CHAPTER TWO

LITERATURE REVIEW

In some regions of the world, cattle are not considered for use in the production of

food. Studies have shown that the work power, fertilizers, milk and the fuel from

dung that the cattle provide. In these regions are more efficient animal products

than meats. Analysis of other cultural practices has often revealed unexpected

efficiency of use fitted to local circumstances.

Environmental influences such as climate also play an important role in the

domestication and use of animals. Water buffalo are use as draft animals to pull

wagons and farm equipment in Southern Asia, where they are adapted to the high

temperature and humidity, while horses, which thrive in moderate climates, were

the principles draft animals in the temperature regions until they were replaced by

treated cattle humid conditions are prevalent in the Southern United States because

they are better adapted to the climates of the region than European cattle.

2.1 Related Works

A study of Swedish precision farmers confirmed the need for a user-centered farm

information system in information- and technology-intensive farms (Norros, et al.,

2009).

1
Norros et al. (2009) utilized the Internet for the communication method with

potential users. A scenario video describing the system’s key functional features

was available to anyone online with the evaluation questionnaire to provide

feedback to the developers. Some for the main results from this survey uncovered

user needs including tools to evaluate the effects of different cultivation practices,

analysis of existing farm data, tools to evaluate the influence of a certain

cultivation practice on the whole farming business, methods to improve farming,

and others (Pesonen, et al., 2008).

Following this, Pesonen et al. (2008), gave recommendations and guidelines for a

novel, intelligent, integrated information and decision support framework for

planting and control of mobile working units which they implemented in the

InfoXT project. They defined farmers’ attitudes towards their work and profession

to formulate the coretask-based system usability claims (Pesonen, et al., 2008).

Their evaluation methods focused on the system performance rather than

individual user interface (UI) components. The system evaluators were expected to

have considerable experience and insight into the domain area (Norros, et al.,

2009). After system validation and extensive analysis, they conclude that

information management systems in mobile plant production environments should

be internet-based with an open interface, and that farm data saved in a central

2
database should be accessible to the farmer through internet servers (Pesonen, et

al., 2008).

The European Union funded FutureFarm project identified the information model

for six field operations (tillage, seeding, fertilizing, spraying, irrigation, and

harvesting), and selected the information model for fertilization for analysis. The

project specified the data provided and the information required for decision-

making and used this to derive the flow of information which, in turn, resolves the

design of the system. The analysis of the information model focused on the farmer

as the primary decision maker (Sorensen, et al., 2010).

2.2 Farm Management

The farm management personnel’s role is becoming progressively more complex

as pressure from trade globalization, sustainability, and complex information and

communication technologies becomes more prevalent (Sindir, 2006; Sorensen, et

al., 2010). An important task of farm management continues to be increasing the

farm’s production and profit and to do so by properly allocating resources and

implementing appropriate operational strategies. With the domestic and global

markets becoming more competitive, farm managers must push their operations to

maximum efficiency.

3
The United States Department of Agriculture (USDA) recognized that a typical

model for farm management does not exist because not all farms are the same size

and type (Sindir, 2006). Instead, the agency acknowledged farm management has

five basic functions to achieve the goals and objectives of the farming business:

1. Planning: creating daily/seasonal priorities and schedules, recognizing areas

needing improvement and finding alternative solutions

2. Organizing: establishing standard operating procedures and a structure of

activities to meet the farm’s goals

3. Managing human resources: recruiting, hiring, training and evaluating workers

4. Directing: delegation of responsibilities, establishing good communication with

workers

5. Controlling: monitoring expenses and income, maintaining records of

operations, comparing rates of productivity, making changes to practices as

necessary, and more (Sindir, 2006).

Sorensen et al. (2010) also developed an engagement web from the farm

manager’s point of view with the circular gray dashed line representing the farm

system boundary, and the darker gray dashed oval representing the system

boundary for the production of crops.

4
While the farm manager is not likely to deal with all of these issues at once, many

of these issues can be arranged into successive and parallel tasks. Many farm

production processes and their sub-processes are diverse in nature and require the

management to balance individual tasks within the seasonal farm plan (Pesonen, et

al., 2008). For example, spring tillage and planting are performed in succession on

a field basis, but they are often performed in parallel across the farm. Spring

planting requires a succession of tasks such as ordering seed and fertilizers, seed

bed preparation, coordinating resources and more. Each task demands the attention

of someone on the farm and can change priority based on weather, resources

available, and the completion of previous tasks.

