Low power wireless sensor networks in industrial environment

ROBERTO FERNÁNDEZ-MARTÍNEZ; ORDIERES J.; GONZALEZ-MARCOS A.

Project Management Area. EDMANS research group (http://api.unirioja.es/edmans/)
Departamento de Ingeniería Mecánica
Universidad de La Rioja
Luis de Ulloa 20. 26004. Logroño. La Rioja
SPAIN
robfermar@gmail.com

Abstract: - The current situation in Spanish municipal solid waste is a reason of concern for the society because
the consumption habits of modern consumer lifestyles are causing a huge waste problem. Last investigations
show generation of polluting emissions, not just during the dumping period but enough time later. Try to
control these emissions or, even develop a way to recycle waste by generating electricity from landfill waste
and pollution needs to have some kind of information.

News technologies allows us to get some different environmental measures, thanks to the great advanced that
this area has had in the last years, how they can be temperature, pressure, light, etc.. To be able to realize a
processing in the right moment and condition inside landfill, can depend on measures read and whether it can
be used later correctly.

To get some information from the environment doesn’t mean that it can be able to be processed and managed.
It is necessary to locate each sensor in the correct place where you would like to have a right measure, and of
course, it is necessary to know in what moment you take this. To install a complex sensor system that can
report events to a central base station, which can take appropriate action, in geographical locations and under
environmental adverse conditions, similar those that can be found on solid waste landfills, can be an expensive
and complicated task.

In those situations is where new devices called micro-electromechanical systems (MEMS) joined to network
wireless technologies are going to help us resolving our problem doing wireless sensor networks (WSN). These
low cost and low power devices are going to allow us to have some sensor connected between them without
needing to do a wired infrastructure, since it won’t be necessary neither to transmit data nor to give them
energy.

The goal, besides saving all those problems in the installation, is to carry out a collection of data of several
physical magnitudes, store them in a central base station and after analysis of this information, a qualified
person can make decisions about the most convenient action to do a correct management of landfill.

Key-Words: - Wireless sensor networks, Landfill monitoring, Solid waste emissions, MicaZ, Heliomote.

