School of Technology
Department of Computer Science
Master Thesis Project 30p, Spring 2013
Wireless Farming: a mobile and Wireless Sensor
Network based application to create farm field
monitoring and plant protection for sustainable
crop production and poverty reduction
By
Elias Edo Dube
Supervisor:
Bo Peterson
Examiner:
Carl Magnus Olsson
2
Contact information
Author:
Elias Edo Dube
E-mail: eliaswedajo@gmail.com
Supervisors:
Bo Peterson
E-mail: bo.peterson@mah.se
Malmö University, Department of Computer Science.
Examiner:
Carl Magnus Olsson
E-mail: carl.magnus.olsson @mah.se
3
Abstract
There is a remarkable growth in the field of Information Communication Technology (ICT) in Developing Countries (DCs). Telecommunication is one of the areas where ICT is recording an ongoing rapid change. Mobile phones are becoming pervasive in daily scenario; and among the beneficiaries of this are farmers. Farmers are using mobile phones in executing their farming business and daily life. At the same time, Wireless Sensor Networks (WSNs) are also showing a result in developed part of our world. WSNs potential in sensing various environmental condition, their affordability and applicability motivated conducting of this master thesis. Therefore, the objective of conducting this master thesis is to investigate and identify how the use of mobile phones in conjunction with WSN enable farmers in Ethiopia monitor and control their farm field. We use firsthand qualitative data we gathered during our field work in Ethiopia to design our proposed prototype. Functional requirements and system design guideless are obtained from observation we make and interviews we carry out on irrigation based farmers around town of Meki in region of Oromia. We use our prototype to demonstrate and evaluate how irrigation based farmers benefit from existence of such system.
Keywords: wireless sensor networks, mobile farming, internet of things, agriculture,
4
Popular science summary
Agriculture is one the important sector that accounts for growth in the developing world. Many developing countries are now focusing on agriculture led development. Yet the sector has its own challenges. The impressive progress of Information Communication Technology (ICT) in the Developing World and extent of Wireless Sensor Networks (WSNs) applicability indicates the possibility of applying this technology in the context of developing countries. In this master’s thesis we investigate and identify how the use of mobile phones in conjunction with WSN enables farmers in Ethiopia monitor and controls their farm field. We carry out an intensive literary review to learn the state of the art and identify similar works done before. Through a field work we conduct in Ethiopia we identify functional requirements to build a prototype. We demonstrate using our mobile and web based prototypes how the use of WSN can enable farmers in Ethiopia monitor their farm field using their mobile phones.
5
Acknowledgement
Foremost, I would like to express my sincere gratitude to my supervisor Bo Peterson for the motivating support of my master thesis project and for the useful academic insights and guidance he provided to throughout the course of this thesis project. Besides my supervisor, I would like to thank my examiner Carl Magnus Olsson for his useful feedback and important comments that helped improve the thesis work. My sincere thanks go to Annabella Loconsolle for the patient follow up of my thesis work and for offering me useful comments and feedback on all the deliverables.
I thank Irrigation based farmers from Meki, Ethiopia, specially - Abreham, Tesfalidet Aron, Tony Burka, Dadi Jara and others who were willing to participate in the interviews I conducted and for cooperating to show their farm field which enabled me conducted an observation and gather visual data.
Last but not least I would like to thank the SPIDER (The Swedish Program for ICT in Developing Regions) and its coordinators at Malmö University for offering me a travel grant which enabled me carry out a field work in Ethiopia and use a firsthand data in my research.
6
Table of contents
1 Introduction... 12
1.1 Approach for solution ... 13
1.2 Motivation ... 14 1.3 Project aim ... 14 1.4 Research Question ... 14 2 Literature Review ... 15 2.1 Use of WSN in agriculture ... 16 2.1.1 Irrigation management ... 17 2.1.2 Precision agriculture ... 17
2.2 WSN and communication technologies ... 19
2.3 Mobile phone use among farmers ... 20
2.3.1 Market price ... 20
2.3.2 Communication medium ... 20
3 Research Methodology ... 21
3.1 Case study ... 21
3.1.1 Planning and conducting interview ... 21
3.1.2 Participant observation ... 22
3.2 Literature review ... 22
3.3 Design and creation ... 22
4 Context of the field work ... 23
4.1 Agricultural management by farmers in Ethiopia ... 23
4.1.1 The East Showa zone ... 23
4.1.2 Type of irrigation farming ... 24
4.2 Interview and interview analysis ... 28
4.2.1 Interview method ... 28
7
4.2.3 Summary ... 33
5 Design and prototype ... 34
5.1 Design guideline ... 34
5.2 Design process ... 35
5.3 System architecture overview ... 37
5.3.1 Hardware Interface ... 37
5.3.2 Software Interface ... 39
5.4 Designing the prototype ... 39
5.4.1 Designing the data sensing and collecting system ... 39
5.4.2 Data logging unit design... 41
5.4.3 Data access system design ... 42
6 Prototype Evaluation ... 45
6.1 Evaluation of Functional requirements ... 45
6.1.1 Data request and verification evaluation ... 45
6.1.2 Data read and send evaluation ... 46
6.1.3 Data saving evaluation ... 46
7 Discussion and Conclusion ... 48
7.1 Limitations ... 48
7.2 Answering the Research Questions ... 48
7.3 Contribution ... 50
7.4 Future work ... 51
7.5 Conclusion ... 52
Appendix I: Interview Questions ... 53
1. Interview: Agricultural management and farm field monitoring by farmers in Ethiopia 53 Appendix II: use cases ... 54
8
9
List of Figures
FIGURE 1:WSN SENSOR NODES GENERAL STRUCTURE ... 16
FIGURE 2:EAST SHOWA ZONE,OROMIA,ETHIOPIA ... 23
FIGURE 3:A FIELD OF HARICOT BEANS AND ONION ... 25
FIGURE 4:FARMERS VISITING A TOMATO FIELD ... 25
FIGURE 5:WATER PUMPING MOTORS ... 27
FIGURE 6:MOTORCYCLES ARE USED BY FARMERS TO ACCESS A FARM FIELD ... 31
FIGURE 7:A PICTURE SHOWING A FARMER OUT IN THE FIELD AND AN IRRIGATION CANAL ... 32
FIGURE 8:A SOIL THAT LOOKS DRY ON THE SURFACE BUT WET INSIDE ... 32
FIGURE 9:DESIGN PROCESS STEPS ... 35
FIGURE 10:OVERVIEW OF ARCHITECTURE OF THE PROPOSED SYSTEM ... 37
FIGURE 11:SENSING UNIT ... 39
FIGURE 12:SENSORS ... 40
FIGURE 13:DATA LOGGING UNIT ARCHITECTURE ... 41
FIGURE 14:ARDUINO UNO BOARD AND ARDUINO ETHERNET SHIELD ... 41
FIGURE 15:DATA ACCESS UNIT ARCHITECTURE ... 42
FIGURE 16:ACCESSING SENSOR DATA USING THE WEB APP ... 43
FIGURE 17:WIRELESS FARMING SYSTEM MENU AND A SAMPLE SCREENSHOT ... 44
10
List of Tables
TABLE 1:SIMILAR WSN PROJECTS... 18
TABLE 2:CONNECTIVITY OPTIONS ... 20
TABLE 3:OVERVIEW OF RESEARCH METHODS ... 21
TABLE 4:BASIC STATISTICS FOR EAST SHOWA ZONE ... 24
TABLE 5: LANDS POTENTIALLY SUITABLE FOR IRRIGATION IN ETHIOPIA ... 26
