Multiple Gas Sensors System for Environmental and Air Quality

Assessments

A Way to Perform Environmental Monitoring in Smart Cities

Zaher Al Barakeh, Valérie Delbart and Fabien Bonnet

Azimut Monitoring, ZA Alpespace, 15 rue Saint Exupery, 73800, Francin, France

Keywords: Sensors Networks, Multiple Sensors Systems, Air Quality Monitoring, Smart City.

Abstract: In recent years, people are getting more and more concerned with air quality and other environmental

nuisance. Whether it’s indoors or outdoors, humans are getting more sensitive to the issue. Implementing a

more accessible commercial low cost environmental surveillance systems have proven to be a rapidly

growing solution. Therefore, we are specially focused on environmental quality measurements, integrating

environmental and gas sensors in wireless communicating systems in order to provide a wide-range

monitoring solutions suitable for several environmental case study. We have also created different

indicators which convey the wellbeing state in encountered situations where user can rely on these

indicators to assess their environment quality. In this work we present a framework for monitoring real time

particulate matter evolution in a construction site and the implementation of different acoustic noise

measurement systems for mapping the noise evolution in a chosen city. Achieved results show the

effectiveness of such systems in nuisance detection and qualifying living environments.

1 INTRODUCTION

Today’s cities are facing two major evolutions.

Firstly, people are slowly increasing their awareness

of environmental issues, global warming, the

decreasing quality of air and water, the increasing

amount of waste, and other current environment

degradations (Hadfield, 1999). Secondly, the data

tsunami, initiated by the internet and amplified by

excessive smartphone use, implies that citizens are

becoming dependent on a direct and real time access

to any information at any given time.

These deep changes in the way people interact

and take into account their environment are directing

cities to a new concept: the smart city concept

(Paskaleva, 2009). Tomorrow’s cities have to deploy

large and dense sensors networks to furnish local,

accurate, comprehensible and relevant information

to its citizens. This is the major challenge that faces

all sensors networks stakeholders.

This work illustrates a way to surpass this

challenge for the precise case of environmental

urban monitoring. We present our developments

work on a wireless and solar powered multiple

sensors systems that monitor noise, weather and air

quality. After presenting the generic platform, we

focus on enlightening the difficulty to match

sensor’s accuracy and cost issues that are needed for

a large and dense sensors network that delivers high

quality data. Finally, we focus on the must be

performed data processing to transform sensors raw

data signals into relevant and easily understandable

information by citizens.

2 GENERIC PLATFORM

The GreenBee is a multiple sensors system

developed to achieve the monitoring of a large range

of parameters in the environmental field. It includes

the acquisition of noise, temperature, relative

humidity, wind (speed and direction), solar

radiation, ozone concentration and 1 µm particulates

concentration. All these measurements are

performed by a wireless system autonomous in both

energy and communication. Data are sent to our

server by several kinds of wireless networks

solutions. Eventually, the whole system is powered

by a solar cell.

360

Al Barakeh Z., Delbart V. and Bonnet F..

Multiple Gas Sensors System for Environmental and Air Quality Assessments - A Way to Perform Environmental Monitoring in Smart Cities.

DOI: 10.5220/0004711303600364

In Proceedings of the 3rd International Conference on Sensor Networks (SENSORNETS-2014), pages 360-364

ISBN: 978-989-758-001-7

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: A GreenBee with weather sensors. Ozone,

particulates and noise sensors are behind the solar panel.

2.1 Multiple Sensors

Multi-sensors systems in environmental monitoring

have to be able to target different kind of

environmental parameters whom mostly are

incompatible like wind speed and acoustic noise.

Each of these parameters is of different nature hence

the need of different type of sensors. Therefore

developing a system capable of harmoniously

control and integrate different sensors from different

types and technologies could become a major

obstacle that could not be easily overcome. To

resolve this problem, an electronic controller had

been standardized and utilized for all GreenBee®

systems. This controller is compatible with different

sensors technologies by having analog and digital

programmable inputs and outputs that can be used

by different sensors. Onboard processors can be

used to communicate with these sensors by insuring

an adapted signal when needed.

2.2 Multiple Network Protocols

This same electronic controller can be used to send

collected environmental data to our servers. Data

can be sent on a daily, hourly, real time or custom

basis. The communications protocol can be chosen

between, GPRS, Zigbee, Home-Rider, CORONIS …

protocols. Whenever communications are down, and

to insure an uninterrupted data collection, unsent

data can be stocked for a very long time before

communications could be restored.

