Read Microsoft Word - 060_111pPA.doc text version

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

Wireless Sensor Network for Variable Rate Irrigation in Citrus

André Torre-Neto, Rafael A. Ferrarezi, Daniel E. Razera, Eduardo Speranza, Wellington C. Lopes, Thiago P.F.S. Lima, Ladislau M. Rabello, Carlos M.P. Vaz Embrapa Instrumentação Agropecuária, P.O. Box 741,13560-970 São Carlos, SP, Brazil [email protected] Keywords: precision agriculture, fixed instrumentation, soil moisture mapping, sitespecific irrigation system Abstract São Paulo State, in Brazil, has the most important citrus production area in the world. The largest part of it is not yet irrigated. Due to recent spreading of some new diseases, like sudden death, this scenario will change and most of the groves will have to be irrigated. Therefore, water source conservation methods must be developed. In the present paper we present the development of a wireless fixed instrumentation (sensor and actuator network) and the related software tools to irrigate perennial crops site-specifically. Citrus crop production is our first goal. This is an on going research and only technical aspects are presented. INTRODUCTION Brazil fruit production was around 38 million of tons in 2002, which meant the third worldwide position, behind China with 133 million and India with 59 million of tons (FAO Statistical, 2002). Citrus production was the most important contribution for that number (49%) and orange juice alone represented 1.87% of the total Brazilian exportation for that year. In spite of that, most of the whole orange grove area (820 thousand hectares) is not yet irrigated. This scenario is changing. Due to the need of disease control, mainly the sudden death, the installed irrigation systems expanded from 1,5% in 1999 to 10,2% in 2004 for the total citrus area. Therefore, conservation of valuable water sources demands that methods must be developed to maximize the use efficiency of irrigation water. Spatially-variable crop production, often known as precision agriculture, has been widely studied and developed to improve agricultural use efficiencies and to reduce environmental impacts. But most of the early research and commercial developments in spatially-variable crop production have been concentrated on variable-rate fertilizer and pesticide application (Schueller, 1997). Experiences with yield maps have convinced many researchers and agriculturalists of the importance of water availability in determining spatial yield patterns. Many think that water availability is the major determinant of yield variation. There has therefore been some significant research into variable irrigation, most of it (such as Fraisse, et al., 1995 and Sadler, et al., 1996) with center pivot or linear move systems. Some efforts to work with variable rate microsprinkler irrigation, the predominant and most efficient type of irrigation used for citrus and similar tree crops, have been carried out by Torre-Neto, et al.in 2001 in Florida, USA. The objective of the present work is to improve those preliminary efforts in Florida using the wireless technology. It has been developed and implemented a demonstration unit of an automated, spatially-variable irrigation system for citrus, based on a proprietary wireless sensor and actuator network. The system was conceived to be low-cost, reliable, and compatible with contemporary local citrus production practices

563

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

and other perennial crops, like coffee. The focus of the present paper is to describe the development of that wireless instrumentation and the related software tools. MATERIAL AND METHODS Commercially available wireless sensor network systems are still not costeffective, mainly for developing countries, and have no appropriate features for the agricultural fields and practices. Therefore, it was developed a proprietary wireless sensor network system adequate for remote monitoring and control in the agronomical area. The system is constituted by Sensor and Actuators Nodes, one or more Field Stations (FS), one Base Station and an installation Kit. Figure 1 exhibits the system architecture. Each Field Station covers 100 hectares. For the local average property size, it means more than one Field Station per farm (around 10 in average). The Base Station main element is a personal computer located in the farm office. So far, four hardware components and nine software modules were developed and will be described. The sensor and actuators Nodes developed up till now are intended for variable rate micro sprinkler irrigation based on soil parameters. There is a soil sensor Node that measures soil moisture and temperature at a specific depth and an actuator Node that controls latching solenoids. They were devised by these premises:

The sensor Nodes must be distributed in a fix grid pattern, spaced 50 meters from each other, and positioned under the citrus tree; The sensor Nodes must be battery operated with no-recharging batteries that must last at least 6 months; The actuator Nodes must be battery operated with recharging batteries associated to photovoltaic cells (unlike the sensor nodes they can be positioned to receive direct sun light); All Nodes must be resistant to agrochemicals; The Nodes installation, operation and maintenance must be as simple as possible not to require specialized staff; The Nodes unit price must be around US$ 50.