TABLE 6:IRRIGATION WATER SOURCES, IRRIGATION EQUIPMENT USED AND METHODS OF WATER DELIVERY OR ABSTRACTION IN THE SELECTED DISTRICT OF EAST SHOA ZONE,JULY 2012. ... 27
TABLE 7:INFORMATION NEED OF FARMERS ... 33
TABLE 8:SYSTEM QUALITY ATTRIBUTES ... 34
TABLE 9:FUNCTIONAL REQUIREMENTS ... 36
TABLE 10:USE CASE 1:WEATHER DATA... 54
TABLE 11:USE CASE 2:SOIL MOISTURE DATA ... 54
TABLE 12:USE CASE 3:SOIL TEMPERATURE DATA ... 55
11
List of acronyms
AEZ Agro-ecological Zone
API Application Programming Interface BS Base Station
DC Developing Countries DS Decision Support ES Expert System
FDRE Federal Democratic Republic of Ethiopia GDP Gross Domestic Product
HCD Human-Centered Design
ICT Information Communication Technology
LIVES Livestock and Irrigation Value-chains for Ethiopian Smallholders MDG Millennium Development Goal
MEMS Micro Electro-mechanical System
OPDEDEZ Office of Planning and Economic Development of East Shewa Zone PA Precision Agriculture
RQ Research Question WFS Wireless Farming System WSN Wireless Sensor Network
12
1 Introduction
Agriculture is one of the important sectors for the economies of African countries. It is a highly relied upon sector for non-oil exporting African countries. The sector contributes for nearly 30% of the continents Gross Domestic Product (GDP) and 70% of the continents population depends on agriculture to sustain their life [1]. Facts from the World Bank about ‘Agriculture in Africa’ indicate that [2] - the agricultural sector has the potential to achieve the Millennium Development Goals (MDGs), reduce poverty, increase rate of employment and increase GDP in sub-Saharan Africa.
Agriculture is a major source of income for Ethiopia and the country’s economy highly depend on it. It accounts for half of the country’s total GDP and more than 80% of the country’s population depends on it [3]. The government of Ethiopia and other key stakeholders involved in agricultural work consider agriculture to be the main source of income, and a key role player for the country’s socio-economic development [4].
However, despite the fact that agriculture accounts to be a source of income and a supply for Ethiopia’s population livelihood, periodic drought and other environmental disasters common happenings many farmers compelled to face. Therefore, to overcome challenges in the sector various policies and strategies have been designed and implemented. These efforts have been made to enable farmers improve their productivity and further to facilitate a preventive methods to avoid risk. For instance, one of the key approaches that are being used by various stakeholders in agriculture to improve the production and productivity of the farmers is having Agricultural Information Systems and make use of Remote Sensing (RS) technologies.
In this master thesis project we plan to conduct qualitative research using a case study, literature review, and design and creation to investigate and identify how the use of mobile phones with existing Wireless Sensor Network (WSN) technologies be applied for agricultural purposes in developing countries taking the case of middle-scale irrigation-based farmers in Ethiopia. Our research tries to understand and demonstrate
13
how such application can benefit primarily farmers and other agricultural stakeholders – both Governmental Organizations (GOs) and/or International Developmental Organizations (IDOs).
1.1 Approach for solution
Nowadays, there are different approaches for solutions that are being used to improve crop production. One of the methods that are showing a good effect in improving crop production and effective resource utilization is Precision Agriculture (PA) [5]. Using PA enables farmers know what amount of fertilizers, seed and other chemicals to use for their land and specific condition. This makes PA an effective way for utilization of resources and improve production outcome.
Wireless Sensor Network (WSN) in agriculture is showing progress [9][10][11][12]. WSNs provide possibilities to sense and gather information of various environmental and crop conditions. However, it is a challenge for a farmer to know real-time data of a farm field and incorporate personalized weather information at same time using mobile phones. Therefore, in this master thesis project we plan to investigate this existing challenge in the field of agriculture by researching how WSNs can be used to monitor a farm and how mobile phones can serve to access the information.
14
1.2 Motivation
1.3 Project aim
1.4 Research Question
There is a rapid growth of Information Communication Technology (ICT) in the Developing Countries. Telecommunication technologies are becoming more available than ever. Mobile phones have become pervasive. Improvements that are happening in the telecommunication field are showing encouraging results. Farmers are among the beneficiaries of this opportunity. Farmers use mobile phones to run their business and stay connected. Almost every household is acquiring a mobile phone. This has created opportunities for farmers to do their job more efficiently and access market information [7][8].
However, there is a lack of research and applications done in the field of Agriculture which could shows using of mobile phones to enable farmers monitor their farm and allow them access existing services - like weather information. Studies [6][3][30][31][32][35][37][41][42][43] show that WSN to be a good opportunity for agricultural development and for researching the field.
Therefore, the growing use of mobile phone in Developing Countries and the affordability, energy efficiency, and ease of use of WSN technologies creates an opportunity to investigate how to enable farmers get real-time agricultural data and weather information on their mobile phones. In addition to this, it creates a chance to show how use of agricultural information can enable better farming decision [36][37][38][39][40][33].
The aim of conducting this project is to investigate and identify how the use of mobile phones in conjunction with WSN enables farmers in Ethiopia monitor and control their farm field.
RQ1: What are the common methods (or ways) used by farmers to monitor a farm field? RQ2: How can Wireless Sensor Networks in combination with mobile phone enable
15
2 Literature Review
Literature review is vital to have an in depth knowledge of one’s intended research area and to learn more about subject matter. Oates[13] describes the importance of carry out a literary review in order to identify research area, review state of the art and learn the area that needs further investigation or contribution. We conduct the literature review in order to identify the research gap and formulate a refined research goal. Therefore, the purpose of this literature review is to know more about the study area of the topic which we are going to research and to learn from previous works done by other researchers in the area. In addition to this we expect to gain a better insight of our own research question in relation to what have already been done.