3 AIR QUALITY SENSORS

The need for an air quality monitoring system is

growing largely. This is mainly caused by

communities’ sensitivity toward the proven

correlation between bad air quality and poor

citizen’s health (Peden, 1996). Mainly, we focus on

two air quality aspects: oxidizing gases like O

and

, and the suspended particles (P.M.).

3.1 Photo Oxidizing Sensor

and NO

gases are generated by different sources.

These sources can be anthropogenic, like gas

emissions from combustion engine and fuel based

energy production as seen for NO

. Or biogenic, like

the rise of temperature and direct contact of air

pollutant with a high sunlight flux, as seen for O

The presence of these gases can be noticed, with

increasing concentration through the last decades, in

every city and urban area around the world. Their

oxidizing effect can be linked to the rise of asthma

and other respiratory health issues (Scoggins, 2004).

Different types of technology already exist to sense

their presence. These technologies can be divided

into two groups: Photoemission, where gas

molecules are stimulated by a photon, and

electrochemical sensing by focalizing on the oxidant

nature to create an electric signal. Sensors using

these technologies defer in resolution, precision and

price. Although photoemission based sensors show

higher performances than electrochemical sensors,

they are hard to miniaturize, are more expensive to

produce, and have higher energy consumption.

Hence their use is in a disadvantage in low cost

multi-sensors units. For these main reasons

GreenBee systems use electrochemical gas sensors.

The following figure shows GreenBee’s

performance in oxide gas detection using an

electrochemical gas sensor in comparison with an air

quality control station operated by the French air

quality surveillance network in the Rhone-Alps

region (Air Rhone-Alps) in the city of Chambery,

Figure 2: Comparison of a GreenBee’s measured O

/NO

concentrations with measured concentrations from an air

quality surveillance system station.

MultipleGasSensorsSystemforEnvironmentalandAirQualityAssessments-AWaytoPerformEnvironmental

MonitoringinSmartCities

361

south-eastern of France. This solution offers an

acceptable precision with a relative error lower than

15%. Such an average precision can be compensated

by a low price and a low electrical consumption,

therefore making these sensors a very propitious

choice to make.

3.2 Particulate Matters Sensor

The term Particulate Matters (PM) signifies a

mixture of solid and or liquid particles suspended in

the air along with surrounding gases. A PM is

characterized by various parameters: chemical,

number concentration, mass concentration, particle

size, granulometric repartition ... Most of the time,

these particles are not homogenous, that is to say it

contains particles of several sizes ranging from a

few nanometers to several tens of microns. By

definition, PM are very heterogeneous in their

chemical composition and are subjected to

equilibrium between gas and particulate phases.

All through the literature we can find different

groups of PM like PM

, PM

… The small number

in the description signify that these particles have an

aerodynamic diameter equals or inferior to the given

number in µm. Hence PM

are a group of particle

with a diameter equal or inferior to 10 µm. The

lower the diameter is, the deeper these particles can

penetrate through the human body, and hence the

more dangerous they are (Polichetti,2009)

The Primequal (French Interagency Research

Program to Improve Air quality at the Local level)

report (2005) points out that the large particles (d> 5

µm) stop in the nasopharyngeal region, particles of 1

to 5 µm in the tracheobronchial region, while fine

particles less than 1 µm, can reach the bronchiolar

and alveolar regions and be persistent. These

particles are the most hazardous for humans;

therefore their monitoring is very interesting for a

life quality point of view.

There are several optical techniques which

provide access to a measure of the amount of PM in

the air. The simplest is the optical absorption with a

visible or infrared light. It is mainly used in the case

of very high concentrations because of its low

sensitivity to low pollution levels. The other main

technique uses scattered lights by PM, it is

nephelometry. The intensity of scattered lights, by

particles of diameter close to the wavelength of the

incident radiation, varies with the number of

particles in the illuminated volume. This technique

has several advantages. It is much more sensitive

than the absorption method and much simpler and

cheaper than the standardized measures. Moreover,

the measurement can be made continuously, a

quality that is sought in the context of such

monitoring. Sadly such systems are cumbersome and

are not miniaturized. Therefore we use this

technology as a calibration and comparison

reference for other type of low cost sensors.

In an objective of measuring health hazardous

particles, we took into ourselves to measure the

number of present particles with a diameter equal or

less than 1µm, hence measuring PM

. Usually, PM

monitoring is made by measuring the mass of these

particles. But in our case we chose to work with

particulates count number in order to obtain

information on the hazardous exposure to these

particles.

