RAJA VARA PRASAD et al.

: REAL TIME WIRELESS AIR POLLUTION MONITORING SYSTEM

REAL TIME WIRELESS AIR POLLUTION MONITORING SYSTEM

Raja Vara Prasad Y1, Mirza Sami Baig2, Rahul K. Mishra3, P. Rajalakshmi4, U. B. Desai5 and
S.N. Merchant6
1,2,4,5
Department of Electrical Engineering, Indian Institute of Technology Hyderabad, India
E-mail: [email protected], [email protected], [email protected], [email protected]
3,6
Department of Electrical Engineering, Indian Institute of Technology Bombay, India
E-mail: [email protected], [email protected]

Abstract bead, photo ionization, and solid-state [3]. The existing

Air pollution has significant influence on the concentration of monitoring system largely uses smart transducer interface
constituents in the atmosphere leading to effects like global warming module (STIM) with semiconductor gas sensors, which uses the
and acid rains. To avoid such adverse imbalances in the nature, an 1451.2 standard. STIM was found to an efficient monitoring
air pollution monitoring system is utmost important. This paper
system but for the power requirements and ability to expand for
attempts to develop an effective solution for pollution monitoring
using wireless sensor networks (WSN) on a real time basis namely large deployment. One of the large scale sensor networks for
real time wireless air pollution monitoring system. Commercially monitoring and forecasting is Environment Observation and
available discrete gas sensors for sensing concentration of gases like Forecasting System (EOFS) [6], but the size of the system and
CO2, NO2, CO and O2 are calibrated using appropriate calibration initial cost is too high. Air pollution monitoring system based on
technologies. These pre-calibrated gas sensors are then integrated geosensor network with control action and adaptive sampling
with the wireless sensor motes for field deployment at the campus and rates proposed in [7] also cannot be vast deployment due to high
the Hyderabad city using multi hop data aggregation algorithm. A cost.
light weight middleware and a web interface to view the live pollution
data in the form of numbers and charts from the test beds was The objective of this work is to come up with cost effective,
developed and made available from anywhere on the internet. Other reliable, scalable and accurate real-time air pollution monitoring
parameters like temperature and humidity were also sensed along system with wireless sensor networks. Commercially available
with gas concentrations to enable data analysis through data fusion electrochemical and resistive heating type sensors were used to
techniques. Experimentation carried out using the developed wireless sense the gases like O2, CO2, CO and NO2. Appropriate
air pollution monitoring system under different physical conditions calibration technologies were developed to calibrate these
show that the system collects reliable source of real time fine-grain sensors, which are then interfaced to wireless sensor motes.
pollution data.
Zigbee based wireless sensor networks with multihop data
aggregation algorithm were implemented to [13] extend the
Keywords:
Wireless Sensor Networks (WSN), Gas Sensor Calibration, Light
range of monitoring area. The calibration technology for the gas
Weight Middleware, Multihop Data Aggregation, Web Interface sensor, system architecture of real time wireless pollution
monitoring system, field deployment and experimentation with
varying physical conditions and various challenges faced during
1. INTRODUCTION the design, development and deployment of the system are
discussed in the following sections.
Air pollution emerged in many parts of the world as a result
of explosive industrial growth. Road transport is also one of the
major contributors of air pollution, which contribute to climate 2. CALIBRATION OF GAS SENSORS
change that has perilous domestic and global consequences [1].
Generation and transport of pollutant materials are governed not Each gas sensor is unique, i.e., though the type and the gas
only by the distributions of their sources but also by the sensed are the same, different sensors may differ from each other
dynamics of the atmosphere. Pollutant clouds are sometimes in terms of the output characteristics. Therefore it becomes
observed traveling along the wind directions [2]. To understand essential to calibrate [5] each and every sensor before interfacing
the involved processes in more detail, we need more thorough to the wireless sensor mote for accurate readings. Calibration of
data on the spreads of fine-grain pollutants and their variations gas sensors are carried out in laboratory by exposing the sensors
with time. An air pollution monitoring system that is to different concentrations of gas.
comprehensive in terms of spatial and pollutant coverage, and is The various stages involved in the calibration process are
relatively inexpensive and autonomous is the priority. shown in Fig.1. A specially design closed chamber (Fig.2) setup
Some of the existing instruments for air pollution monitoring is arranged with temperature at 25o C and at 45 to 50% RH
are Fourier transform infrared (FTIR) instruments, gas humidity. Chamber has a provision for sending gas and taking
chromatographs, and mass spectrometers. These instruments the electrical output. Gas is sent from highly precise equipment
provide fairly accurate and selective gas readings. However high which can maintain constant flow rates with MASS Flow
cost, large size & maintenance cost made them unfavourable for controllers. Precise gas chromatography (GCC) equipment is
monitoring applications on large scale [3]. A gas sensor that is used to measure the PPM (parts per million) of gas which is
compact, robust, with versatile applications, and low cost could flowing through the air tight chamber. The measured raw output
be an equally effective alternative [4]. Some of the gases voltage from the sensor is found to be highly unstable and of
monitoring technologies are electrochemical, infrared, catalytic very low magnitude. Signal conditioning circuits were designed