We conducted web searching to obtain the literatures and assessed them in our literature review. We retrieve several articles from different online sources. Among the online sources we obtained the literatures are: IEEE1, Google Scholar2 and ScienceDirect3 via Malmö University’s proxy. The Institute of Electrical and Electronics Engineers (IEEE) provides access to a wide range of literatures published on more than 100 peer-reviewed journals in the field of electrical and electronics engineering and computer science. Google Scholar allows conducting academic related searches using Google’s search engine. Physical or digital copies of articles that are available online or in libraries can be searched using Google Scholar. Malmö University’s library allows signing in to Google Scholar using the library’s proxy which allows students access to a wide range of articles for free. The other main source we used to look up for articles is ScienceDirect. We used ScienceDirect to search for articles related to our topic from its full-text scientific database which offers access to journal articles and book chapters from more than 2,500 journals and almost 20,000 books.
We used different keywords related to our topic, research question and goal when searching for the literatures. We first review abstracts of articles carefully and keep track
1 http://ieeexplore.ieee.org.proxy.mah.se/Xplore/home.jsp 2
http://scholar.google.se.proxy.mah.se/
16
of each article accordingly. We also use bibliographies and references from articles to find more related articles to our topic. We finally review and assessed the articles we obtained. Of course, the process of our literature review was continuous throughout our project.
2.1 Use of WSN in agriculture
Wireless Sensor Networks (WSN) emerged from advancements in the areas of micro-electro-mechanical system (MEMS) technology, wireless communication, and digital electronics. WSNs devices are small in size, low cost, and require low power to work. The basic structure of WSN sensor nodes as identified by Chebbi et al. [50] shown below (see figure 1).
Figure 1: WSN sensor nodes general structure
According to Chebbi et al. [50] there are four main components that make up a sensor node. The parts are namely: a sensing unit, a processing unit, a transmission unit and a power unit. Depending on the type of application a sensor node may have additional parts such as a position finding system, mobilize and a power generator. Sensing unit usually takes the burden of sensing and gathering sensor data and then passes the data to the processing unit. The processing unit receives the sensed data and processes it according
17
to a set procedure or program. A transmission unit connects the sensor not with a network. The power unit supplies power required to run a sensor node.
Akylidiz et al. [14] identified five areas of application of WSN. These are for military applications, environmental applications, health applications, home applications and other commercial applications in offices and buildings. Other studies [10][11][12] conducted show that WSN technologies can be used in the area of agriculture as well. There is a growing demand for technological application in the developing world [27] and expansion of telecommunication infrastructure is growing rapidly [7][8]. The rapid growth in the field of telecommunication in the developing world is making farmer get access to Information and Communication Technological (ICT) Infrastructures. The potential of applicability of WSNs in the agricultural sector couple with the increasing growth of ICT in developing countries create an opportunity to explore the benefit of such application for farmers.
Below we discus some of the areas the use of WSN tried or studied in the field of agriculture.
2.1.1 Irrigation management
One of the promising areas that WSN could be used is in the field of agriculture is irrigation farming. Irrigation farming is a way of farming that uses various water sources to farm. The utilization of available water sources can be scarce, since this farming method allows farmers to farm throughout the year. Even though this method allows farmers to farm more than once a year, which raises profitability of farmers, over using of available water sources can lead to shortage of water. In this regard, WSN can applied for effective water management along with crop and soil condition monitoring [36][37][38][39][40][33].
2.1.2 Precision agriculture
One of the application area of WSN Precision Agriculture (PA) (see Table 1:Similar WSN projects) [32] [35] [37] [41]. A precision agriculture method is where the amount of inputs (like seed, fertilizer, pesticides, water etc.) given to a specific farm field is important to control and monitor the condition of the farm to determine the amount of the
18
crop output. Shah et al. and Coates et al. [15][16] describe sensor network based precision irrigation and problems and challenges of agricultural water management and how to improve water use effectiveness. The application of WSN in precision agriculture is one possible way to determine and control the inputs by monitoring the farm field for an improved and efficient precision irrigation [6][3][30][31] [42][43].
The table below (Table 1:Similar WSN projects) contains some projects and researches that used WSN components to improve different aspects in agriculture. The table shows list of projects and researches, their description, sensor used, crops involved in the study or project and devices used to access the information. We carried out a thorough investigation of these projects and researches in order to get an idea of what has been done and what is lacking.
Table 1:Similar WSN projects
Name Description Research
Prototyp e
Actual Product
Sensors used Crops Devices Refer ence
An
Enegy-efficient WSN for farming
Using WSN for PA in an energy efficient way
PC [9] Application of WSNs for Greenhouse parameter control in PA Use of programmable system on chip tech. to control the parameter of greenhouse for PA ZigBee sensor network(temprat ure, pressure, light, humidity,CO2, wind speed, and wind direction) PC [10] WSN in Agriculture: for potato farming Application of WSN to improve potato crop production Water depth, soil water tension Potato NA [11] WSN application in Agriculture
The future of farming and how it would be improved by using WSN
NA NA NA [12]
19
• Soil Moisture • Electrical conductivity • Soil temperature mationSoilNet A Zigbee based soil
moisture sensor network
Soil moisture NA PC [20]
COMMONSense Using WSN data for
effective water management VMC/VWC with soil moisture sensor, soil-matrix potential(SMP) with water mark sensor,
temperature sensor
NA PC [21]
Wireless Farming Using WSN to monitor a farm field and assist farmer make a decision Soil moisture Soil temperature Humidity temprature Onion, Tomato PC, Mobile
2.2 WSN and communication technologies
A WSN is composed of several sensor nodes that have the capacity of sensing and gathering data. The sensor nodes can sense varying types of parameters and send it to a central gateway [14] [50].
WSN sensor and processing boards have the capability of working with various communication technologies. WSN can be linked to external servers or services both with wires or wirelessly (see Table 2: Connectivity options). Amongst others some of connection options could be using Ethernet connection, Wi-Fi, Bluetooth, or
GSM-20
GPRS. In our case we plan to use GSM connectivity option both to link the sensor network to external database and/or to communicate with farmers mobile devices.
Table 2: Connectivity options
Communication ways Communication distance coverage
Bluetooth 30-100m
GPS-GPRS Network carrier coverage (km)
Wifi 100-300m
2.3 Mobile phone use among farmers
2.3.1 Market price
2.3.2 Communication medium
One amongst the routine daily activities of farmers is to keep track of their various businesses by assigning responsible person who could look over for them. So in order to keep themselves updates farmers call to person who is acting on behalf of them. Again the use of mobiles is high among farmers to communicate with their families and colleagues when they are not around [17][18].