1. Introduction In spite of the great quantity of existent landfills and

Nowadays it is necessary to carry somewhere all
municipal solid waste that have been generated by
human being, and it has driven us to make landfills.
Spain is one of European countries where landfills
more have been used to resolve the problem with
municipal solid waste. However, only a little part of
those belongs to the group of the controlled
landfills.
the efforts carried out by the governments to try to
palliate the environmental effects and to be able to
obtain a reuse from landfill, like it can be the control
of biogas to generate energy, they are still few
landfills that they make a control.
So that they are controlled correctly, besides having
to follow the correct methodology when you treat
waste inside the landfill, as including methods of
contention of grouts, or doing a compression of the
garbage to increase their density, or covering this to
avoid their dispersion, you need information of
system if you want to manage situation correctly.
These days some landfills have gas extraction
systems, where can get gas from biodegradable
materials. In those systems gas is pumping outside
of landfill using perforated tubes and later it will be
burned in a gas motor to generate energy. Burn this
gas is better solution than release it directly to
atmosphere, since this way methane consummation
is carried out, that it is a greenhouse gas much more
noxious than dioxide of carbon.
To help to do these tasks, obtaining some
measurements that can give environmental
information of the area like can be temperature,
atmospheric pressure, quantity of light and other
variables that allow detecting the quantity of certain
gases like can be the methane or O2, can be really
important when you have to make decisions.
2. Wireless sensor network (WSN)
A sensor network is a flexible and powerful
instrument that allows monitoring complex systems,
where to locate the sensors can be impossible in any
other way. The aim of data gathering for sensors in
monitoring is the obtaining of data having as only
limitation sensors features.
Traditionally the obtaining of data coming from the
monitoring was carried out by an individual
installation located in a concrete place that picked
up data every sort period of time. When it was
considered that the number of samples was the
appropriate one, a person was moved to the location
and stored data were picked up for its study.
Then this technology appears and not just allows
picking up the data, but also allows processing and
transmitting them to other place without necessity of
making a movement. Besides thanks to that each
node can be programmed, these can be adapted to
the conditions that environment demands them. And
against what can seem, this technology is less
expensive than old system.
A WSN is a network with several devices
distributed widely that use sensors to control diverse
conditions in different points. Each node is typically
equipped with a wireless communication device, a
small microcontroller with a little memory and some
measurement elements, which allow having the
ability to observe and to react to events and
phenomena in a certain environment.
There are four basic components in a sensor
networks:
• A group of located and distributed sensors.
• An interconnection network, in our case a
wireless network.
• A central station where information are sent.
• And certain computational resources where be
able correct data, make consultations in order to
see the state and even apply mining of data.
Environmental monitoring represents a kind of
sensor network applications with enormous
potential benefits, which is a highly active area of
research devoted to developing the tools and
techniques needed for monitoring. There are a lot of
projects that work in this research [1] [2] [3] [4].
3. MicaZ technology.
In a study of the available possible technologies to
develop a WNS, the Micaz technology was chosen.
This decision was made because this family was the
one that more accessories had to carry out the work,
besides it had very good features.
Fig. 1 Wireless device MicaZ
MicaZ nodes have a chip from chipcon, model
CC2420 that works under IEEE 802.15.4 standard.
It also has a low power microcontroller, an
Atmega128L, where programs can be executed from
available internal flash memory [5].
It is an evolution of mica technology and as it has an
in/out connector with 51 pines, many applications
and sensor boards available for this family can be
used for MicaZ.
These devices have been designed to work with
feeding of batteries, specifically with two type AA
batteries, however other sources can be used
whenever feeding is between 2.7 and 3.6 VDC. The
given batteries are alkaline with a load of 2000
mAh.
4. Sensor boards fed, is also limited because this tension will be
As it was said before, MicaZ technology has one of obtained from feeding system of the mote, in this
the widest ranges of sensors, and here we have case two type AA batteries.
studied several of them.
With MDA300CA happens something similar. It is
possible to connect analogical sensors that will go to
different channels of the converters, as digital
sensors that will connect to the digital channels.
Fig. 2 MTS/MDA Sensor Board
This board has a greater tension range from 0 V to 5
V. Can be used a greater range of sensors, but it
With all these boards and with the sensors integrated
continues having the same problems, because it is
in them, we can obtain useful measures that can be
complicated to work in those tensions ranges.
used for our aim. The problem appears when we
want to make measures with other sensors and that
they are part of our sensor network. 5. Low power networks
In these kinds of networks, it is important to keep in
Sensor board Sensors and features mind the energy consumption, mainly for getting
that life of devices be highest. So our network nodes
Light, Temperature, Acoustic, Sounder, Dual-
Axis Accelerometer and Dual-Axis
won't be connected to a power system and their
MTS310CB
Magnetometer. work mainly with batteries, it has tried to reduce and
to optimize energy consumption inside the node. In
Ambient Light Sensor, Relative Humidity & spite of this, should be kept in mind how long our
MTS400CB Temperature Sensor, Dual-axis Accelerometer node will be correctly fed and when it will stop to
and Barometric Pressure Sensor. work.

