You are on page 1of 4

Progress In Electromagnetics Research Symposium Proceedings, Stockholm, Sweden, Aug.

12-15, 2013 653

Benefits of Using Wireless Sensor Networks to Predict Plagues in


Vineyards
I. Cuinas, S. Cervera, and J. A. Gay-Fernandez
Department Teora do Sinal e Comunicacions, Universidade de Vigo, Spain

Abstract Wireless sensor networks (WSN) are experienced an important development, as


their applications grow and grow. Vineyards are one of the most popular scenarios to deploy
WSN in order to tracking various botanic parameters. Agronomists could use the sensing capa-
bilities to check the evolution of many parameters. Some of the measured parameters are closely
related to various plant plagues, as mildew in vineyards. The usefulness of such tracking in
plague prediction, and the possible benefits of controlling plagues by predicting their cycles with
automatic data capture are the objective of the present work, which highlights both the botanic
and the economic advantages of a fully performance WSN in a vineyard.

1. INTRODUCTION
Wireless sensor networks (WSNs) have been proposed as a key tool in improvement of agriculture
goods quality, as a mean to control different parameters related to the biological activity and
growing rhythm of the plants. Wine sector is especially interested in such technology: the final
product (the wine) has important added value, but its quality is strongly dependent to the vines
health and development during several months a year.
This paper is intended to describe the experience in installing and operating a WSN with
different meteorological sensors in a vineyard at a Spanish denomination of origin area, as well as
the processing of collecting data to analyze the most troublesome plague in the region, the mildew.
As described in Section 2, the basic system consists of a number of motes, each provided by a
number of sensors (temperature, humidity, leaf wetness, and so on), a coordinator, and a gateway
to link the data to the World (i.e., the Internet). This set of sensors is collecting data each a
defined time gap, constructing what could be seen as time varying signals that could be processed
and analyzed.
There are various mathematical models to predict plagues at vineyards, as indicated in Section 3.
The application of methods as EPI, POM, Goidanich, RAI, Milvit, or Plasmo allows us to provide
the biologists taking care of the vineyards with a mildew alert system, in order they to decide the
actions or products to apply the plants each time. This scientific alert procedure could reduce the
number of applied treatments, as it breaks the traditional period-based treatment and supports
the decision of cure when it is need. Then, a reduction in costs could be obtained: both economics
(products and work-time to apply them) and ecological (less chemical products incorporated to the
soil and plants), as explained in Section 4.
2. FUNDAMENTALS OF WIRELESS SENSOR NETWORKS
The concept of WSN is an evolution of wireless networks with a clear orientation: the information
transmitted from node to node in the network mainly come from sensors connected at each node.
Thus, each node has a double mission: collecting the data provided by the plugged in sensors, and
transmitting the own information or that received from other nodes to the adjacent ones. The
goal is to minimize the capabilities of each node, reducing it to the minimum needed to perform
its function: a small processing ability and short range transmitter/receiver. As a consequence,
instead of talking about nodes, in WSN they are better known as motes. This is a word with a clear
meaning: small, weightless, and inexpensive. The objective is to have a technology that allows the
deployment of networks constructed by hundreds of motes at a low cost.
Whereas the motes assure the interconnectivity, the sensors provide the data to be transmitted.
For agriculture applications there are a large collection of sensors, designing to measure a variety
of parameters: temperature and humidity both at soil and air, solar irradiation, rain, leaf wetness,
wind speed and direction, barometric pressure, rainfall and soil water content.
The motes carry the data towards the coordinator node. This coordinator works as a master
node, and it is connected to a gateway that allows the transmission of data far away the location
of the WSN. Depending on the installation, the gateway could be connected to a computer or
server, to a modem to access terrestrial cellular networks, or to a dedicated radio link. Anyway,
654 PIERS Proceedings, Stockholm, Sweden, Aug. 1215, 2013

independently of the physical communication method, finally data is stored in a data base, being
available for agronomists, biologists, managers, and other people caring the crop under tracking.
Besides, some WSN, as those based on IEEE 802.15.4 [1], have the ability of re-configuring its maps
of links when some of the paths do not work. This feature provides robustness to the complete
system.
3. PREDICTION OF PLAGUES ON VINEYARDS
WSN provides an exhaustive set of data, and up to date actualization, which is very good infor-
mation for those persons that need to monitor the status of the crop. However, it is possible to go
further: collected data could be analyzed by experts, but it also could be the input to mathematical
models that would help in plague detection tasks.
Concretely, there are some well-known plagues attacking vineyards, and farmers are very afraid
of them. Among those, mildew is probably the most studied, as it produces important economic
losses in vineyards and, consequently, in wineries. Mildew is the result of the presence of a fungus
in the green elements of the vines, and its life cycle is well defined [2]:
1. Contamination phase: in spring, zoospores are transported by rain and wind, reaching the
vine green parts and colonizing them. The needed conditions for contamination are in fact a
set: an air temperature around 10 C, a young plant of at least 10 cm, and a rain rate of at
least 10 mm/day.
2. Incubation phase: it takes from 4 to 21 days, depending on atmospheric conditions.
3. Sporulation phase: during this step, the fungus is propagated along the vine.
4. Propagation phase: dissemination to nearby plants with the help of rebounding rain drops
and wind. The optimal condition for ambience temperature is between 20 and 25 C, with the
help of rain during night.