Soil Sensor Node A capacitive soil moisture sensor was designed and built to fit the inside space of a PVC plastic tube measuring 25mm in diameter and 20cm long. This piece was attached to one edge of another PVC tube measuring 18mm in diameter and 50cm long and having at the other edge the circuitry, sealed inside an aluminum case with the antenna installed at its top. The antenna is a 6 dBi gain double dipole. The circuit was developed around a nanowatt technology microcontroler (PIC 16F88 from Microchip Inc.), a low power digital radio transceiver (DR-3000 from Radio Frequency Monolithic), a LM35 integrated circuit as the temperature sensor and a few other electronic components (Amp. Ops., transistors, etc). There are a main board with the microcontroller and the communication circuits and a daughter board with the sensor interface and signal conditioning circuits. The soil sensor Nodes are powered by a 3.6 volts, 1000mAh lithium battery. The Node duty cycle is controlled by a Real Time Clock. They were programmed to remain most of the time in sleep mode and wake up every 15 minutes for data acquisition. The communication with the Field Station occurs every hour. The data communication protocol operates in a pool-response mode and was adapted from the SNAP protocol (Scalable Node Address Protocol, 2005). The microcontroller is able to switch power individually for the radio and for the analog circuits in order to save power whenever it is

564

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

possible. The microcontroller software was developed in assembly language and the resulted program runs stand alone (with no operational system). Valve Actuator Node The actuator Node has similar hardware, software and construction to the sensor Node. The main change is the daughter board that has the circuit interface to Latching Solenoids and to Flow Meters. An H-bridge was used to perform the switching of a twowire type Latching Solenoid and a simple debounce circuit attached to the microcontroller counter input performs the Flow Meter interface. Different from the sensor Node, this Node is powered by a pack of four 1.2 volt, 800mAh NiCd batteries associated to a 1 Watt photovoltaic panel. The actuator Nodes duty cycle, acquisition interval and data communication are the same as the sensor Nodes. Field Stations The Field Stations are the data sink of this wireless sensor network. Their primary function is to receive, process and temporarily store data in the field until a data request from the Base Station. They also perform the valve actuation and operation supervision scheduled by the operator at the Base Station. Another important function executed by the Field Stations is the sensor network routing. Each Field Station radio signal coverage range is around 100 hectares what means 400 sensor Nodes spaced 50m from each other. The data communication protocol is prepared to handle 65 thousand nodes. The Field Stations are composed by a CPU, a mini-weather station, a WLAN interface and the system data communication driver, all powered by a set of two 55Ah lead acid batteries recharged by a 70-Watt photovoltaic solar panel. The CPU is a PC104 single board computer based on a 300 MHz NS Geode GX1 processor and NS CS5530A chipset (PCM-3550 from Advantech), 256 Mbytes RAM and a 512 Mbytes Compact Flash as the system disk. It was installed the Linux Slackware 9.1 and kernel 2.4.22 as the operational system and all software was developed in ANSI C Language. The weather station is the WMR-918 from Oregon Scientific connected to one of the two PC104 serial RS-232 interfaces. The wireless LAN interface is a Tecom model WL24U plugged to the PC104 USB port and having an external 18 dBi gain directional antenna to establish the link to the base station. The usual TP/IP services like FTP, SMTP and SSH are used to send and receive data, file and commands to and from the base station. Configuration and data file are generated in the XML standard. The data communication driver is the only non-commercial part of the FS. It was devised with another microcontroller interfaced to one DR-3000 radio module to transmit and another one to receive data from the Nodes, besides one serial RS-232 interface to communicate with the CPU PC104. It runs the low level portion of the data communication protocol. The transmission power level is increased 20 times related to the Nodes in order to the FS reach 500m range. Base Station The Base Station is used to centralize all system operation. It is composed by a personal computer and a WLAN access point connected to an external omnidirectional antenna that performs data communication with all Field Stations. A space-temporal data base was created and management software was developed with geographical information system capabilities to operate the system and help to make decisions about the variable-rate irrigation. The data base has all system entities information, like the Nodes geographical coordinates, configuration and operation modes,