The same five basic environmental sensing To make an energy consumption study in a sensor
MTS420CC parameters than MTS400CB and an GPS
module.
network by means of direct measure is not feasible,
due to the great number of nodes and different
consumption level in each of them. Investigations
MDA100CB Light, Temperature and general prototyping area.
made on this area are rather few and most is based
on software simulations [7].
Light, Relative Humidity an extremely versatile
MDA300CA data acquisition board.
Some energy consumptions in these type wireless
devices are:
The boards studied with capacity to connect sensors
are MDA100CB and MDA300CA. In MDA100CB Processor operating current
Normal operation 8 mA
besides having sensors integrated in it, we have a Full operation 12 mA (7.37 Mhz)
prototyping area where we can connect external Sleep operation < 15 μA
sensors. This area supports connection to all eight
channels of the Mote’s analog to digital converter Transmission system operating current
Receive mode 19.7 mA
(ADC0–7). It also has more holes without Transmission mode -10 dBm 11 mA
connection that are used for breadboard of circuitry Transmission mode -5 dBm 14 mA
if it was necessary [6]. Transmission mode 0 dBm 17.4 mA
Idle mode 20 μA
Sleeping mode 1 μA
A problem that can be found is the difficulty of
finding sensors that have features that data Memory operating current
acquisition boards need. It is not possible to connect Write 15 mA
signals that are greater than VCC (3V) or less than 0 Read 4 mA
Sleep 2 μA
V to any of the holes, because it could damage that
board. Sensor operating system
Temperature 0.4 mA
Not only obtained tension from sensor is a problem Humidity 1.5 mA
Light 0.030 + 0.002 * v mA
for sensor board, the tension that the mote should Gas 174.618 + v * 0.011
give to the external sensor so that it works correctly
The MicaZ technology incorporates a fed module
that uses two batteries type AA. The batteries
usually express in mAh, which means that a battery
of 1000 mAh should support a consumption of 10
mA during 100 hours, although in the theory it is a
little less.
Technologies in that batteries are based can be
several, alkaline, lithium, nickel-metal hydride. To
understand the cycle of life of the node is necessary
to see the graph of life of an alkaline battery [8],
since for defect the nodes MicaZ come with this
kind of batteries.
Fig. 3 Test Panasonic Industrial Alkaline batteries
MicaZ node is designed to work to 2.7v, it makes
that it works with two AA batteries. But like can be
seen in the graph, although time depends on the
resistance, voltage given quickly falls below that it
needs.
To solve fed problem is chosen solar energy and it
work system. At the moment, a solar panel can give
more than 15 mW for square centimetre and if this
energy is stored correctly it would be enough to
consume at night that it stored during the day.
Fig. 4 Heliomote
The device selected to solve the problem is one
from Atla Labs and it is called Heliomote. This
device is a system of solar feeding for small devices
and low power. Its main purpose is to harvest
energy from the sun, store the energy in two AA
NiMH batteries, and supply power to the sensor
node.
This device is housed in an environmentally strong,
IP-67 rated enclosure, which is shockproof, dust-
proof, and water-proof up to 6 ft. of water. This
enclosure allows the Heliomote to survive in the
harsh, outdoors conditions [9].
MicaZ technology can adapt perfectly to this device,
being connected through its connector of 51 pins
because it has a material that avoids the static,
protects to the node and adapts it to the size and
position of mote.
The figure 5 shows the Heliomote solar panel at
various insolation conditions ranging from moderate
night to bright day. The peak power output of the
solar panel is 94mW. In a given day the Heliomote
will supply approximately 1800J of energy. A
MicaZ at full operating power is approximately
0.3W. The problem is that a MicaZ at full power
consumes more power than a Heliomote can
produce. It is necessary optimize the software on the
mote so that it doesn't consume more than it can
generate [10].
Fig. 5 Heliomote insolation conditions
6. Base Station
Inside our sensor network it is necessary to have a
device that can pick up data generated by sensors
and can also store them in order to use them later.
Making a study of the possible available
technologies, Stargate SPB400 from Crossbow was
chosen. It is a powerful single board computer with
enhanced communications and sensor signal
processing capabilities [11].
 Microphone

415
410
405
400
395
390
385
380
375
370
365
360

171541
180455
185409
194322
203254
212301
221215
230129
235042
003956
012909
021823
030737
035650
044604
053517
062431
071345
080258
085212
Fig. 6 Stargate SPB400

This device has a Linux operating system that can Fig. 9 Microphone from MTS310
run software that allows reading all the picked up
data from sensors. These data are analyzed and
stored in a database for their later study. Barometric Pressure mbar.