Mobiles are widely used among farmers in developing countries to access market. Farmers make calls or send SMS to merchants at central markets in order to learn market prices and negotiate prices. The existence of mobile network coverage in most rural parts is allowing farmer to communicate with each other and others which makes them more informed about market conditions than before. The farmers who are ones used to be taken advantage of by merchants because of their lack of information, are now in turn taking advantage of the existence of helpful infrastructures like mobile and internet usage [17][18][30].
21
3 Research Methodology
3.1 Case study
3.1.1 Planning and conducting interview
We use Oate’s [13] book for planning and conducting interviews. Semi-structured interviews are planned and conducted on six farmers, one extension worker, and three other stakeholders. This facilitated to obtain data by enabling our interviewees to speak Oate’s [13] describe the process of conducting a research as a process which starts from setting a research question and all the steps that occur throughout the sequence of approaching and answering the question raised. According to Oate’s [13] a research process initially starts with set of questions to research and follows with series of activities which take the initial question to an answer or set of answers in order to present an evidence and conclusions to an academic audience and thereby show the creation of new knowledge or contribution.
In our qualitative research approach we used three methods – case study, literature review, and design and creation - to investigate the research area and answer our two research questions we raised. Table 3: Overview of research methods gives an overview of the research methods employed in answering each of our research questions throughout the process of our research activities.
Table 3: Overview of research methods
Research question Literature review Case study Design and creation
RQ1
RQ2
Case study is important research method to conduct an observational study to learn about an activity or a project and collect data [13]. In order to investigate about our instances we used farmers in East-showa zone in Ethiopia to conduct our case study. In order to generate data we used both interview and observation. This gave us an input to learn in detail about the activities and processes of farming and the farmers.
22
with more details on the issues we raise and allow them to raise their own issues that they think are relevant [13] (see Appendix I: Interview Questions). The interviews are recorded in audio formats and kept for analysis.
3.1.2 Participant observation
As part of our data gathering process we conducted participant observation. We observed various activities of farmers in East-showa zone, Ethiopia near a town called Meki. We gathered data by taking notes and taking pictures of the participants and the environment in addition to interviews we conducted.
3.2 Literature review
3.3 Design and creation
Literature review is vital to have an in depth knowledge of one’s intended research area and to learn more about subject matter (see 2 Literature Review). We conduct the literature review in order to identify the research gap and formulate a refined research goal. Therefore, the purpose of this literature review is to know more about the study area of the topic which we are going to research and to learn from previous works done by other researchers in the area. In addition to this we expect to gain a better insight of our own research question in relation to what have already been done.
In this study we use the Design and Creation method to design a prototype for monitoring a farm field using mobile phones and wireless sensor networks. The Design and Creation approach is useful to create a new IT product [13][22]. After identifying requirements from the data obtain from interviews and observation, the next process we carry out is to design a prototype that can be evaluated. Since this study is not concerned solely about the creation of a prototype, we plan also to make our work a factor that contributes to existing knowledge. Therefore, the contribution to knowledge shall be shown using the data gain from literature review and case study. The prototype will be used to demonstrate this.
23
4 Context of the field work
4.1 Agricultural management by farmers in Ethiopia
4.1.1 The East Showa zone
The East Showa zone is located in central Oromia region, Ethiopia [25]. The Central Eastern Rift valley crosses the zone. Majority of the lakes of the region are found in East Showa. The zone records annual temperature of 15 to 27 and a mean annual rainfall of 410 to 820mm. The main soil types of East showa zone are Andosols(36.47%), Vertisols(16.12%), Rendizome and Phaeozomes(22.94%) and Fluvisols(2.05%)[25][45]. Agro-ecological Zone (AEZ) type of the region are mainly Dry Weina dega and Moist Weina Dega[44].
24
Table 4: Basic statistics for East Showa Zone
Source: Profile of East Showa Zone [43]
4.1.2 Type of irrigation farming
Huib et al.[26] identifies the crops grown in the region to be vegetables (eg. onion, tomatoes, haricot beans) and flowers. We also observed this to be true from observation we had during our field work. The cultivation of these crops depends on irrigation water [28][29]. Flowers that are grown in greenhouses are estimated to use between 2000 and 4000 m3 water per ha; while vegetables grown in open field require an estimation of 500-800mm ha-1 (5000-8000 m3 ha-1) of irrigation water.
Selected Districts
Total
Bora Dugda Lume
Rain fed crops 27805 54175 46030 128010
Irrigated crops 7292 5965 - 13257
Communal /opening graze 1911 3964 44654 50529
Private grazing - 7361 - 7361 Forests/woodlots - 3172 4200 7372 Plantation - 239 - 239 Land covered by irrigated vegetables (ha.) Onion 6819 3008 1745 14894 Tomato 1688 Hot pepper 97 Cabbage 597 Eggplant - Green pea 477 Land covered by irrigated fruits (ha.) Papaya 365 88 Watermelon - Mango 6 Orange - Banana - 4
25
Figure 3: A field of haricot beans and onion
26
Lakes and rivers located in Rift Valley Basin are major source of irrigation water for crops developed in the region. In the East Show zone, Dugda woreda - where we carried out our field work – the main water sources used for irrigation farming are Lake Ziway, Meki River and Ground Water. Farmers said sometimes it is difficult to identify between where the good water comes from and where source of harmful water is. For instance they told us that at some of the sites the water is salty and that affects badly the crop growth and soil fertility.
Table 5: lands potentially suitable for irrigation in Ethiopia
No River Basin Area ('000ha)
1 Abbay 5800 2 Awash 406 3 Baro-Akobo 1100 4 Genale-Dawa 660 5 Mereb 38 6 Omo-Ghibe 348 7 Rift Valley 80 8 Tekeae 1383 9 Wabi-Shebele 335 Total 10150
Source: FDRE Ministry of Water and Energy[24]
Motor pumps are irrigation equipments mostly used by farmers. The motor pumps are used to pump water from available source of water near to the plantation. Most farmers access the Lake Zeway water that crosses their field through the major irrigation canal in the district. Some farmers use river Meki water. While some other use water from their own well.
27
Figure 5: Water pumping motors
Table 6: Irrigation water sources, irrigation equipment used and methods of water delivery or abstraction in the selected district of East Shoa Zone, July 2012.
Characteristics Lume Bora Dugda
Irrigation water sources
Koka Dam( 1150 ha) (3 PAS) River 1186 ha in 3 PAS Lake Zeway 5914 ha in 16 PAS River 476 ha ( 11 pAs) Private owned shallow wells 120 ha (7 PAs) River 1242 ha in 8 Pas
private owned shallow wells 120 ha (7 PAs)
Private owned shallow wells 3319 ha 13 PAs Irrigation equipment
used Motor pumps( 1745 ha)
Motor pumps( 7184 ha)
Drip irrigation set Motor pumps Water delivery
/abstraction Gravity
Pumping (motor
pump) Pumping (motor pump)
Pumping (motor pump) Gravity Gravity Irrigation potential
(ha) 2075.13 9009 14050
Actual irrigated area
(ha) 1744.75 7184 9523
28
4.2 Interview and interview analysis
This section presents the result of a field work carried over a period of 20 days in Ethiopia around the city of Meki (Dugda, Oromia)[44][45]. The objective of conducting the interview is to identify requirements of the farmer. Carrying out the interview allows learning ways farmers used to monitor and control their farm field and common challenges they experience. This will be used as guidance while identifying requirements (see Table 9: Functional requirements) and making design decisions.
