Computers and Electronics in Agriculture 98 (2013) 233–241

Contents lists available at ScienceDirect

Computers and Electronics in Agriculture

journal homepage: www.elsevier.com/locate/compag

A fuzzy logic based irrigation system enhanced with wireless data

logging applied to the state of Qatar
Farid Touati a,⇑, Mohammed Al-Hitmi a, Kamel Benhmed a,b, Rohan Tabish a
a
Department of Electrical Engineering, Qatar University, Qatar
b
Ecole Nationale d’Ingenieurs de Gabes (ENIG), Gabes, Tunisia

a r t i c l e i n f o a b s t r a c t

Article history: In arid regions, developing environment and crop-specific irrigation scheduling that reduces water lost
Received 12 November 2012 via evapotranspiration is a key to a sustainable and better managed irrigation. This paper presents a prac-
Received in revised form 26 June 2013 tical solution based on intelligent and effective system for a field of hyper aridity in Doha–Qatar. The sys-
Accepted 19 August 2013
tem consists of a feedback fuzzy logic controller that logs key field parameters through specific sensors
and a Zigbee–GPRS remote monitoring and database platform. The system is easy to deploy in existing
drip irrigation systems without any physical modification. For a given crop, the fuzzy logic controller
Keywords:
acquires data from these sensors and then applies well-devised fuzzy rules to produce appropriate time
Fuzzy logic
Intelligent drip irrigation
and duration for irrigation. All variables are fuzzified using trapezoidal and triangular membership func-
Evapotranspiration tions. In this fuzzification, Max–Min inference engine and Mamdani-type rule base is adopted in order to
Wireless monitoring make the best decision for each situation. Typical data in summer and winter showed that the controller
Arid region ensures maintaining the soil moisture above a pre-defined value with non-abrupt oscillations. The sys-
tem compensates the amount of water that is lost through evapotranspiration as predicted by Pen-
man–Monteith model and hence allows predicting future water consumption. A local station first
processes and saves real-time data received from the field controller via wireless Zigbee protocol to
finally transmit these data to a remote station via a GPRS link. This enhancement enables tracking system
performance in real time and creating a database for analysis and improvement. It follows that the
deployment of fuzzy control combined with remote data logging would foster better management of
irrigation and water resources in hyper-arid lands such as Qatar.
Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction improve the efficiency of water usage in agriculture will have a

The importance of water is indisputable around the world and
particularly in arid and hyper arid lands. With the rapid industrial
and social development, and the massive increase in Qatar’s popu-
lation, conventional water resources have become seriously de-
pleted. With an average of less than 250 m3 available per person
per year, Qatar falls far below the internationally recognized ‘water
poverty line’ of 1000 m3 per person per year (Al-Mohannadi et al.,
2003; Hasim, 2009; Prakash and Chenb, 2010; Kanzari et al., 2012).
Without proper management, water will become a severe con-
straining factor in the socio-economic development of the country.
Currently, agriculture accounts for about 60% of the total net de-
mand on fresh water in Qatar (Alexandridis et al., 2008; FAO Water
Reports, 2008; Feliu-Batlle et al., 2009; John, 2011). This demand is
expected to increase even further while ground water is continu-
ously in depletion. It is thus obvious that any measure taken to
⇑ Corresponding author. Tel.: +974 66239471; fax: +974 44034201.
E-mail address: touatif@qu.edu.qa (F. Touati).
significant impact towards achieving water supply sustainability,
and secure the availability of this very valuable resource to the
community.
In Qatar, wastage of water in irrigation is mainly caused by;
first: the use of traditional techniques which are based on timers
such as basins and furrows irrigation (Gillies and Smith, 2005),
and second: the water loss through ground evaporation and crop
transpiration (so-called evapotranspiration ET). In the first case, re-
search has shown that people over-irrigate crops due to the misun-
derstanding of seasonal water need or the impracticality of
updating the irrigation schedule to reflect actual water needs of
the landscape (Haley et al., 2007). In this scenario, people generally
adjust timers by observing the crop and irrigating when it looks
stressed (qualitative). On the other hand, the loss by evapotranspi-
ration is inevitable and accentuated by the hyper-arid environment
under untapped ambient temperature and solar radiation of
6 kW h/m2/day where daylight is about 4449 h/year (Touati et al.,
2013). Here, there is a need for automated irrigation systems that
are able to deliver the exact quantity of water required by the crop
for proper irrigation while reducing ET losses.