959
The Stargate also has connection to the exterior, 958
957
which makes possible to consult data from any 956
place, only having an Internet connection. 955
954
953
952
7. Data examples 951
950
In the following graphics is observed a sample of 949
the data picked up for several of studied sensors. 948
171535
172252
173017
173741
174506
175231
175956
180721
181445
182210
182935
183700
184425
185150
185907
190632
191356
192121
192838
193603
Temperature ºC
Fig. 10 Barometric Pressure from MTS400
30

25
Relative Humidity %
20
43,5
15
43
10
42,5
5
42
0
41,5
171541
180401
185221
194040
202900
211831
220651
225510
234330
003150
012009
020829
025648
034508
043327
052147
061007
065826
074646
083505
092343

41

Fig. 7 Temperature from MTS310 40,5

171535
172252
173017
173741
174506
175231
175956
180721
181445
182210
182935
183700
184425
185150
185907
190632
191356
192121
192838
193603

Light mV
Fig. 11 Relative Humidity from MTS400
3000

2500
Light lux
2000
160
1500
140
1000 120
500 100
80
0
60
171541
180549
185557
194604
203724
212732
221739
230747
235754
004802
013810
022817
031825
040833
045840
054848
063855
072903
081911
090936

40
20
Fig. 8 Light from MTS310 0
171535
172252
173017
173741
174506
175231
175956
180721
181445
182210
182935
183700
184425
185150
185907
190632
191356
192121
192838
193603

Fig. 12 Light from MTS400

8. Conclusion Networked Sensor Systems. Proceedings of
the 2nd international conference on
Nowadays, the use of wireless sensor networks is Embedded networked sensor systems, 2004,
increasing considerably thanks that every day these pp 214 – 226.
devices have low cost and better features. [5] MPR-MIB Users Manual,
http://www.xbow.com
Every day sensors, wireless communications [6] MTS/MDA Sensor Board Users Manual,
devices and processors are integrated in smaller http://www.xbow.com
devices and consume less energy, what allows using [7] Hyunwoo Joe, Jaebok Park, Chaedeok Lim,
this technology in new applications where before it Duk-Kyun Woo, and Hyungshin Kim,
could not be used. Instruction-Level Power Estimator for Sensor
Networks, ETRI Journal, vol.30, No.1, Feb.
Technologic developments, like bigger duration in 2008, pp.47-58.
batteries life or the possibility of harvesting energy [8] Panasonic Industrial Alkaline batteries,
from environment with solar panels, make that http://www.panasonic.com/batteries
nodes life can be considerably longer. [9] Heliomote User manual,
http://www.atlalabs.com/
In spite of high quantities, these data can be stored [10] Technical Support, http://www.atlalabs.com/
and used later. Thanks to the advances in [11] Stargate Developer’s Guide Rev. B,
computation, it is possible to pick up and study as http://www.xbow.com
many data as the application needs.

All this joined makes that the future of monitoring

goes unavoidably by the wireless sensor networks
and one of these applications can be monitoring a
landfill.

References:
[1] Krishna Chintalapudi, Tat Fu, Jeongyeup
Paek, Nupur Kothari, Sumit Rangwala, John
Caffrey, Ramesh Govindan, Erik Johnson, and
Sami Masri, Monitoring Civil Structures with
a Wireless Sensor Network, Internet
Computing, IEEE, Vol. 10, Issue 2, March-
April 2006 pp. 26 – 34.
[2] Shamim N. Pakzad, Sukun Kim, Gregory L
Fenves, Steven D. Glaser, David E. Culler and
James W. Demmel, Multi-Purpose Wireless
Accelerometers for Civil Infrastructure
Monitoring, The Proceedings of the 5th
International Workshop on Structural Health
Monitoring (IWSHM '05), Stanford, CA,
September 2005, ed. F-K Chang, pp. 125-132.
[3] Alan Mainwaring, Joseph Polastre, Robert
Szewczyk, David Culler and John Anderson,
Wireless Sensor Networks for Habitat
Monitoring. International Workshop on
Wireless Sensor Networks and Applications,
Proceedings of the 1st ACM international
workshop on Wireless sensor networks and
applications, 2002, pp: 88 – 97.
[4] Robert Szewczyk, Alan Mainwaring, Joseph
Polastre, John Anderson and David Culler, An
Analysis of a Large Scale Habitat Monitoring
Application, Conference On Embedded