370
ISSN: 2229-6948(ONLINE) ICTACT JOURNAL ON COMMUNICATION TECHNOLOGY: SPECIAL ISSUE ON NEXT GENERATION WIRELESS NETWORKS AND
APPLICATIONS, JUNE 2011, VOLUME – 2, ISSUE – 2

to stabilise and amplify the measured signal from the sensors The test gas is a composition of 10,000ppm Carbon dioxide gas
during the calibration process. The sensors along with the with balanced N2 and synthesized air with 10% Hydrogen. The
conditioning circuit are placed in the chamber and readings are sensors are calibration with CO2 concentration between 350ppm
noted down for regular PPMs of the gas. Each sensor produces a (atmospheric concentration) and 2000ppm. This type of resistive
voltage value corresponding to the input concentration of gas. sensors is sensitive to temperature changes and room
These observed values are plotted and a characteristic equation temperature of the calibration environment was maintained at
is formulated to map voltage signals into corresponding around 25 degree. The flow rate was maintained at 200 ml/min
concentrations in PPM. The calibration process for O 2 and CO2 in closed chamber during experiment. The gas concentration was
is discussed in detail in the following sections. crosschecked with gas analyser. The gas analyser reported a
good match for values between 350ppm to 1200pm and above
that a 2-5% deviation of gas concentration was observed. Five
sensors of the same types were calibrated. Before testing each
concentration, the sensor has been exposed to atmospheric CO 2
concentration 350ppm for 10 minutes. After each tested the gas
concentration will return to 350ppm before the next test starts.
Fig.1. Various steps in the gas sensor calibration process The measured sensor signal strength was low, typically of
the order of 220mV for the atmospheric concentration of
350ppm for CO2 sensor and was highly unstable. To stabilize
and amplify the measured signal, sensor signal conditioning
circuits with amplifier and filter were used during the calibration
process Fig.4.

Fig.2. Gas chamber used for calibration process

2.1 CALIBRATION OF CARBON DIOXIDE

Fig.4. Signal conditioning circuit for CO2 sensor – TGS 4161
Figaro’s TGS4161 (Fig.3) is a solid electrolyte (a type of
solid state sensor) CO2 sensor with detection range of 350 to
5,000ppm [8]. The sensitive element of the sensor consists of a
solid electrolyte formed between two electrodes, together with a
printed heater substrate. CO2 concentration is measured by
monitoring the change in electromotive force (EMF) generated
between the two electrodes. This type of solid electrolyte sensor
features selectivity towards target gas, small in size, low cost
and has long life expectancy (>10 years)[ 9][15].

Fig.5. Calibration result for Figaro TSG4161 CO 2 sensors

Fig.5. shows the calibration results of the five different

TSG4161 CO2 sensors. It was observed that the output of
Fig.3. CO2 and O2 gas sensors TGS4161 is linear in semi-log scale up to 1000ppm and after
that it slightly deviates and is approximately linear. By curve
fitting, the equations for each of the sensors are arrived at which
Calibration of the TSG4161 sensor was performed in a
is given in Fig.5.
laboratory environment with a sealed gas chamber shown in
Fig.2. The sensors were placed in the container for calibration. Calibration equation for CO2 sensor is

371
RAJA VARA PRASAD et al.: REAL TIME WIRELESS AIR POLLUTION MONITORING SYSTEM

For Sensor1, y = p1*x + p2,

where,
y is the measured voltage value
x is the concentration of CO2
Coefficients: p1 = 0.011992, p2 = -3.6973
For the rest of the sensors, the equation is given in Fig.5.