In total we had Ten participants in our interview which of this - eight of them are middle scale irrigation based farmers, one is an extension worker and one is agricultural officer. The farmer participants had a varying level of experience in their farming business. Some are new to the farming business while some have been in the farming business for a while and have large scale farms. After we found the contact of participant of our interview we arrange time and place to interview them. After each interview we also conduct a field visit of each farmer to have a look on the crops they are growing and learn the way they access information about their farm and what kind of ways they use to monitor the crop and their farm.
4.2.1 Interview method
Oates [13] describes interview as a way of data generation technique by getting information from the interviewee. An interview needs to be planned before proceeding to conduct one. A well planned and conducted interview is a key to gaining significant information about issues we want to get information. The IDEO HCD toolkit [22] provides also important techniques for conducting interview for Human-Centered Design. The toolkit provides advises on how to prepare interview questions and get prepared for the interview. It is recommended to make interview questions less abstract and use examples.
4.2.1.1 Defining, planning and conducting interview
We use Oates’ [13] guideline to define and plan our interview questions. Interviews are designed to generate data. The defined interviews are semi-structured interview. This means that all interview questions are not pre-determined before actually conducting the
29
interview. This provides a chance to capture relevant issues that may arise while conducting the actual interview.
We use IDEO HCD toolkit [22] as a guide when conducting our interview. In the process of conducting the interview we followed the guide provided by IDEO how to find people for our interview putting into consideration the level of understanding both we have about the topic we are investigating and the knowledge and understanding our interviews have about the technologies we are using. We start by acknowledging existing knowledge of our interviewees about the topic. By creating such a knowledge base of the topic and getting an understanding of the level of knowledge our interviews have about the topic, we ask for questions we prepare and take note of relevant issues that arise while conducting the interview.
4.2.2 Farm field monitoring methods used by farmers
A usual vegetable plantation has mainly the following production cycles. These are land (or field) selection, seed preparation, seed plantation, growing the crop, gathering crop when it is ready and putting the product on market. One full production cycle takes three to four months of duration. During each production cycle a farmer has different concerns and activities to achieve a good production.
During land selection process a farmer want to know whether a soil is suitable for the desired seed type the farmer is planning to plant or not. During this process one activity that needs to be carried out is to check whether the soil is salty or not. Farmers we interviewed expressed they try to identify this using indigenous knowledge about the environment. This means they rely on the information they obtain from a person who has used that specific land or another land close to it. Usually they observe characteristics of the land and effects noticed on plants to determine the nature of the land; that is to determine whether soil is salty or not. However, they fail often in identifying whether a certain soil is good or bad.
The next action a farmer takes after selecting a land suitable for the desired planting is preparing the seed to be planted. This usually does not require a lot of effort. The only thing that concerns is the quality of the seed chosen. The quality of a seed chosen determines the growth potential, the production outcome, and its resistance to plant
30
diseases. At the same time a seed is being prepared for plantation; the land for plantation is plowed and gets ready for irrigation farming.
When the seed is ready for plantation, it is taken to a field prepared for the plant. From this time onwards the farmer starts preparing irrigation schedule and close follow up and monitoring. The farmer continuously visits the plant and checks if it needs watering or not. A farmer also sees the weather condition and humidity of the field and decides if the plant needs pesticides or not. With such continuous follow up and monitoring the farmer takes care of a plantation until production is ready for the market.
The primary ways of accessing farm field is by using their motorcycle and a mobile phone. A farmer frequently travels to different sites to checks status of crop. Having an eye on a farm field is a significant part of farmers’ daily activities. Most of the farmers we interviewed told us they visit their farm field at least twice a day to check the status. Since mostly they have plantations on more than two sites, only checking status of their field takes a lot of time, energy, and resource. Since the only way this farmers control and monitor the condition of their farm field is by being physically present to every site they have, they buy motorcycles. All farmers we contacted during our interview and observation owned their own Motorcycles. Another way of getting information about a farm is using mobile to call someone who is at a farming site to ask about status of a field or a crop.
31
Figure 6: Motorcycles are used by farmers to access a farm field
Every time a farmer visits a farm field diagnoses soil of the field, the environmental condition and the crop status. The first thing a farmer does is to look if the farm has been watered or not. If not, the farmer checks by looking at the soil if watering is needed or not. Usually farmers mistakenly consider that the farm needs watering when they see a dry soil on surface while the soil is wet 5-10cm deep. This causes over watering of the plant. This affects the growth of the crop planted. Another observation a farmer conduct is to look at the leaf of the plant and soil to distinguish whether the water contains saltines or acidity (PH). And subsequently, a farmer checks air moisture and temperature, humidity, and temperature to decide whether the farm needs fertilizers and/or pesticides. We will show how we used this information to make our design choices in Chapter 5.
32
Figure 7: A picture showing a farmer out in the field and an irrigation canal
33
4.2.3 Summary
The information need of the farmers we contacted is similar. Most of the farmers we interviewed expressed their need of information regarding soil type and crop follow up methods. As mentioned (on section 4.2.2) above, farmers are concerned about how to choose a good farm land suitable for seed they are planning to plant and how to prevent a plant from a potential plant diseases. Farmers do this using their knowledge they got from personal experiences and indigenous knowledge. Table 7 below shows farmers need of information in order to monitor their farm field and take necessary action to prevent plants from plant disease.
Table 7: Information need of farmers
Information need Reason Action taken
Soil moisture identification To identify the need for watering
watering
Soil pH To find out soil saltines or acidity
Decide use of fertilizers
Air moisture and air temperature
Risk for plant diseases Pesticides and use fertilizers
humidity Plant disease Use preventive
pesticides
Weather condition Rain forecast Make watering and
34
5 Design and prototype
5.1 Design guideline
We use guideline from IDEO Human-Centered Design approach [22] and additional quality attributes (see Table 8: System quality attributes) of the hardware equipments we used in our project. By involving farmer in a semi-structured and user centered interview (see 4.2.1.1 Defining, planning and conducting interview) we have tried to learn existing trends farmers use and their level of understanding both about their situation and available technologies. Combining the qualitative input from the interviews conducted and information we acquire about the state of the art (2 Literature Review), we define quality attributes (see Table 8 below) to achieve a human-centered, feasible and usable design.
