email protected] [email protected]

Abstract—With an increasing air pollution, we developed II.PREVIOUS WORK

an Advanced Air Quality Monitoring System using
Raspberry Pi 3B+. The system involves monitoring the In [1], graphlab is used for monitoring the air quality in an
air quality by considering parameters like Suspended indoor environment. In [2], the concept of IoT [8] is used to
particulate matter(SPM), Carbon dioxide, Ozone, monitor the air pollution through Raspberry Pi 3B+ and
Carbon monoxide, Smoke, temperature and humidity. takes into account the parameters of PM, CO, CO2,
Particulate matter being a very important parameter Temperature, Humidity and air pressure. However,
gives a clear indication of pollution in that particular pollutants like Nitrogen Dioxide, Sulphur Dioxide, and
time in the area. These pollutant data are extracted Ozone are not taken in it. This may produce faulty results,
using sensors like MQ7, MQ135, MQ9, DSM501A, as the concentrations of these pollutants may be high in
DHT11, etc. Most of these sensors produce analog some areas. [3] Shows the use of web socket and Raspberry
output so an Analog to Digital converter is required Pi 3B+ to monitor pollution but the challenges faced are
before supplying the data to the Raspberry pi 3B+ security of data and Storage Management. In [4], the author
microcontroller. Using software and coding of the explains the raspberry interfaced monitoring process. Most
Raspberry Pi 3B+, the data are analyzed and a graph to of the previous works lack sensors of all kind of pollutants.
show the changes in the locality and time in which the In addition, data security and management needs to be taken
experiment is plotted. The results obtained were verified. into account. Movable objects help in better extraction of
The experiment is conducted in a locality at Rajarhat, pollutant data.
New Town, Kolkata and the results were compared with
that from a local environment control authority. This III.LITERATURE SURVEY
system would help to take real-time decisions and very
effective in today’s situation of high air pollution in most The first system ever made, measures CO, CO2, SO2 and
of the Indian cities. NO2, were calculated using different infrared sensors. The
outcome of the system were stored in a different data server.
Environmental parameters with the help of some aerometric
Keywords—Raspberry pi, Python, Gas sensors, Matplotlib sensors and gas sensors were measured using the
PIC18F87K22 microcontroller. Sensor nodes are set up in
different areas for real time monitoring of environment. The
I.INTRODUCTION results were posted through the city map. Raspberry pi 3B+
was used for implementing the system. Parameters like C02,
Air is one of the important component of our biosphere. The C0, temperature and pressure ware measured but no stress
quality of air is changing day by day. All the was given on particulate matter, which made the work
industrialization affects the air quality so the components of incomplete.
air, such as nitrogen, carbon dioxide, carbon monoxide, Recent development of a nearly perfect air quality
Sulphur dioxide etc. are changing. These changes for gases monitoring system using raspberry pi was done but it did
are affecting the ecosystem. To get an idea of this changing not include any real time sensor for SO2, NO2 and ozone
of quality and the quantity of pollutants over time our that makes a gap in the model.
project is used. This change in quality of the air is affecting
humans as well as other animals and plant health. Due to In our model, we have tried to make an air quality
this high industrialization, large amount of toxins is being measuring system using Raspberry pi 3 B+ microcontroller.
introduced in the air, which are foreign. Here we have used MQ2, MQ7, MQ9, MQ135, DHT11 and