0168-1699/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.compag.2013.08.018
234 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241
There are several studies discussing the pros and cons of open-
loop and closed-loop control systems (McCreadya et al., 2009;
Rahangadale and Choudhary, 2011; Obota and Inyama, 2013).
According to (Wade and Waltz, 2004; Jaume et al., 2012), the most
deployed method of irrigation control is the closed-loop which
splits into two categories; feed-forward and feedback control. In
the feedback control, the idea is to maintain soil moisture (i.e.
plant’s water stress) within a specific range by measuring crop’s
needs from soil moisture levels using instruments such as tensiom-
eters or dielectric probes (Javadi et al., 2009). However, in the feed-
forward control (known as ET control), controllers use the crop’s
reference evapotranspiration (ETo) to schedule irrigation compen-
sating then for ET water loss through the water balance technique.
Climatic conditions have direct influence on ETo (Davis and Dukes,
2010), which can be calculated by using Penman Monteith model
as this has been officially adopted by the FAO (Allen et al., 1998;
Roy et al., 2009; Yang et al., 2010). In hyper-arid lands like Qatar
for example, ETo can reach 10 mm day1 in summer which corre-
sponds to 10 L per square meter per day (Hasim, 2009). For both
feedback and feed-forward approaches, in order to minimize ET
losses and hence system’s overrunning that would result in wast-
age of water and energy, the controllers should schedule irrigation
either in the morning, around sunrise, or even at night (Speetjens
et al., 2008; Car et al., 2012). This reveals the complexity of the irri-
gation decision and stresses the need for a robust and effective irri-
gation strategy especially in harsh environments where
evapotranspiration is acute. To address this complexity, much
attention has been paid to fuzzy theory in order to improve the
ability to make correct decisions.
Fuzzy logic interprets real uncertainties and becomes ideal for
nonlinear, time-varying and heuristic knowledge to control a sys-
tem. As contrasted to conventional feedback control systems, it is
gaining importance due to its flexibility in handling imprecise sub-
jective data and hence very effective for real-world decision-mak-
ing problems (Zhang et al., 1996; Mirabbasi et al., 2008). The
functionality of fuzzy logic has been extensively tested in a wide
range of applications. Jia et al. (2011) designed a field integrative
irrigation controller based on fuzzy logic and programmable logic
controller. A fuzzy logic-based multi-criterion decision making ap-
proach was applied for selecting the best-performing irrigation
subsystem in India (Raju and Kumar, 2005). A Fuzzy Logic Feed-
back Controller (FLC) prototype based on a Mamdani controller
and simulated on MATLAB software was shown to be more effec-
tive when compared to feedback controllers of simple on/off and
on/off with hysterics (Javadi et al., 2009). A fuzzy logic based ben-
efit-cost approach proved to be very suitable in decision making
between three alternative irrigation projects (Anagnostopoulosa
and Petalasb, 2011). In Mexico, the development of a fuzzy irriga-
tion control system using a field programmable gate array (FPGA)
to control greenhouse fertigation was found cost-effective and easy
to implement (Melendez et al., 2011). Rahangadale et al. (2011) re-
ported that fuzzy logic control improves the performance of auto-
matic irrigation systems by smoothing system ON/OFF and having
Fig. 1. Block diagram of smart irrigation system.
the potential for saving water when compared to conventional
controllers. Fuzzy logic was used to develop a model for crop water
stress index (Al-Faraj et al., 2011). In Qiu et al. (2007), a fuzzy irri-
gation decision-making system established by using virtual instru-
mentation platform of sensors, test instruments, data logger, and
LabView was presented. Generally, published studies use ON/OFF
controllers where the inherent complexity of irrigation process
made it difficult to achieve optimal results. Few studies used ET
controllers, however, inputs for such systems which are used to
calculate the theoretical irrigation requirement (scheduling) are
complex and subject to a lot of uncertainties rendering the sched-
uling efficiency and irrigation adequacy a difficult task (Davis and
Dukes, 2010). Systems using feedback (or soil moisture) controllers
have been used successfully. For example, an irrigation controller
which has been developed based on controlling soil moisture
(Muñoz-Carpena et al., 2004) reduced irrigation water by 70% on
drip irrigated tomato in South Florida. Research shows that feed-
back control based systems give promising results in terms of
water savings as high as 70% compared to drip approach with no
negative impact on crop yields (Nogueira et al., 2003; Dukes
et al., 2003; Dukes and Scholberg, 2005).
Most commercialized systems are ‘‘on/off with hysteresis’’ sys-
tems, where the controller continually compares one input with
two preset values to provide a decision either start or stop irrigat-
ing. This approach is simple but cannot handle efficiently complex
systems with multiple inputs like irrigation. As discussed above,
fuzzy logic is very flexible, robust by nature, has a clear reasoning
logic mimicking human brain (Domingo et al., 2011) and hence
well fit to this application.
In this study, we developed a site-specific standalone smart irri-
gation system (SIS) which uses FLC-based feedback ON/OFF control.
The SIS is designed to efficiently schedule irrigation for cucumber
under drip irrigation system in open fields in order to keep the soil
moisture above 17% and avoid times where evapotranspiration is
high. In Qatar, the soil moisture that is adequate for cucumber is
around 66%, however, it should not go below 17% to avoid water
stress (shortage) and bitter taste in the yield (Hasim, 2009). The
FLC is supposed to provide the time and duration of irrigation as
needed by the crop in a 100-m2 field in Doha (coordinates:
25.375°N, 51.490°E). Various sensors were deployed to log critical
factors, such as soil moisture, ambient temperature, solar radiation
and amount of water consumed. The readings from the various sen-
sors were then fed to the FLC to apply well-devised fuzzy rules to
control irrigation. In order to monitor and save measured data in real
time for further analysis, the system was also enhanced with Zig-
Bee–GPRS module for wide-range wireless monitoring and data log-
ging platform. To the authors’ knowledge, this is the first study
published on designing such irrigation system applied to Qatar.
2. System design
The system structure of SIS is shown in Fig. 1. It consists of two
main subsystems: (1) the sensing and control, and (2) long-range
 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241 235