2.2 CALIBRATION OF OXYGEN SENSORS

Figaro’s Oxygen Sensor KE-25 (Fig.3) is a unique galvanic
cell type oxygen sensor, which was developed in Japan in 1985.
It provides a linear output voltage signal relative to percentage
of oxygen present in a particular atmosphere. The sensor
features long life expectancy, excellent chemical durability, and
it is not influenced by CO2, CO, H2S, NOx, H2 [10]. Moreover, it
Fig.7. Calibration Result for Oxygen sensor KE-25
operates at normal ambient temperature and requires no warm up
time making, it ideal for oxygen monitoring for portable
The calibrated equation for each of the sensor obtained after
applications.
the experiment is given in Fig.7. Having calibrated the sensors,
The setup for the calibration of KE-25 was the same as the the system architecture to which these sensors are interfaced is
calibration of CO2 sensors, except for the oxygen used was discussed in the following section.
generated by an oxygen generator and is diluted with synthetic
air. Two MFCs were used and the K-factor for Oxygen gas with
H2S calibrated was 1.24. K-factor for synthetic gas with
3. SYSTEM ARCHITECTURE OF REAL TIME
Hydrogen calibrated was 1.48. The calibration was performed at WIRELESS AIR POLLUTION
room temperature of around 15oC and the flow rate was MONITORING SYSTEM
maintained at 300 ml/min in sealed gas chamber. The sensor is
tested with different concentrations at the interval of 5% from The real time wireless air pollution monitoring system was
15% to 50%. designed and developed to obtain the fine-grain pollution data of
We have observed that the effect of temperature and the gases like CO2, O2, NO2, CO along with other parameters
humidity has negligible effect on the sensor output. KE-25 gives like temperature, humidity, and pressure. The architecture of the
a very stable output voltage with amplitude between 11-14 mV monitoring system is shown in Fig.9. The design and
in ambient air. We have employed sensor conditioning circuit development of the pollution monitoring system constitutes the
with high impedance stage follow by an amplification gain of following stages.
100 in the calibration process Fig.6. 1. Calibration of gas Sensors (Discussed in section II)
2. Configuring Wireless sensor nodes for air pollution
monitoring
3. Development of middleware
4. Field deployment.

3.1. CONFIGURING WIRELESS SENSOR NODES

FOR REAL TIME AIR POLLUTION
MONITORING
Fig.6. Signal conditioning circuit for O2 sensor – KE-25 The pre-calibrated commercially available gas sensors are
interfaced to wireless sensor motes/modules through the gas
Fig.7 shows the calibration result of the three different KE- sensor board, which are programmed for air pollution
25 O2 sensors. It is observed that the output of the oxygen sensor monitoring application. Libelium WASP motes are used as the
(KE-25) is linear for all range of oxygen concentration. The basic wireless communication module, which comprises of the
relation of the output voltage and gas concentration can be processing unit and the communication unit Fig.8. ADC (analog
represented by to digital converter) ports of the wireless nodes are programmed
O2 = (Va – V0)/ (V100 – V0 ) (1) to periodically sample the various gas sensors interfaced to the
sensor board on a rotational basis. The collected samples are
where, packetized and sent to base station [12] [16] at regular intervals
O2 = measured concentration of O2 gas from each of the sensor node, which forms the mesh network
Va = Output Voltage of the sensor at tested concentration. (Fig.9). To increase the monitoring range, multi hop data
V0 = Output Voltage of the Sensor at 0% oxygen concentration aggregation algorithm [13] was implemented. To configure RF
Xbee module, gain of signal conditioning and other modules on
V100= Voltage of the Sensor at 100% oxygen concentration the WASP system refer [11]. The real time pollution monitoring
test bed was developed and deployed with five node network.

372
ISSN: 2229-6948(ONLINE) ICTACT JOURNAL ON COMMUNICATION TECHNOLOGY: SPECIAL ISSUE ON NEXT GENERATION WIRELESS NETWORKS AND
APPLICATIONS, JUNE 2011, VOLUME – 2, ISSUE – 2