978-1-7281-0070-8/19/$31.00 ©2019 IEEE

DSM501A sensor for measuring flammable gas, Carbon of gas concentration. It has good sensitivity to ozone in wide
dioxide (CO2), Carbon Monoxide (CO), other minor range. MQ 9 is useful for gas detection such as LPG, CO
pollutant gases(like methane), temperature and humidity and and CH4, i-butane, smoke and so on. The sensitivity of the
suspended particulate matter. Most of the previous models sensor [7] can be adjusted by using potentiometer. MQ 7 is
have shown results using sensors but some challenges we used in detecting Carbon dioxide. All of these produce
there regarding the ozone sensing. Here, ozone monitoring analog outputs so an analog-to-digital converter is used.
was done using MQ 131 sensor and the system shows DSM 501A is the sensor that is used for sensing the
effective results. Particulate Matter (PM) in the environment which is a very
important parameter of air pollution. It operates on a wide
IV.SYSTEM DESIGN range of temperature and produces effective result. DHT 11
is used for sensing temperature and humidity.
The Raspberry pi 3B+ is used, with different sensors
connected to it. The system needs to collect data through the
different sensors for a particular locality to monitor the air
pollution. The design involves both hardware and software
components. The parameters of ozone, particulate matter
(PM), Carbon dioxide (CO2), Carbon monoxide (CO), gas,
temperature and humidity are taken into consideration. The
sensors of the MQ series are used, i.e., MQ-131, MQ-7,
MQ-2, MQ-9, MQ-135. In addition to these, DSM-501A,
DHT-11 is used for measuring Particulate matter and
temperature & humidity respectively. These sensors are
connected with the Raspberry pi 3B+, then analyzed by
Python language, and then displayed on a monitor to
understand the air quality of a particular locality.
The sensors mentioned above were already used in various
previous models in this field. However, none of the models
used the ozone sensor along with other sensors to measure
the ozone level of a particular area. Here, in our project, we
have implemented an ozone sensor to make the measuring
system more concrete and trustworthy since ozone sensing
has become an important area in pollution monitoring. We
have used MQ131 sensor to measure the ozone level. As a
whole, this system is a more complete and advantageous
than any previous model, which had been made to measure
the air quality of a certain area. The proposed system
provides low cost, low power, compact in nature and highly Fig:1:Block Diagram of the Air Quality Monitoring
accurate system for monitoring the various parameters of the
environment. System

A. Raspberry Pi 3B+: The left most part shows the sensing devices and they are
connected to the ADC and then to the Raspberry Pi 3 B+
Raspberry Pi 3B+ microcontroller has feature modifications microcontroller. The display is obtained from it to show the
in the memory and in supporting peripheral devices. This results.
version of Pi has got the Ethernet and USB ports. The
Ethernet adapter is connected to an additional USB port. In
the previous models of this microcontroller, i.e., A, A+, and V. SOFTWARE CONFIGURATION
Pi Zero models, the USB port is attached to the system on a
chip. In the model of Raspberry pi used in the pollution Matplotlib
monitoring system the USB or the Ethernet chip contains a
five-port USB hub, but out of the five ports only four ports Matplotlib is a 2D plotting python library. The matplotlib
are available for use. The operating systems supported by contains commands for plotting different types of graphs.
Raspberry Pi are Raspbian, windows 10 and Ubuntu. Here we have used the library to plot real time data with the
Raspberry Pi 3 B+ has quad-core processor. time on x-axis and quantity of the element and compound
present in y-axis.
At first, we downloaded the library of the particular sensor
B. Sensing Devices: for GitHub then install it. After that, import the sensor
library for example for sensor DHT11,
MQ series sensors are used in the system to obtain the data >>>import Adafruit_DHT
that are interfaced with the raspberry pi 3 B+ to determine >>>
the results at last. MQ 131 is the ozone sensor, which Then if on pressing entre the next arrow comes black then it
converts change of conductivity to correspond output signal was ready to go. Then we executed the code for the sensors
with a specific delay of 60 seconds for update. Then on
perfect execution, we open LibreOffice where we opened NO2 PM10 SO2
the saved file, which displayed the values from the sensors. Stations
µg/m³ µg/m³ µg/m³
The matplotlib library is then downloaded and installed.
Then matplotlib is imported and the python code is written
and a real time graph is obtained from the sensors output. Barasat ** Not operated on this date

Barrackpore ** Not operated on this date

DumDum ** Not operated on this date

Khardah ** Not operated on this date

Rajarhat ** 39.00 149.67 4.67

80(D) 100(D) 80(D)
Standard ***
40(A) 60(A) 50(A)

D for daily (24 hours); A for Annual; 8 hours for any 8

hours of a day; 1 hour for 1 hour of a day; NM for not
measured
**This air quality station is operated manually, twice
every week, as per directions of the Central Pollution
Fig.2: Plot showing the variation of pollutant in the Control Board. Data reported are 24 hours average
Morning ending at 6 AM on 27/01/2019
***National Ambient Air Quality Standard-2009`