Table 8: System quality attributes
Quality attribute Description
Availability: The system shall be available to be accessed by its users using available mobile networks in the area. Lifetime: Installed Wireless Farming system shall last for more
than two production seasons.
Affordability: The mobile and web applications; and the wireless sensor system all together shall be affordable for the farmers to purchase and use.
Security: The system shall be secure enough both software and hardware.
Usability (or desirability): The mobile and web applications shall be easy to use and understand.
feasibility and viability The overall system shall maintain feasibility and viability
35
5.2 Design process
The overall design process goes through three main phases – Hear, Create and Deliver - identified using the IDEO HCD toolkit [22]. These three phases served as a key in our human-centered design process which enabled us to hear the needs and desires of farmers involved in our research, create innovative and inclusive solutions which meets the needs of farmer we are designing our product for, and finally enables delivering a solution that is affordable, accessible, sustainable, secure, feasible and viable system (see Table 8: System quality attributes). Therefore, the design solution emerges from overlap of the human-centered design process (see Figure 9: Design process steps) we follow and the overlap of the quality attributes we identified (see Table 8: System quality attributes).
Figure 9: Design process steps
Figure 9: Design process stepsabove shows the designs process our system. Firstly, we start our human-centered design process with hearing from farmers to identify problems. During this phase we dedicate most of our research time to hear from farmers what they have to say about challenges they face. This provides a means to capture relevant information to the identification of challenges they have regarding farm field monitoring, management and agricultural problems they have. In order to obtain all the information we need necessary for our design, we conduct a semi-structured interviews and carry out observations (see 3.1 Case study and Appendix I: Interview Questions). In addition to
Create
Hear
Deliver
Problem Indentification Method: Interview Observation Brainstorming Functional requirement Prototype development Prototype evaluation Design choice36
hearing from farmers, we also interview agricultural experts and extension workers which provide extra information to what we already obtained from farmers.
The second phase of our design process is – Create. In this phase we use the information we obtained in the first phase to identify functional requirements of our proposed system. In order to achieve this in a human-centered approach we carry a brainstorming session with farmers. During our brainstorming sessions we let farmer discus freely what they think would be interesting design and what the proposed system should let them do in order to monitor and manage their farm fields. We use paper sketches, create user scenarios and paper prototypes in order to illustrate and visualize ideas. The outcome of brainstorming sessions provides a means to define functional requirements (see Table 9: Functional requirements) and design choices. We also take into account information we obtained from interviews conducted during first phase of our design process. In addition to this, we use functional requirement definition provided by Sommerville [23] in order to get more insight on how define functional requirements.
Finally we use the results obtained from the first two phases of our design process to deliver a solution using a prototype of our proposed system. We use the identified quality attributes and functional requirements in the process of making the design decision of our proposed prototype.
Table 9: Functional requirements
Data.Request
Data.Request.Verification
Data.Request.Verfification.Fail
The mobile application shall let a user access a sensor data from the Wireless Farming system
The system shall check whether the user has a verified user (or a register user)
If a user is not a verified or authorized user, the system shall not allow the user access the sensor data
Data.Read The system shall read sensor data over some interval of time and log it to a cloud service provides by xivity.com [24]
37
Data.Value.Send The system shall send the read sensor values to the user
Data.Value.Save While sending read sensor values to a user the system shall also save access history to a database
5.3 System architecture overview
Our proposed system is comprised of both hardware and software components (see Figure 10: Overview of architecture of the proposed System) below. We make our design decision using the design choices we obtain from our brainstorming session (see Figure 9: Design process steps) and identified quality attributes (see Table 8: System quality attributes). Therefore, we used open source hardware and software interfaces to design our proposed system.
Figure 10: Overview of architecture of the proposed System
5.3.1 Hardware Interface
• Micro-controller: Arduino uno board – this microcontroller is based on ATmega328 datasheet. It has 14 digital input/output pins, 6 analog inputs, a USB connector, a power connector and built in clock speed resonator [51].
38
Summary
Microcontroller ATmega328 Operating Voltage 5V
Input Voltage (recommended) 7-12V Input Voltage (limits) 6-20V
Digital I/O Pins 14 (of which 6 provide PWM output) Analog Input Pins 6
DC Current per I/O Pin 40 mA DC Current for 3.3V Pin 50 mA
Flash Memory 32 KB (ATmega328) of which 0.5 KB used by bootloader
SRAM 2 KB (ATmega328)
EEPROM 1 KB (ATmega328)
Clock Speed 16 MHz
• Ethernet shield: Arduino Ethernet shield – this Ethernet shield enables to connect and Arduino to internet using a RJ45 cable [52].
Summary
• Requires an Arduino board
• Operating voltage 5V (supplied from the Arduino Board) • Ethernet Controller: W5100 with internal 16K buffer • Connection speed: 10/100Mb
• Connection with Arduino on SPI port • IEEE802.3af compliant
• Low output ripple and noise (100mVpp) • Input voltage range 36V to 57V
• Overload and short-circuit protection • 9V Output
• High efficiency DC/DC converter: typ 75% @ 50% load • 1500V isolation (input to output)
39
• Sensors: Temperature sensor, photo sensor, soil moisture sensor, humidity sensor, PH sensor
5.3.2 Software Interface
• Arduino Development software • Xively API
• Acosm
• Android development kit
5.4 Designing the prototype
In this section we present main functionalities of the proposed system. The proposed system performs three main functionalities. The first and foremost activity of the proposed system is to sense and collect environmental data. Then the next process is to log the collected data to a cloud server. The final activity in the process is to make the logged data available for visual access. Therefore, our design of the proposed system corresponds to these identified processes of the system.
5.4.1 Designing the data sensing and collecting system
Figure 11: Sensing unit
The data sensing unit consists of sensor nodes which are placed in a farm field to sense and monitor different environmental conditions (see Figure 11: Sensing unit above). The sensors monitor farm field conditions like – soil, humidity, temperature and
40
weather conditions and send sensed data to base station. The specific type of necessary sensors for our proposed system are identified using the functional requirements (see 5.2 Design process) generated and literature reviews conducted on wireless sensor networks (see 2.2 WSN and communication technologies ).
Data is generated within some interval of time that is set or when a user sends a request to the system. In the first case of data generation, the sensor senses the environment and sends the data to the web application. This occurs according to a schedule that the system uses to put itself in active or sleep mode in order to save power and lifetime. During the second case – when a user sends a request – the system collects sensor data and sends it back to user’s mobile. This enables a farmer to send a request to base station to check status of farm field. A base station processes the request received and replies to request with sensor data read. This means, for instance, if a farmer sends a request for soil moisture statues, a BS checks the sensor reading for the corresponding request and replies with the collected data value.