communication protocol and monitoring. The control part is devel- 5 Volts

oped to be near to the field and a local (relay) station communi-
cates with the controller and remote station via ZigBee and GPRS
R1=10 k
communication protocols, respectively. A more detailed descrip-
Buffer
tion of the methodology in developing these subsystems is given Vout Microcontroller
in the following subsections.
RLDR
2.1. Sensing and control

The controller consists of specific field sensors, an FLC-embed-

ded data logging and wireless communication board, valve and
flow sensor, and power supply system. Fig. 3. LDR voltage divider circuit.

2.1.1. Field sensors and calibration

Three sensors were used to log the environment, namely, solar
radiation (LDR), soil moisture (VG400RevD), and ambient temper-
ature (LM35). These sensors are deemed most critical in the irriga-
tion process and their readings are used by the FLC to devise the
different fuzzy rules (Boman et al., 2006). Although there are other
parameters like relative humidity and wind speed, in arid lands
these parameters have little effect (Liao et al., 2008).
Since the FLC relies on readings from sensors, these have to be
checked for integrity. As sensors drift with use and time, they have
to be calibrated before use whenever deemed necessary. The cali-
bration results of the different sensors are discussed below.

2.1.1.1. LDR calibration. The resistance of the LDR decreases as the

light impinging on it increases and thus can be used to create a
simple and low-cost photometric sensor. The LDR was calibrated
using a light meter (RS 180-7133). The calibration results are de-
picted in Fig. 2 which spans a wide light intensity window from
dark to full sun light. The relationship between ambient light and
LDR resistance is non-linear, and shows a straight line if drawn
in a logarithm scale. A good fitting curve is given by following
equation:
log 10 ðRLDR Þ ¼ 0:5  log 10 ðLightÞ þ 4:23
When placed in a voltage divider circuit of Fig. 3, the LDR output
voltage (Vout) which is given by Eq. (2) reflects the field solar radi-
ation as follows.
V out ¼ 5  RLDR =ðRLDR þ R1 Þ
Then, combining (1) and (2) gives Eq. (3) below which expresses
light as a function of Vout. It follows that the controller (i.e. micro-
controller) can calculate the light intensity in the field by reading
Vout and plugging it into equation.
!!
104 þ V out
light ¼ 10 ^ 8:46  2  log
5  V out
Fig. 4. Experimental and estimated output voltage of moisture sensor vs. soil
moisture.
2.1.1.2. VG400RevD calibration. The VG400RevD sensor delivers a
maximum of 3 V for 100% moisture when powered with 5 V. Cali-
bration measurements in Fig. 4 show a non-linear behavior be-
tween output voltage (Vout) and soil moisture (vwc). A good
estimate of a fitting curve between Vout and moisture is given by:
ð1Þ V out ¼ 2:98  ð1  eðv wc=7Þ Þ ð4Þ
The sum squared error between the experimental data and Eq.
(4) was estimated to be about 0.182, which we consider acceptable
in this application.
ð2Þ In addition, Fig. 4 shows that the dynamic range of the soil
moisture sensor used is around 35%. This means that an acquisition
system cannot discern between moistures above 35% since the soil
moisture sensor will provide same output voltages for moistures
above 35% as shown in Fig. 4. In this application to schedule irriga-
tion the FLC uses threshold moisture of 17% (see Fig. 5) which is
within the sensor range.
ð3Þ
2.1.1.3. LM35 calibration. The LM35 does not require any external
calibration or trimming to provide typical accuracies of ±3/4 °C
over a full 55 to +150 °C temperature range. The LM35 has a lin-
ear +10.0 mV/°C scale factor, to detect a maximum of 100 °C. The
calibration curve is given by:

V out ¼ 0:05  temperature ð5Þ

To reduce conversion errors (conversion, linearity, drift), a con-
ditioning block was used to match the sensor output to the analog
full-scale range of the microcontroller ADC. Eqs. (3)-(5) were used
at the local station to convert received data back to light, soil mois-
ture and ambient temperature, respectively.