565

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

links, as well as the farm and staff information. The management software was developed in C++ and the geographical information capabilities are from a free library called Terralib developed by INPE, the Brazilian National Institute for Space Research. Installation Kit The installation kit was created to make the system installation and maintenance as easy as possible. It is basically a Palm-PC connected to a GPS receiver and two software modules, one at the Base Station and the other at the Palm-PC. The system installation begins with a virtual installation of the Nodes using the Base Station software module that has geographical information tools to establish Nodes and FS position and configuration. After this step, all that information about the Nodes and FS installation is synchronized with the palm-PC. At the field a Node to be installed is picked from the Palm-PC screen and navigation software leads the operator to the exact Node point. The Node is switched to installation mode and receives from the Palm-PC all its data operation, like ID, routing information, acquisition time interval, communication time interval, among other parameters. In this way, the Nodes go to the field all cleared, with no need of previous configuration. The software module for the Palm-PC was developed in Java by the SuperWaba, a special software development tool for PDAs. RESULTS AND DISCUSSION A six hectares pilot unit was installed and put in operation to evaluate the system operation in field condition. It was selected a ten-year-old orange grove with trees about 3m tall. There are nine sensor Nodes and one Field Station in operation since December 15th, 2004. The system was programmed to acquire data every 15 minutes and transmit data to the Field Station every hour. Figure 2 shows one soil moisture map obtained by interpolation of the nine sensors data at an arbitrary moment. Operational problems are being solved before installing all the 25 predicted sensors for this pilot unit. They are mainly related to the Nodes microcontroller program and routing at the Field Station. Besides these minor problems the system is quite reliable to map soil moisture in almost real-time. Next steps include expanding this pilot unit to 25 hectares, what means 100 sensor Nodes installed in a this size plot. Soil moisture map sequences for the whole coming draught season will be obtained for that plot. The expected result from those maps analyses is to have the grove divided in five or more management zones with different needs of irrigation water. Those zones will reflect the spatial variability for the combination of different factors like, soil texture, terrain topology and individual tree needs. The irrigation system will be modified to site-specifically irrigate the crop accordingly to those zones. For each zone it will be addressed an automated valve and established a control loop with its own set point. This set point will be compared to the average soil moisture for that zone obtained by the sensor network subset concerned to that zone. The water consumption will be measured by an automated flow meter for the entire plot and will be compared to the consumption of a conventionally automated plot. ACKNOWLEGMENTS Funding: FAPESP - Fundação de Amparo à Pesquisa no Estado de São Paulo Process 03/07998-5; Embrapa - Empresa Brasileira de Pesquisa Agropecuária Colaborative Institutes: FUNDECITRUS - Fundo de Defesa da Citricultura; INPE- Brazilian National Institute for Space Research

566

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

Colaborative Universities: USP ­ Universidade de São Paulo; UNB- Universidade de Brasília; Unicamp ­ Universidade de Campinas Colaborative companies: Enalta Inovações Tecnológicas; Fischer S/A Agropecuária A lot of people should be mentioned here. In special we acknowledge the efforts of John Schueller, Dorota Haman, José Camargo da Costa, Jacobus Swart, Odílio B. G. de Assis, Andre Petronilho, Fabio Martin, Marcel Fleury, Alexandre Carbone, Fernanda Kalil, Marcelo Zorzan, Eliedson Pandorf, Marcos Catalfo, Ana Maria Felicori, Gilmar Victorino, Luiz Godoy, Jorge Novi, Cleber Manzoni, Eder Giroto, among others. Literature cited FAO Statistical Databases. 2002. http://apps.fao.org Fraisse, C.W., Heermann, D.F., Duke, H.R. 1995. Simulation of variable water application with linear-move irrigation systems. Trans. ASAE. 38(5):1371-1376. Sadler, E.J., Camp, C.R., Evans, D.E., Usrey, L.J. 1996. Irrigation system for coastal plain soils. Proc. Precision Agriculture: 3rd International Conference. Minneapolis, USA. vol.1 p.827-834. Scalable Node Address Protocol. 2005. www.hth.com/snap/ Schueller, J.K. 1997. Technology for precision agriculture. European Conference on Precision Agriculture. vol.1 p.33-44. Torre-Neto, A.; Schueller, J.K.; Haman, D.Z. 2001. Automated System for Variable Rate Microsprinkler Irrigation in Citrus: A Demonstration Unit. Proc. Third European Conference on Precision Agriculture. Montpellier, France June 18-20. p.725-730. CDROM.

567

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

Figure 1 - Variable-rate Irrigation System Architecture.

Figure 2 - Soil map obtained by ordinary kriging of nine measurement points in an arbitrary moment.

568

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

Réseau de capteurs sans fil pour l'irrigation à débit variable de citronniers

Mots clés : agriculture de précision, instrumentation fixe, cartographie de l'humidité du sol, irrigation site-spécifique Résumé L'état de Sao Paulo au Brésil a la plus importante surface de production de citronniers au monde. La plus grande partie de cette surface n'est pas irriguée mais du fait de la progression de nouvelles maladies, telle que la mort subite, cette situation va évoluer et la plupart des plantations vont être irriguées. C'est pourquoi des méthodes de sauvegarde de la ressource en eau doivent être développées. Cet article présente le développement d'une instrumentation fixe sans-fil (réseau de capteurs et d'actionneurs) ainsi que les outils logiciels correspondants pour irriguer des cultures pérennes de manière adaptée à chaque site. Les productions de citrons sont le premier objectif. Il s'agit de travaux en cours et seuls les aspects techniques sont présentés.

569

Precision Agriculture

Information and Technology for Sustainable Fruit and Vegetable Production

FRUTIC 05, 12 ­ 16 September 2005, Montpellier France

570

Precision Agriculture

Information

Microsoft Word - 060_111pPA.doc

8 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

543311


You might also be interested in

BETA
Microsoft Word - Engg Abs SLITE EMBEDDED 2011.doc
Microsoft Word - 060_111pPA.doc