41
5.4.2 Data logging unit design
The data logging unit is used to send the sensed data passed to the base station (see Figure 13: Data logging unit below). Base stations are used to process and send sensor readings to the internet.
Figure 13: Data logging unit architecture
Data sensed using the sensors is passed to the Arduino micro-processor. Arduino then receives the sensor data and logs it to Xively cloud server using the Xively API [46][49]. Data is logged to the server on certain interval of time that is set. The Arduino micro-processor that serves as a data logging unit is also responsible of executing sensor readings. This makes it serve as base station of the system. For our prototype we programmed data to be logged every second to simplify the process of evaluating the system at later times.
42
5.4.3 Data access system design
The next step we took in our design the proposed system is to decide the way of making the sensor data available for the user. The sensor data uploaded to the internet using the data logging unit can be accessed from both personal computers (PCs) and mobile phones (see Figure 15: Data access unit architecture). We considered the presentation of sensor reading to give meaning to the user. That is current sensor reading values should be displayed along with a graphical visualization of sensor reading. The user will be able to view also the physical location of a farm field and other details. The sensor reading presentation is accessible in two ways. One option to access sensor reading is by using the web. The other option is to access sensor reading using mobile phones.
Figure 15: Data access unit architecture
5.4.3.1 Accessing data using web visualizer
The sensor readings are continually uploaded to Xively [46][49] cloud service and made available for access from any web browser using internet. We use Xively’s API service to feed our sensor data to channels we created on their cloud service. It is possible to view current sensor reading value both visually and numerically. The web application provides a graphical presentation of sensor readings over some period of time – which can range from current time up to three months of reading history. Using the web application a user can view geographical location of a field a sensor is located.
43
Figure 16: Accessing sensor data using the web app
5.4.3.2 Accessing data using mobile application
Mobile phones have become a part of our daily activities. With the rapid growth of telecommunication technology in developing countries [17][18][47][48], the availability of mobile phone services and their use among farmers is growing at the same time. Our research aims how existing WSN technologies be used in conjunction with mobile phones. It is also our researches objective to see ways how the combination of these two technologies could be used to enable farmers monitor and control their farm field.
As part of one our interview question (see Appendix I: Interview Questions) we ask farmers for a type phone they have. We used the answer of this interview question to decide the platform on which we shall develop mobile based prototype.
The figure below (Figure 17) shows a screenshot of the main menu showing the menu items and a sensor reading. The main menu items we included in our prototype are humidity, soil temperature, soil moisture and farm temperature data streams.
44
Figure 17: Wireless Farming system menu and a sample screenshot
In the figure below (Figure 18), shows a graphical view of a sensor reading of one day. The value on the horizontal bar shows the sensor reading while the vertical bar represents the time span. In addition to this sensor readings are shown on basis of an hour, four days and over 30 days. The graphical representation of the sensor readings makes it easy to visualize current status and sensor reading history over period of time.
45
6 Prototype Evaluation
In this chapter we present the evaluation of our prototype. Our evaluation takes into account only evaluating the mobile part of the proposed system. The purpose of evaluating the proposed prototype is to present our solution to participants and get their feedback. The participants whom we used to evaluate our prototypes are the farmers we used for interview. Besides getting users feedback about usability of the proposed system, evaluating our prototype helped check whether the system fulfils the function requirements specified (see Table 9: Functional requirements).
6.1 Evaluation of Functional requirements
6.1.1 Data request and verification evaluation
• The mobile application shall let a user access a sensor data from the Wireless Farming system
Accessing a sensor reading using the mobile phone application is possible. The mobile phone application (Acosm) efficiently retrieves sensor data reading uploaded to Xively. Users are prompted to try to use it. We observed a users have slight level of difficulty to understand accessing the application for the first time. After a while of trial all users we checked the product with were comfortable using the application. At the end of every user’s evaluation we ask for their opinion.
• The system shall check whether the user has a verified user (or a register user) The mobile application verifies and authenticates when a user tries to use the application. A user must provide the required credentials in order to access the system. This makes every application to be accessed only by an authorized user. This provides a specific application to be accessed only by a specific user that the system is created for.
46
• If a user is not a verified or authorized user, the system shall not allow the user access the sensor data
The application does not allow a user access the system unless the user provides correct authentication information. This is important to keep the system secured and prevent an authorized access. Therefore, in order to access sensor reading data a user should have a registered account.
6.1.2 Data read and send evaluation
• The system shall read sensor data over some interval of time and log it to a cloud service provides by xively.com [24]
The Arduino based system we designed is capable of reading data on a given interval of time and uploads it to Xively’s data stream. Sensor data is uploaded to a specific data stream using API key and feed ID of the data stream.
• The system shall send (or upload) the read sensor values to server
The proposed system frequently reads sensor values and uploads it to cloud. This sensor readings can be accessed using the web or from a mobile phone application. Data is uploaded within a given time interval defined for the system.
6.1.3 Data saving evaluation
• While sending read sensor values to a user the system shall also save access history to a database
At this version of the prototype the system saves sensor readings on Xively cloud service. A user can access sensor reading records up to three months. Since one plantation season for vegetable crop lasts for three months, it could be enough data to monitor throughout the plantation season. However, it is important for a farmer to keep farm field information for longer time. It helps to make decision for consequent
47
plantation seasons. In order to decide on what crop type to plant next, and identify which season and soil is best for a certain crop type, it is important for a farmer to get access of sensor data that have been saved for long time.
48
7 Discussion and Conclusion
7.1 Limitations
The proposed system’s mobile based prototype has been evaluated with farmers. This was conducted in order to get feedback from future users of the system (see 5.1 Design guideline and Figure 9: Design process steps) and confirm the requirements identified have been met in the system (see Table 9: Functional requirements ). By doing so we have are able to get insight for future work – both to improve the system and enhance the usability. However, ensuring the usability and functionality of the mobile part of our proposed system’s prototype was important, more test and evaluation needs to be carries out on the hardware part of the proposed prototype as well. The functionality of the designed hardware prototype have not been tested and evaluated under actual environmental conditions out in the field. This might lead to a failure if the system is implemented without testing the hardware components of the proposed system out in the field. Furthermore, it is the mobile phone application part of the prototype that we evaluated. We gave higher priority for the mobile part over the web part of our application since the main aim of our research is to investigate how mobile phones and WSNs are used to enable farmer in Ethiopia control and monitor their farm field. The budgets allocated to conduct the field work in Ethiopia is also of limited amount to allow us do everything.
7.2 Answering the Research Questions
In this section we provide how we answered our research questions.
RQ1: What are the common methods (or ways) used by farmers to monitor a farm field?