2.1.2. FLC-based data logging and wireless communication board

The board is constructed around the ATmega32 microcontroller
which makes the system compact, deployable, adaptive, and scal-
Fig. 2. Experimental and estimated LDR resistance vs. light intensity. able. The main tasks of this board are data acquiring, processing,
236 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241

Table 1
Begin RULES for duration of irrigation (in min) at wet soil moisture.

Air temperature (°C) Solar radiation (Lux)

Light Medium Dark
Read: Temperature,
Cold Zero Zero Zero
Radiation and Moisture Medium Zero Zero Zero
Hot Zero Zero Zero

Moisture > 17% Yes Table 2

RULES for duration of irrigation (in min) at medium soil moisture.

Air temperature (°C) Solar radiation (Lux)

No
Light Medium Dark

Calculate Duration from Cold Short Short Short

Medium Very short Short Short
Fuzzy Logic
Read Water Flow Hot Zero Very short Long

Start irrigation Table 3

RULES for duration of irrigation (in min) at dry soil moisture.

Air temperature (°C) Solar radiation (Lux)

Read Water Flow
Light Medium Dark
Cold Very long Very long Very long
Medium Short Long Long
Wait for Duration Hot Zero Very short Very long

Stop irrigation was added to the system. This flow sensor output is binary and
hence a frequency-to-voltage (F/V) converter was added. The out-
put of F/V converter was fed to the microcontroller analog input
Send data via Xbee port. By recording the duration of irrigation, the local and remote
stations calculate the amount of water being consumed. This
amount is displayed and saved along with ambient temperature,
Wait 30 minutes soil moisture, and solar radiation readings. The flow sensor also
alerts in case of shortage of water.

Fig. 5. Flow chart software. 2.1.4. Solar power supply

The control hardware was developed over a small printed cir-
cuit board and consumes a low current (max of 0.5 A) under a
logging, and transmitting. An application-specific FLC was devised
12-V supply. The whole system was run from a single solar
and loaded to the microcontroller. MAT Lab simulations and sys-
100 W-PV (from PTL Solar Co.) charging a battery via a charge opti-
tem reduction were performed on the FLC beforehand. Fuzzy sets
mizer (PTL/CC-T/30). The solar power supply was designed so that
and logic can handle real-life uncertainties hence ideal for as such
it can power the control subsystem continuously. The panel was
nonlinear, time-varying and hysteretic control system (Bellazzi
tilted at an angle to match the latitude of the location. In order
et al., 1998; Hahn, 2011). Detailed description of the FLC is given
to avoid panel heating and damage due to high irradiation levels
in the subsection below. Also, the board is enhanced with Zigbee
(6 kW h/m2/day, untapped world’s greatest solar power potential),
modem for short-range wireless connectivity. In this application,
a slight shift from that angle is recommended. Since Qatar is lo-
the board is configured to transmit all sensors’ reading to the local
cated in the tropical cancer at latitude of 25° from the equator,
station every 30 min. The local station saves and in turn transmits
the system was installed with a tilt angle of 23°. This set up al-
processed data to the remote station. This feature enables remote
lowed replenishing the battery enough to account for energy need
monitoring and creating database of irrigation parameters, envi-
of the system during the day. A critical unit in the SIS is the FLC
ronmental conditions, and system performance for analysis and
which is described thoroughly in the subsection below.
better management.