A low cost optical sensor has been chosen. Tests

were conducted by burning incense and introducing

different PM emission sources. Sensor’s

measurements were compared with a TSI aerotrak

9306 nephelometry based instrument. The latter

instrument is capable of measuring different size of

particulate matters. Measurements are shown in

figure 3. A high correlation can be noticed between

the prototype sensor particulates count and the TSI

instrument particulates’ count for particles with a

diameter of 1µm and less. Hence the chosen sensor

has shown its potentials for PM

particle sensing.

This prototype sensor had been integrated in a

GreenBee system and put in a construction site

where we were intending to monitor dust and

aerosol emissions. Figure 4 shows the measured

particulates count during 4 days of testing. The first

two days being weekend days, where construction

works were on hold, measured particulates count

was stagnant and therefore we could consider it

Figure 3: Particles count comparison between a TSI and

sensors prototype.

SENSORNETS2014-InternationalConferenceonSensorNetworks

362

Figure 4: PM

evolution in a construction site as measured

by the prototype sensor.

equal to the background emissions. On the other

hand, the following two days witnessed dynamic

measurements where particulates count where

increasing and decreasing along with construction

works.

The prototype sensor response shows a dynamic

sensitivity towards particulate matters emitted by a

construction site. The chosen sensor targets

particulates with a diameter range close to 1µm.

Hence this sensor is suitable for our desired

application on hazardous PM

monitoring.

4 DATA PROCESSING

Low cost multi sensors systems are well in the reach

of common people. Therefore data should be

processed in order to be rapidly understood by users

despite their educational background. Hence a road

map should be drawn in order to define data

utilization, diffusion, and targeted users (pro/public).

Therefore data processing should be rigorous and

unified. Indicators could be used to convey different

information about the severity of a variable. Also

Colors and letters could be used as an easy display

of these indicators in order to obtain an easy

recognition. Incorporating these indicators in a map

could allow for nuisance mapping to pinpoint the

source of certain events. These indicators are

calculated based on time exposition or concentration

exposition when dealing with hazardous gases. The

interval period for this calculation could be a real

time basis or a specified time period (hour, day,

week …).

For example, several GreenBee units were

placed in a city in southern of France. Figure 5

shows an example of 2 units placed on a map, where

each unit is represented in a color and an

alphabetical index describing the severity of the

nuisance surrounding it. In this example, monitored

nuisance is the acoustic noise pollution, and shown

indicators are calculated on a monthly basis. Hence,

using these indicators, users could easily distinguish

a noisy part form a calm part of the city.

Figure 5: Acoustic noise mapping in a city, example using

2 units with color and alphabetical indicators.

Different indicators could be obtained from a

GreenBee unit depending on the number of

monitored parameters. Therefore, we could

aggregate some of these indicators in order to

construct a more general indicator, for example: a

wellbeing indicator.

5 CONCLUSIONS

Advancements in sensors technology like

miniaturization and industrial production are

pushing the development of relatively low-cost

multi-sensors systems to a new limit. These systems

have begun changing the way we live and

communicate with our surrounding environment.

Using these advancements, we have developed a

continuously evolving multi-sensors system for

environmental monitoring. These systems are

autonomous with the help of a solar panel. They are

small in dimension and can communicate in any

communication protocol. They could be used

anywhere to monitor environmental nuisance, even

the hazardous ones. The relative low cost of these

systems could lead to a more realistic pollution or

nuisance mapping in a city or a given urban location.

Such mapping could lead to a better localization and

understanding of pollution and nuisance dynamics in

these locations which hopefully could lead in itself

to a better life quality.

MultipleGasSensorsSystemforEnvironmentalandAirQualityAssessments-AWaytoPerformEnvironmental

MonitoringinSmartCities

363

REFERENCES

Hadfield, L., Seaton, R.A.F., 1999. “A co-evolutionalry

model of change in enviromental management”.

Futures. Vol.31.

Paskaleva, K, 2009. "Enabling the smart city:The progress

of e-city governance in Europe". International Journal

of Innovation and Regional Development : 405–

422(18).

PEDEN, D. B., 1996. “Effect of air pollution in asthma

and respiratory allergy”. Otolaryngology - Head and

Neck Surgery 114. 242-247.

Polichetti, G., Cocco, S., Spiniali, A., 2009, “Effects of

Particulate Matters (PM10, PM2.5, and PM1) on the

cardiovascular system”. Toxicology Volume 261,

Issues 1–2, 30 June 2009, Pages 1–8.

Scoggins, A., Kjellstrom, T., Fisher, G., 2004, “Spatial

analysis of annual air pollution exposure and

mortality”. SCI total Environ. 321(1-3): 71-85.

SENSORNETS2014-InternationalConferenceonSensorNetworks

364