Here we have measured the amount of pollutant in the air of

a particular area of Rajarhat, New Town, Kolkata and
calculated the level of pollution will exceed from the
standardized limit of air pollution or not. After the
calculation has done, we have plotted the graph based on
generated results. The data obtained from the sensing
devices are compared with the standard [9] and actual data
available. These provide the basis of the change in pollution
levels from what is considered normal. The system we have
made is a low cost and flexible system and modularity of
this model is very high and the generated output is very
accurate than the other contemporary models. A perfect
collaboration between sensors and the mini-computer has
been established by using a compact analog-to-digital
converter, which by using eases the process of data
collection. Using this model, we have gathered data from
crowded and non-crowded places, for example, University
campus and open areas and each time the accuracy of data
was satisfactory. Through this model, we are able to
Fig. 3: Plot showing the variation of pollutants in the measure almost all the common air pollutants in our
Evening environment. Thus, it is one of the compact air-quality
monitoring systems in recent times.

VII. CHALLENGES FACED

VI. CONCLUSION
In our system, we have used Raspberry pi 3B+
1. Collecting Useful Data
microcontroller along with different real time sensors to
Initially collecting the data and accumulate them ware a big
analyze the environmental parameters and then take
task to be performed. Here the data are various kind of gas
appropriate action based on the generated results.
pollutants, which are always varying depending upon the
surroundings. Each molecule has unique feature. Therefore,
it was very difficult to locate them all in one place. The use
of many independent sensors makes our work easy and [7] Erik Wilhelm, Sandra Siby, Yuren Zhou, Xavier Jayaraj
efficiently gathers the data from different areas and helps to Siddarth Ashok, et al., Wearable Environmental Sensors
process them to plot the graphs accurately. And Infrastructure for Mobile Large-Scale Urban
Deployment “, IEEE SENSORS JOURNAL, VOL. 16,
2. Storage Scarcity NO. 22, NOVEMBER 15, 2016.
As per the demand, the storage unit is under tolerance in this
model. However, in future further storage will be needed [8] Sean Dieter Tebje Kelly, Nagender Kumar Suryadevara,
when the amount of data will exceed than its default storage And Subhas Chandra Mukhopadhyay, “Towards the
capacity. That time we need to use additional storage unit Implementation of IoT for Environmental Condition
with the system in order to make the system works Monitoring in Homes” IEEE sensors journal, vol. 13,
flawlessly. no. 10, October 2013.

[9] http://emis.wbpcb.gov.in/airquality/showaqdatanxt.do
VII. FUTURE DEVELOPMENTS

In future, further development that can be done on this

model is by placing the sensors on an object, which is
movable in nature. This model can be arranged near any
industrial or any area where the level of air pollution is so
high that it exceeds the standardized level of air pollution,
so that it can be used to detect the pollutants and then a
websites can be developed which will inform the authority
to reduce the amount of pollution. The sensors, which are
used in our project work, can also be implemented in
various types of drones, which can move around and
identify the polluted agents of the environment. This model
can be also implemented in domestic air conditioning
devices so that it can monitor the condition of air inside a
home. If the amount of dust particle is higher than the
adequate level then it can take appropriate action to measure
the dust particles and reduce the level as much as possible.
Automatic air purification device can also be implemented
by using this prototype model.

REFERENCES

[1] Indoor air quality management using raspberry pi and

Graphlab by Somdeep Acharyya, Sayan Hazra, Atanu
Pal, Nabarun Sengupta.

[2] Air Quality Monitoring System Based on IoT using

Raspberry Pi International Conference on Computing,
Communication and Automation (ICCCA2017).

[3] Air Quality Monitoring System using Raspberry Pi and

Web Socket International Journal of Computer
Applications (0975 – 8887) Volume 169 – No.11, July
2017.

[4] IOT based Indoor Air Pollution Monitoring using

Raspberry Pi International Journal of Innovations in
Engineering and Technology (IJIET).

[5] WHO, “WHO guidelines for indoor air quality: selected

pollutants,” WHO, 2010.

[6] Yifei Jiang, Kun Liz, Lei Tian, Ricardo Piedrahita,

Xiang Yun et al,”MAQS: A Personalized Mobile
Sensing System for Indoor Air Quality Monitoring”,
UbiComp'11 / Beijing, China.