2.2. Fuzzy logic control

2.1.3. Water valve and flow sensor
For water supply, conventional on/off electro-valve was se-
lected because it consumes less current than proportional solenoid
valves. The on/off electro-valve was connected to a drip irrigation
pipe line and controlled by the FLC. The current drawn by the elec-
tro-valve is about 200 mA under 12-V supply. Therefore, a buffer
circuit that ensures driving the electro-valve from one of the
microcontroller output port was used.
In order to record the amount of water being dispensed, a flow
sensor (model FS-1F-SZ) which measures the water flow in L/min
Fig. 5 shows the flow chart of system operation. At the begin-
ning, the FLC acquires readings from ambient temperature, solar
radiation and soil moisture sensors to compute the fuzzy rules.
The soil moisture sensor indicates if irrigation is needed. In this
case study, if moisture falls below 17% then irrigation is triggered.
However, the duration of irrigation is made by the FLC after carry-
ing out a fuzzy reasoning and decision using the information from
these sensors (see paragraph below). In any situation, the outcome
will be the time and duration of irrigation according to the
 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241 237

Input Data

Fuzzifier

Inference Rule Base

Defuzzifier

Output Data

Fig. 6. Fuzzy logic controller structure.

Temperature

Fig. 8. Photos of the system: field controller and sensors with solar PV system.

rule-base (see Tables 1–3). It is worth mentioning that the process

of aspiration of water through the soil was estimated to be about
Solar Radiation (Lux) 30 min, thus the system was programmed to wait the same period
before starting next cycle. This accounts for the 30-min frequency
of data transmission. The program allocates 6.5 kilobytes in flash
memory, equivalent to 20.3% of the total memory (32 kilobytes)
only.
Fig. 6 explains the structure of the FLC implemented here. All
variables are fuzzified using trapezoidal and triangular membership
functions. The membership functions are distributed according to
the possible values of each variable after fuzzification. The fuzzifi-
cation process, Max–Min inference engine and Mamdani-type rule
base produce the required decision for each situation. After appli-
cation of a centroid defuzzification, the controller produces the de-
Soil moisture (%) sired output.
Three inputs (i/p) and one output (o/p) are used in this work:

Soil moisture (i/p#1): gives the water stress in soil. It is the most
important information to know whether or not irrigation is
needed and estimate the amount of water that should be dis-
pensed to the crop.
Outside temperature (i/p#2): measures ambient temperature.
This input enables the FLC reducing water evaporation by
avoiding irrigation at high ambient temperatures.
Radiation (i/p#3): measures solar radiation in the field. Likewise
the ambient temperature input, since solar radiation acceler-
irrigation duration (minutes) ates evapotranspiration, it helps the FLC reducing water loss
through evapotranspiration.
Duration (o/p): the FLC provides the time and duration of irriga-
Fig. 7. Inputs and output membership functions. tion through the output port.
238 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241

Fig. 9.1. Real-time remote temperature and radiation data (May 2012).

Fig. 9.2. Real-time remote duration of irrigation and temperature data (May 2012).

Fig. 9.3. Real-time remote soil moisture and duration of irrigation data (May 2012).

The devised inputs and output membership functions of FLC are
shown in Fig. 7.
For the purpose of avoiding memory leaks and improving the
program execution speed, three significant membership functions
were used for each input variable. The output variable needs to
be extended with five membership functions to cover the variation
of water need.
Tables 1–3 show the designed fuzzy logic based rules that were
implemented in the system. As an example of these rules:
 ‘‘If the soil is wet, then regardless of the temperature and radi-
ation values, the irrigation duration is set to Zero’’.
 ‘‘If the soil moisture is medium, temperature also medium and
radiation is light, then the irrigation duration is set to Very
Short’’.
 ‘‘If the soil moisture is dry, temperature is hot while the radia-
tion is light, then the irrigation duration is set to Zero’’.
2.3. Long-range communication protocol and monitoring
End-to-end transmission of SIS data is assured by a Zigbee–
GPRS platform. The communication between the field controller
and local station is done over a short-range (100 m line-of-sight)
IEEE802.15.4 compliant Zigbee protocol (Xbee Series 1 chip) by vir-
 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241 239

Fig. 9.4. Real-time remote data of water consumption (May 2012).

Fig. 10.1. Real-time remote temperature and radiation data (January 2013).

Fig. 10.2. Real-time remote soil moisture and duration of irrigation data (January 2013).