Table 1: Distribution of sensors plugged in to the nodes.

nodes
Sensor
1 2 3 4 5 6
Air temperature and humidity X X X X X
Soil temperature and humidity X X X X X X
Solar radiation X X X
Soil water content X X X
Leaf wetness X X X X X X
Weather station: barometric pressure,
wind velocity and direction, rainfall, solar X
radiation, air temperature and humidity

Figure 1: Evolution of mean air temperature temporal series, 2011 season.


Progress In Electromagnetics Research Symposium Proceedings, Stockholm, Sweden, Aug. 12-15, 2013 655

Taking into account this process, most of the farmers apply preventive treatments to the com-
plete vineyard, commonly at fixed dates selected with no scientific criteria. Generally, this system
leads to a fixed number of treatments, in many cases more than needed, and not applied at the
most effective instants.
In parallel, some models have been developed, based on the biology of the fungus and the
evolution of meteorological parameters. Some of these models are IPS (Infection Potential State) [3],
POM (Prevision of Optimal Maturation) [4], Goidanich [5], RAI (Risk Assessment Index), Milvit [7],
and Plasmo [8]. They use different meteorological data, or various lengths in the time series, but
they also look for helping the farmers to detect in advance the presence of mildew.
4. EXPERIENCE IN ACTUAL TESTS
A pilot WSN was installed in a vineyard of 210,670 square meters within a demarcated region,
planted with vines of four different varietals. The yard is zoned and each zone is planted with only
one vine varietal.
The network consists of six nodes, a coordinator, a gateway, and a link to GPRS public network,
in order to use the Internet to transmit the data to the database at University facilities. This
network has been working for more than two years, providing data each 5 minutes [9]. Table 1
summarizes the sensors connected to each node.
Some of the results could be observed in Figures 1 to 3. These data have been analyzed using
both IPS and Goidanich methods. The IPS model indicates that the risk of mildew development
was very low in 2011 season, as the index resulted to be below 10 in spring. However, using data

Figure 2: Evolution of mean rain falling temporal series, 2011 season.

Figure 3: Evolution of humidity temporal series, 2011 season.


656 PIERS Proceedings, Stockholm, Sweden, Aug. 1215, 2013

from previous years, in spring 2009 the risk was very high. Goidanich model also confirms a null
risk of mildew infection during 2011 season. The tests developed allow us to conclude that both
methods are complementary.
5. CONCLUSIONS
The use of WSN in vineyards could allow the farmers to improve the control of the vines health,
and also to reduce costs, both economics and environmental.
A good knowledge of plague developments (the real-time condition of the vines, the exact days
of mildew contamination risks, etc.) would allow the farmers to apply only the strictly needed
treatments, instead of treating the plants in a calendar basis, which is pre-defined independently
of the meteorological conditions. The scientific procedure provides important advantages for the
agriculture exploitation:
1. Reduction of costs in chemical products: the treatments have to be applied only when they are
strictly necessary. As an example, in 2011 season, no treatments were needed in the vineyard
under test.
2. Reduction of chemicals in the soil: this redounds in less external elements in the environment,
improving the ecological footprint of the crop.
3. Reduction of costs in terms of person-hours: each treatment that is not needed represents a
number of working hours that could be used in other tasks.
4. Improvement in the quality of biologists work: they have up to the date information (i.e., each
5 minutes), what is impossible if they have to physically visit the vineyards.
ACKNOWLEDGMENT
This work was partially funded by Spanish Ministry of Economy and Competitiveness, State Sec-
retary for Research under project TEC2011-28789-C02-02, that was co-funded using European
Regional Development Funds (ERDF), as well as the Galician Regional Government under project
CN 2012/260 Consolidation of Research Units: AtlantTIC.
REFERENCES
1. The Institute of Electrical and Electronics Engineers, IEEE 802.15.4-2003 Standard, 2003.
2. Goncal Barrios, J. R., Modelizacion del Mildiu en la Vid, Phytoma, No. 164, December 2004.
3. Molot, B., La Modelisation du mildou de la vigne, Progres Agricole et Viticole, Vol. 103,
No. 1516, 1986.
4. Tran, C. and S. Strizyk, Simulation of the date of maturity of Plasmopara viticola oospores
to predict the severity of primary infections in grapavine, The American Pytopathological
Society, 1990.
5. Goidanich, G., Manuale di Patologia Vegetale, Edizione Agricole, Bologna, 1964.
6. Bendek, C. E., P. A. Campbell, R. Torres, A. Donoso, and B. A. Latorre, The risk assessment
index in grape powdery mildew control decisions and the effect of temperature and humidity
on conidial germination of Erysiphe necator, Facultad de Agronoma e Ingeniera Forestal,
Pontificia Universidad Catolica de Chile, 2007.
7. Spencer-Phillips, P. and M. J. Jeger, Advances in Downy Mildew Research, Vol. 2, Kluwer
Academic Publisher, 2004.
8. Orlandini, S., F. Rossi, and M. Magli, Plasmo project, web http://plasmo.bo.ibimet.cnr.it/.
9. Gay-Fernandez, J. A. and I. Cuinas, Deployment of a wireless sensor network in a vine-
yard, International Conference on Wireless Information Networks and Systems, WINSYS
2011, Seville, Spain, 2011.

You might also like