pollution monitoring test bed deployed at IITH campus which is

far away (30 km) from the urban polluted area is shown in
Fig.10. Nodes were placed at different locations of the campus
to monitor the gases like CO2, O2, CO and NO2. Along with the
gases concentration parameters like temperature, humidity, and
pressure were also monitored at these locations. Real time plots
of these gases along with other parameters are plotted with the
Fig.8. Libelium WASP mote and the sensor board with different middleware and the GUI. At IITH campus most of the times,
gas sensors the observed data from all of the motes seems to reflect normal
atmospheric concentration of the gases, which is for CO2 – 350
ppm, O2 – 21%, NO2 – 0.7 ppm and CO-0.1 ppm. Experiments
were performed by exposing some of the sensor nodes in the
network to different physical conditions. Accordingly, one of the
sensor nodes (mote1) was placed near the exhaust of the motor
bike. The collected data from this node shows CO2 concentration
of 900 to 1000 PPM with variation in acceleration of bike. It’s
also been observed that when the motor bike is switched off CO 2
concentration come back to its normal atmospheric level of 350
to 380 PPM after some time. Other important observations are
O2 concentration is decreasing along with temperature and
humidity variations Fig.11. One more experimentation was
carried out were number of people near one of the motes was
increased and the CO2 concentration was found to increase as
Fig.9. Multihop mesh network system architecture for the real shown in Fig.12.
time wireless pollution monitoring system.

3.2 LIGHT WEIGHT MIDDLEWARE AND WEB

INTERFACE
Base station or the sink node receives data at regular
intervals of time from the deployed network. Light weight
middleware is developed for effective storage and retrieval of
data. An application to read data from serial port and convert to
appropriate format is developed using visual studio. Parsed data
is logged in to the database in the form of tables along with the
time stamp of each packet. A web based graphical user interface
(GUI) is developed to view the live data in the form numbers
and charts, which is made accessible from anywhere on internet.
Data is sent to the base station through multihop network.
Data repository with effective, efficient and secured data
transactions is required. This repository should have the facility
of web based service requests as well as sms based mobile
services for effective data access. Data from repository should
also be targeted for sensor web enablement may be a ―sensor Fig.10. Wireless air pollution monitoring system deployed at
wiki‖ for sensor information with geospatial information. IITH campus
Standardization in terms of gas sensor systems is to be
developed like TML [17] protocol for exchanging live Data collected from the next deployment at Kukatpally
streaming or archived data to (i.e. control data) and/or sensor which is highly dense traffic area is shown in Fig.13.The CO 2
data from any sensor system. Data repository with levels are almost at 1600-1800 PPM, we observe minor changes
interoperability (technologies like WIFI, Bluetooth etc..) is also in NO2 concentrations but are under safety limits. Oxygen levels
incorporated. are not very much varying except 0.1 to 0.2 % variations.
Temperature was observed to be low and humidity was slightly
3.3. PILOT DEPLOYMENT OF THE REAL TIME higher. Level of pollutant was significantly high at Kukatpally as
AIR POLLUTION MONITORING SYSTEM AND on the day of measurement there was rain.
OBSERVATIONS:
Pilot deployment of the wireless air pollution monitoring
system was carried out at two different places, namely IIT
Hyderabad campus and at high trafficking area of Hyderabad
city – Kukatpally. The objective of the deployment is to collect
the fine grain pollutant data in these areas. The five node

373
RAJA VARA PRASAD et al.: REAL TIME WIRELESS AIR POLLUTION MONITORING SYSTEM

nodes should be utilized optimally. The sensed data from every

node is sent at regular intervals of time. Data compression and
data modeling algorithms helps in saving the energy of node by
reducing the redundant data. Data should have geospatial
information for better visualization or analysis of data. Sensor
nodes should have the capability of estimation, quantification of
the data. Sleep modes also to be included considering the
statistics of the data received.