tue of its low-power feature. This would support long power sys-
tem autonomy since the control subsystem is solar-powered. In
addition, IEEE 802.15.4 ‘‘Zigbee’’ compliant chips present unique
features for conventional and precision agriculture (Garcia-San-
chez et al., 2011). The local station receives, processes, and saves
data from the field and then routs this data to a remote station
(i.e. a PC) via wide coverage GPRS/GSM modules on both sides.
The GPRS/GSM module is interfaced to the remote station via a
microcontroller over UART. All data are saved and displayed in
real-time at the remote station using the graphical software Lab-
View. The data at the local and remote stations when compared
for conformity, revealed no differences.
In fact, in an actual scenario, there is no local monitoring for
large systems because it is not practical to deploy monitors (e.g.
PCs) within every 100 m range. The purpose of the local station
is to provide connectivity between the short-range and long-
range networks. This local station is generally AC-powered from
the sector and meant to remain turn-on all the time. Also, the
local station permits system setting and testing during deploy-
ment phases.
240 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241
Fig. 10.3. Real-time remote data of water consumption (January 2013).
Fig. 8 depicts shots of the proposed developed system including
the field controller, sensors and solar PV supply. The system can be
easily implemented into existing drip-type irrigation systems
without any logistic or physical modifications.
3. Results and discussion
To examine the system’s behavior, typical real-time data col-
lected remotely at two different seasons; summer (May 2012)
and winter (January 2013), are illustrated in Figs. 9.1–9.4 and
10.1–10.3, respectively. In May and June, the average temperature
is around 40 °C and 20 °C and the duration of luminance is about
14 and 10 h of daylight, respectively. The curves show actual mea-
surements of a typical field plant in Doha–Qatar. The reliability of
these data was verified by comparing them to on-site and local sta-
tion data.
As expected, Figs. 9.1–9.3 and 10.1–10.3 showed that the FLC
avoids operating the system during high temperatures, which cor-
responds to high solar radiation (Figs. 9.1 and 10.1). This contrib-
utes in reducing evapotranspiration losses throughout the day.
Also, the FLC selects the appropriate time and duration by referring
to the base of fuzzy rules (Figs. 9.2 and 10.2). The actual soil mois-
ture tracks the desired value (17%) with non-abrupt oscillations as
compared to traditional feedback systems, avoiding frequent sys-
tem runoff of such ON/OFF systems (Figs. 9.3 and 10.2).
Fig. 9.4 shows that in 6 days the total amount of water con-
sumed was 6252 L, that is, an average of 1042 L per day over an
area of 100 m2. It follows that our control system irrigated the field
with an average of 10.42 L/day per square meter, that is,
10.42 mm day1. Interestingly, this is similar to the evapotranspi-
ration value for arid lands in summer; 10.00 mm day1 as reported
in (Hasim, 2009). The same calculation was made using the data in
Fig. 10.3 (January) which resulted in an average irrigation of
6.70 mm day1 whereas the evapotranspiration calculated by
FAO Penman–Monteith model in January is around 3.72 mm day1
(Hasim, 2009). This indicates that in winter the system compen-
sated ET losses with an over irrigation. Over-irrigation has also
been reported for systems using ET controllers as reported by Davis
and Dukes (2010). Overall, the above results show that the fuzzy
control-based irrigation system developed here compensates effi-
ciently ET losses in hyper-arid regions like Qatar for a well-watered
and actively-growing crop. The system enables predicting future
water needs and hence fosters better irrigation management. This
would require the development of appropriate artificial intelli-
gence algorithms (e.g. ANN) that are trained over wide time win-
dows. Our team is currently embarking on developing such
algorithms along with scaling up the system to larger fields.
A preliminary study showed that compared to the current irri-
gation techniques in Qatar (surface flooding and furrow), the cur-
rent SIS system saves up to 80% of water consumption, equal to
around 1521320.00 L per year over an area of 100 m2. For the user,
this translates locally to around $1825.6 of net profit per year for
highly subsidized price of about $1.20/m3, about 35% of the real
cost (Darwish and Mohtar, 2013). This looks very beneficial vis a
vis a total system cost of around $400.00. The developed system
is low-cost, efficient, and easy to implement. This is in addition
to the benefit of preserving the groundwater aquifer which is cur-