To answer this research question, we conducted a field work in Ethiopia in which we interviewing farmers, done field observation (see section 3.1: Case study and section 4.2: Interview and interview analysis). Then we used the interview and observation results to learn and understand common methods farmers employ to
49
monitor and manage a farm. By identifying the common ways of farm field monitoring used by farmers and learning the parameters a farmer use to monitor a plant, we identify user requirements and check the need for quality attributes(see 5.1Design guideline).
RQ2: How can Wireless Sensor Networks in combination with mobile phone enable farmers monitor a farm field?
For answering this research question we used the result we obtained from the interview we conducted. We asked the participants in our interview (middle scale irrigation based farmer in Meki,Oromia, Ethiopia) what type of mobile phones they use ( see Appendix 1.Interview: Agricultural management and farm field monitoring by farmers in Ethiopia). In addition to this, we did a literature review to examine existing mobile application development environments that work with WSN technologies (see section 2.1Use of WSN in agriculture and section 2.3: Mobile phone use among farmers). We used the results to answer this research question by designing a prototype (see chapter 5Design and prototype). Our prototype shows the possibility of managing farm field and accessing services using mobile phones.
The data we gathered show that most of the farmers use Samsung galaxy S2 and Iphone 4 mobile phones. There are also few farmers who have Blackberry mobile phones. The diverse type of mobile phone usage among farmers made it challenging to identify a platform to develop our prototype on. However, after comparing the functional requirements we identified and technical information we obtained from our literature review on WSN, we decided to make an Android based application.
We used the feedbacks we received from our prototype evaluation with farmers in Ethiopia. After each time a user evaluates our prototype, we recorded textually users experience and level of satisfaction to examine the usability and usefulness of our proposed system. We also prompted users if they find such a
50
system to be of use to assist them in their daily activities. The results we achieved from our assessment show that farmers think such system to be of greater help to carry out their daily farming activities. It is in this manner we approached to answer this research question using the feedbacks we obtained from farmers during our prototype evaluation (see section 6 Prototype Evaluation).
7.3 Contribution
In this thesis research we have studied similar products (and researches) that existing out in the market (see section 2 Literature Review). This enabled us to get significant inputs to conduct our research and identifies areas that need knowledge contribution as well.
We found most projects on WSN in agriculture to be research oriented (see Table 1:Similar WSN projects). Our literature review and field work in Ethiopia indicate lack of research on use of mobile phones and WSNs technologies to enable middle-scale- irrigation-based-farmer in the country monitor and control their farm field. For this reason, we conducted a research which investigates ways of using mobile phones in conjunction with WSNs to enable farmers in Ethiopia monitor and control a farm field. We used functional requirements we identified from a firsthand data we obtained from our field work and extensive literature review we conducted to address the issues we raised and answer our research questions (see section 1.3Project aim).
Furthermore, our research motivates the possibility of carry out an extensive research in the area of WSN and mobile phone technologies; and the role this could play on improvement of agricultural methods in the context of developing countries.
51
7.4 Future work
The research presented in this thesis indicates some more areas to exist for future research.
Firstly, a study of how to use alarming system in a Wireless Farming System (WFS) can be investigated. Including an alarming system in WFS involves sending a warning or alarm to a farmer’s phone when a certain condition occurs. For this, more research needs to be conducted to identify the parameters and learn the conditions that are severe for plants in context of DCs.
Secondly, we believe that such a system designed for farming use requires including a Decision Support (DS) unit. Most of the farmers who participated in our interview expressed they use their indigenous knowledge and own experience to make decision while using fertilizers and pesticides. This exposes crops under treated or use of more resource than require. In short, further study that investigates a cost effective and scientifically helpful ways on how to associate sensor data reading with professional explanation. One way to do this is to conduct a study of designing and Expert System (ES) which supports decision making by associating sensor readings with professional (scientific) information.
Finally, what we see as another potential of future work is integrating existing services in WFS. For instance, a farmer can be enabled access weather forecasts from within the same system. This creates the condition not to leave the application alone but makes system suitable for the farmer to make a irrigation schedule while viewing other farm field conditions. Commodity exchange information can be also included. Therefore, it is relevant to study available existing services and APIs that could be of significant importance for farming.
52
7.5 Conclusion
In conclusion, with a user focused design approach, our research show that mobile phones can be used as a tool to enable farmers in DCs monitor a farm field. We have identified the main concerns farmers have regarding carry a diagnosis of different aspects of their farm field and following up the status of a plantation. we relied on data we obtained from our field work and literatures we reviewed to create a context to assess to what extent WSNs be used to monitor farm field conditions and work with existing mobile phone platforms. We identified user requirements common among middle scale irrigation farmers in Developing countries taking the case in East Showa zone, Ethiopia. We used inputs we got from our field work and literature review to formulate a design guideline and make a design decision on how to create a prototype. In addition to learning about requirements we have evaluated our prototype with users to validate its functionalities meet their requirements. Furthermore, we have managed to answer our research questions. And the result of the proposed Wireless Farming System (WFS) shows the possibility of enabling farmers in DCs monitor their farm field using their mobile phones.
53
Appendix I: Interview Questions
1.
Interview: Agricultural management and farm field
monitoring by farmers in Ethiopia
The objective of conducting this interview is to gather a firsthand data from farmers in Ethiopia to learn how farmers manage and monitor their farm field. We do this to identify system design requirements. All the information you provide will be used for the master thesis report only.
Name:
Where you Work:
1. What are your main daily activities? 2. What type of crops do you plant?
3. How big is your farm? How much investment does it require? How is the profit margin?
4. What are the common risks of production loss?
5. Can you briefly explain how you keep track of your farm field during the day? 6. How do you access your farm field (both physically and remotely)?
7. How do you use your mobile phone in your farming business? What type of mobile phone do you have? Can you please show me?
8. If a sensor based application is installed to enable you monitor your farm field using your mobile phone and internet services, would you be willing to pay for that?
54
Appendix II: use cases
Table 10: Use Case 1: Weather data
Use Case ID 1
Use Case Name Weather data Created by Elias
Actors Users/farmers
Description This action allows a farmer to send a weather information request to the system and gets weather condition detail
Preconditions ·The user must have a functional WSNFarming system.
· The user must have the WSNFarming system app on his mobile phone.
· There should be a mobile network coverage both where the WSNFarming system is installed and at the location of the user Postconditions The user will get a weather information detail on his mobile phone. Normal Flow ·A user logs in to the app on his phone.
·Selects weather information option from the menu.
· If the user has mobile network and his number is register on the system the app will send a request for sensor data to the system ·The system then reads the sensor data and also checks other weather
stations information
·The system sends the sensor data to the user’s mobile phone. ·The system updates access history on the database.
Alternative Flows
1. Use the web based application.
Priority Medium
Frequency of Use
Depends on the user’s preference.
Table 11: Use Case 2: Soil Moisture data
Use Case ID 2 Use Case Name