Fig.11. Gas concentration for mote-1 near the exhaust of

motorbike

Fig.13. Data collected at Kukatpally using wireless pollution

monitoring system

4. CONCLUSIONS
The importance of real-time wireless air pollution monitoring
system is investigated considering the vital technical and
Fig.12. CO2 gas concentration for a mote surrounded by more economic issues for vast area deployment. Commercially
number of people available gas sensors were calibrated using the appropriate
calibration technologies. These pre-calibrated sensors are then
3.4. ISSUES FACED DURING THE DESIGN AND interfaced with the wireless sensor motes forming multi hop
DEVELOPMENT OF THE SYSTEM mesh network. A light weight middleware and web based
interface were developed for online monitoring of the data in the
Air pollution sensors like CO2, CO and NO2 are resistive form of charts from anywhere on internet. Pilot deployment of
heating based sensors. They consume a lot of energy from the the system was carried out at the campus and at the Hyderabad
battery of wireless nodes which is detrimental to network life city. Experimentation carried out using the developed wireless
time. Chemical /MOSFET sensors need very less power but the air pollution monitoring system under different physical
cost is too high. The effect of temperature and humidity on conditions show that the system collects reliable source of real
resistive type gas sensors is to be considered for accurate time fine-grain pollution data.
readings. Calibration at periodic intervals is necessary but it is
difficult to do for large no of sensors in field. Life of the sensors REFERENCES
is very short (typically 6 to 9 months)[14]. Also the costs of
Libelium motes were too high. In the next phase of the project, [1] Department of Transport, ―Transport ten year plan‖ The
we plan to design and develop cost effective architectures for Stationary Office, London, 2000.
pollution monitoring systems. Energy of the wireless sensor

374
ISSN: 2229-6948(ONLINE) ICTACT JOURNAL ON COMMUNICATION TECHNOLOGY: SPECIAL ISSUE ON NEXT GENERATION WIRELESS NETWORKS AND
APPLICATIONS, JUNE 2011, VOLUME – 2, ISSUE – 2

[2] J. C. St. John and W. L. Chameides, ―Environ. Science [11] Wasp mote technical guide-libelium system,
Techno 31‖, pp. 2797-2804, 1997. http://www.libelium.com/support
[3] IST, Chapter 1—Introduction ―International Sensor [12] Sung-Hwa Hong, Byongguk Kim, and Doo-Seop Eom, ―A
Technology‖, pp.1–4. Base-station Centric Data Gathering Routing Protocol in
[4] D. D. Lee and D. S. Lee, ―Environmental gas sensors,‖ Sensor Networks Useful in Home Automation
IEEE Sensors J., Vol. 1, No. 3, pp. 214–215, 2001. Applications‖, IEEE Transactions on Consumer
Electronics, Vol. 53, No. 3, pp. 945 – 951, 2007.
[5] Nihal Kularatna, and B. H. Sudantha, ―An Environmental
Air Pollution Monitoring System Based on the IEEE 1451 [13] Mohd Adib B. Sarijari, Rozeha A. Rashid, Mohd Rozaini
Standard for Low Cost Requirements‖ IEEE Sensors Abd Rahim, Nur Hija Mahalin, ―Wireless Home Security
Journal, Vol. 8, No. 4, 2008. and Automation System Utilizing ZigBee based Multi-hop
Communication‖ IEEE 2008 6th National Conference on
[6] Xu, N, ―A Survey of Sensor Network Applications,‖ IEEE
Telecommunication Technologies and IEEE 2008 2nd
Communications Magazine, Vol. 40, No.8, pp. 102-114,
Malaysia Conference on Photonics, Putrajaya, Malaysia.
2002.
[14] S.Bicelli , A.Flammini , A. Depari , D. Marioli , A.
[7] Young Jin Jung, Yang Koo Lee, Dong Gyu Lee, Keun Ho
Ponzoni , G.Sberveglieri , A.Taroni, ―Low-Power Carbon
Ryu, Silvia Nittel ―Air pollution monitoring system based
Monoxide MOX Sensors for Wireless Distributed Sensor
on geosensor network‖, IEEE International Geoscience
Networks‖ Instrumentation and Measurement Technology
and Remote Sensing Symposium, pp.III-1370–III-1373,
Conference - IMTC 2007 Warsaw, Poland, 2007.
2008.
[15] Wen Wang, Taehyun Kim, Keekeun Lee, Haekwan Oh,
[8] Gas board technical guide – Libelium system,
and Sangsik Yang, ―Development of a New Wireless
http://www.libelium.com/support
Chemical Sensor for CO2 detection‖ IEEE SENSORS
[9] Muna Acosta, Designing with Electro-Chemical Sensors, Conference, 2007.
National Instrument Application Note AN-1798. [10]
[16] Gongbo Zhou, Zhencai Zhu, Guangzhu Chen, Ningning
―Electrochemical CO sensors‖, Targeted Wind Sensing.
Hu, ―Energy-Efficient Chain-type Wireless Sensor
[10] http://tws.unh.edu/index.php?option=com_content&task=vi Network for Gas Monitoring‖ Second International
ew&id=63& Itemid=7 Conference on Information and Computing Science, 2009.
[17] http://www.opengeospatial.org/standards/html.

375