rently considered a ‘‘first-order resource’’, that means it is a re-
source which is becoming scarcer in the country.
4. Conclusions
In this paper, feedback fuzzy control and wireless monitoring
were presented as a strategy to develop an irrigation approach that
fosters water conservation and better irrigation management in
hyper-arid lands like Qatar. The developed fuzzy controller, based
on Mamdani fuzzification using trapezoidal and triangular member-
ship functions, efficiently set the time and duration of irrigation
for a given crop. The use of fuzzy control helped maintaining the
soil moisture above a pre-set value with smooth oscillations pre-
venting hence frequent system’s run-off and preserving water
and energy. The system compensated the amount of water that
was lost by evapotranspiration as predicted by Penman–Monteith
model in summer and winter with a noticeable over-irrigation in
winter. Therefore, the system enables predicting future irrigation
needs. In order to monitor system performance in real time and
create a database, a wide-range ZigBee–GPRS based wireless sys-
tem was also developed. Key environmental and climatic factors
along with the time and duration of irrigation as well as the
amount of water being sourced were recorded. Currently, this is
helping a great deal in data analysis and system improvement.
The system is easy to implement. A preliminary cost-benefit anal-
ysis showed that the system is economically justifiable.
Acknowledgment
This research has been supported by Qatar University project
Grant (QUUG-ENG-DEC-10/11-13).
References
Al-Mohannadi, H., Hunt, C., Wood, A., 2003. Controlling residential water demand in
Qatar: an assessment. The Royal Swedish Academy of Sciences 32, 362–366.
Alexandridis, K., Zalidis, C., Silleos, G., 2008. Mapping irrigated area in
Mediterranean basins using low cost satellite Earth Observation. Computers
and Electronics in Agriculture 64, 93–103.
 F. Touati et al. / Computers and Electronics in Agriculture 98 (2013) 233–241
Allen, R.G., Peeria, L.S., Racs, D., Smith, M., 1998. Crop Evapotraspiration (Guidelines
for Computing Crop Water Requirements). FAO Irrigation Drainage. Paper 56,
Rome, Italy.
Anagnostopoulosa, K.P., Petalasb, C., 2011. A fuzzy multicriteria benefit–cost
approach for irrigation projects evaluation. Agricultural Water Management
98, 1409–1416.
Bellazzi, R., Ironi, L., Guglielmann, R., Stefanelli, M., 1998. Qualitative models and
fuzzy systems: an integrated approach for learning from data. Artificial
Intelligence in Medicine 14, 5–28.
Boman, B., Smith, S., Tullos, B., 2006. Control and Automation in Citrus
Microirrigation Systems, Agricultural and Biological Engineering Department.
University of Florida, Institute of Food and Agricultural Sciences (UF/IFAS)
CH194, Circular 1413.
Car, N., Christen, E., Hornbuckle, J., Moore, G., 2012. Using a mobile phone Short
Messaging Service (SMS) for irrigation scheduling in Australia–Farmers’
participation and utility evaluation. Computers and Electronics in Agriculture
84, 132–143.
Darwish, M.A., Mohtar, R., 2013. Qatar water challenges, desalination and water
treatment. Desalination and Water Treatment 51 (1-3), 75–86.
Davis, S.L., Dukes, M.D., 2010. Irrigation scheduling performance by
evapotranspiration-based controllers. Agricultural Water Management 98,
19–28.
Domingo, G.M., Alvaro, L.L., Gilberto, H.R., Carlos, F., Enrique, R.-G., Carlos, O.O.,
Daniel, A.L., Teobaldis, M.F., Sven, V., 2011. Fuzzy irrigation greenhouse control
system based on a field programmable gate array. African Journal of
Agricultural Research 6 (11), 2544–2557.
Dukes, M.D., Simonne, E.H., Davis, W.E., Studstill, D.W., Hochmuth, R., 2003. Effect of
sensor-based high frequency irrigation on bell pepper yield and water use. In:
Proceedings of 2nd International Conference on Irrigation and Drainage, May
12–15, Phoenix, AZ, pp. 665–674.
Dukes, M.D., Scholberg, J.M., 2005. Soil moisture controlled subsurface drip
irrigation on sandy soils. Applied Engineering in Agriculture 21 (1), 89–101.
FAO Water Reports, 2008. Irrigation in the Middle East Region in Figures.
Feliu-Batlle, V., Rivas-Perez, R., Castillo-Garcia, F.J., 2009. Fractional order controller
robust to time delay variations for water distribution in an irrigation main canal
pool. Computers and Electronics in Agriculture 69, 185–197.
Garcia-Sanchez, A., Garcia-Sanchez, F., Garcia-Haro, J., 2011. Wireless sensor
network deployment for integrating video-surveillance and data-monitoring
in precision agriculture over distributed crops. Computers and Electronics in
Agriculture 75, 288–303.
Gillies, M.H., Smith, R.J., 2005. Infiltration parameters from surface irrigation
advance and run-off data. Irrigation Science 24, 25–35.
Hahn, F., 2011. Fuzzy controller decreases tomato cracking in greenhouses.
Computers and Electronics in Agriculture 77, 21–27.
Haley, M.B., Dukes, M.D., Miller, G.L., 2007. Residential irrigation water use in
Central Florida. Journal of Irrigation and Drainage Engineering 133 (5), 427–
434.
Hasim, M.A., 2009. Water, Agriculture and Environment in Arid Lands–Water and
Agriculture Vision for Qatar by 2020. Friends of Environment Centre (FEC),
Doha, Qatar.
Jaume, C., Mercè, M., Jordi, M., Joan, G., 2012. A general algorithm for automated
scheduling of drip irrigation in tree crops. Computers and Electronics in
Agriculture 83, 11–20.
Javadi, K.P., Tabatabaee, F.A., Omid, M., Alimardani, R., Naderloo, L., 2009. Intelligent
control based fuzzy logic for automation of greenhouse irrigation system and
evaluation in relation to conventional systems. World Applied Sciences Journal
6, 16–23.
Jia, X., Yao, L., Zhang, Y., 2011. Design of field integrative irrigation control system
based on fuzzy control and PlC. Communications in Computer and Information
Science 237, 295–301.
241
John, F.R., 2011. The impact of mechanization on agriculture. National Academy of
Engineering 41, 22–29.
Kanzari, S., Hachicha, M., Bouhlila, R., Battle-Sales, J., 2012. Characterization and
modeling of water movement and salts transfer in a semi-arid region of Tunisia
(Bou Hajla, Kairouan) – salinization risk of soils and aquifers. Computers and
Electronics in Agriculture 86, 34–42.
Liao, L., Zhang, L., Bengtsson, L., 2008. Soil moisture variation and water
consumption of spring wheat and their effects on crop yield under drip
irrigation. Irrigation Drainage Systems 22, 253–270.
McCreadya, M.S., Dukesa, M.D., Millerb, G.L., 2009. Water conservation potential of
smart irrigation controllers on St. Augustinegrass. Journal of Agricultural Water
Management 96, 1623–1632.
Melendez, D., Lambraño, A.L., Ruiz, G.H., Fuentes, C., Garcia, E.R., Olvera, C.O.,
Lumbrerasc, D.A., Fernández, T.M., Verlinden, S., 2011. Fuzzy irrigation
greenhouse control system based on a field programmable gate array. African
Journal of Agricultural Research 6 (11), 2544–2557.
Mirabbasi, R., Mazloumzadeh, S.M., Rahnama, M.B., 2008. Evaluation of irrigation
water quality using fuzzy logic. Research Journal of Environmental Sciences 2
(5), 340–352.
Muñoz-Carpena, R., Dukes, M.D., Li, Y.C., Klassen, W., 2004. Design and field
evaluation of a new controller for soil moisture based irrigation. ASAE Paper No.
04–2244. Presentation at ASAE/CSAE Annual International Meeting, 1–4 August,
Ottawa, Canada.
Nogueira, L.C., Dukes, M.D., Haman, D.Z., Scholberg, J.M., Cornejo, C., 2003. Data
acquisition and irrigation controller based on CR10X datalogger and TDR sensor.
Proceedings Soil and Crop Science Society of Florida 62, 38–46.
Obota, M.E., Inyama, H.C., 2013. Soil moisture based irrigation control system for
rice cropping using wireless sensor network. The International Journal Of
Engineering And Science 2, 37–43.
Prakash, N., Chenb, D., 2010. Modification of a spatially referenced crop model to
simulate the effect of spatial pattern of subsoil salinity. Computers and
Electronics in Agriculture 74, 313–320.
Qiu, Z., Tong, X., Shen, J., Bao, Y., 2007. Irrigation decision-making system based on
the fuzzy-control theory and virtual instrument. Nongye Gongcheng Xuebao/
Transactions of the Chinese Society of Agricultural Engineering 23 (8), 165–169.
Rahangadale, V.S., Choudhary, D.S., 2011. On fuzzy logic based model for irrigation
controller using Penman-Monteith Equation. International Journal of Computer
Applications, 22–25.
Roy, D., Pandab, S.N., Panigrahic, B., 2009. Water balance simulation model for
optimal sizing of on-farm reservoir in rainfed farming system. Computers and
Electronics in Agriculture 65, 114–124.
Speetjens, S.L., Janssen, H.J.J., Straten, G., Gieling, Th.H., Stigter, J.D., 2008. Methodic
design of a measurement and control system for climate control in horticulture.
Computers and Electronics in Agriculture 64, 162–172.
Raju, K.S., Kumar, D.N., 2005. Fuzzy multicriterion decision making in irrigation
planning. Irrigation and Drainage 54, 455–465.
Touati, F., Al-Hitmi, M., Bouchech, H., 2013. Study of the effects of dust, relative
humidity and temperature on solar PV performance in Doha: comparison
between mono-crystalline and amorphous PVs. The International Journal of
Green Energy 10, 680–689.
Yang, Y., Wilson, L.T., Wang, J., 2010. Development of an automated climatic data
scraping, filtering and display system. Computers and Electronics in Agriculture
71, 77–87.
Zhang, Q.X., Wu, C.H., Tilt, K., 1996. Application of Fuzzy Logic in an Irrigation
Control System. In: Proceedings of the IEEE International Conference on
Industrial Technology (ICIT’96), 2–6 Dec., Shanghai, China, pp. 593–597.