Farm management requires numerous decisions of various kinds, and most have

many implementation options (Sindir, 2006). Typically the farmer deals with very

complicated information flow paths since information comes from several sources

(Pesonen, et al., 2008; Sorensen, et al., 2011).

2.3 Livestock Farm Management Information Systems

Operating an agriculture enterprise requires the management to make and

implement numerous decisions across the operation throughout the season, and

many times, away from the office. The management must process the necessary

data in terms of recording, conditioning, and correlating (Sindir, 2006). The

5
decision-makers also use information from a variety of resources, but the most

valuable is often the source with information specific to the farm’s operations,

which often includes financial and operation records (Sindir, 2006). LFMIS can be

designed to deal with these issues and to support strategic and operational

decisions (Nurkka, et al., 2007).

Nurkka et al. (2007) indicated that the management of information and decision-

making are the core issues for successful farming. Nurkka et al. (2007) and

Murakami et al.(2007) indicated data acquisition is not a hindrance for precision

agriculture. While Sorensen et al. (2010) point out that the use of computers and

the internet has improved the task of handling and processing information, it

remains a demanding task for the farm managers. They emphasize that there is

potential of integrating various data sources when suitable information systems are

developed and these should improve management practices. A variety of LFMIS

have been available to farmers, but unfortunately, the adoption of these systems

has been relatively slow compared to other popular consumer-grade information

and communication technologies.

2.4 Requirements of LFMIS

Multiple studies and surveys have been conducted in attempts to better understand

the needs of the growers. Sorensen et al. (2010) noted the importance of

6
understanding how the farmer views current information management methods

including what he/she thinks is working well and what is not working well. It is of

great importance for the LFMIS developer to understand what the farmer needs to

make his/her daily working life easier and what would help the farm run more

effectively (Sorensen, et al., 2010).

The following LFMIS requirements were found to be the most important by

Murakami et al. (2007) and several of these also directly relate to adoption:

 A system for the specific needs of the farmers

 A simple UI

 Simple or automated methods for data processing

 A user controlled interface allowing access to processing and analysis

functions

 Integration of expert knowledge and farmer preferences

 Improved integration of standardized computer systems

 Enhanced integration and interoperability

 Scalability

 Interchangeability between applications

 Low cost

7
Fulton et al. (2013) echoed many of these same requirements based on a survey of

farmers from the Midwestern and Southern United States and agriculture

professionals nationwide during the winter of 2012/2013. They also elaborated on

some desires and requirements from growers:

 Automatic wireless data transfer between machines and with cloud storage so

data is stored in a single location

 Resources to find local support and training

 Web-based LFMIS so data can be accessed from an internet-connected device

 Quick-start guides to make sure the technology is being set up correctly

 Standardized data formats and compatibility between different machines and

operating platforms Murakami et al. (2007) recommended an open software

platform as an appropriate solution rather than a single proprietary system

because it is unlikely that any single complex and comprehensive solution could

meet all the requirements listed above.

When considering the core tasks of farm recordkeeping software, Pesonen et al.

(2008) identified the following steps when focusing on managing field operations:

o Creating the operation plan

o Delivering detailed task plan to the field

o Setting up mobile working units to execute the plan

8
o Managing, controlling and recording the operation

o Documenting the as-applied operation for recordkeeping

In recognizing the various information sources, Pesonen et al. (2008) stated that

these various sources needed to be easily integrated and combined for different

analyses.

Murakami et al. (2007) noted that a simple UI was an important requirement when

designing an LFMIS. More specifically, Haapala et al. (2006) concluded that the

information presentation and consistency with UI components was critical. They

recommended that only information necessary to carry out a task should be

presented to the user, and information needed to be presented in a logical order.

They also recommended that icons needed to be designed such that they had a

clear meaning with respect to the task at hand. They found that inconsistency and

lack of clarity, as well as poor choice of icons and language were likely sources of

usability problems (Haapala, et al., 2006).

2.5 Identification of Technologies Available for Implementation

Technologies are identified here which build on the idea that new services and

technologies can be added to the system as they become available. Independently

developed services from other interested parties, assuming they are built upon the

9
same concepts and hosted externally, can complement an open-nature of system as

envisioned in this research project.

2.5.1 Data Sources

Data specific to the farm is possibly the most valuable source of information to

support decision-making (Sindir, 2006). Thus, the keys to the success of any

LFMIS are accurate and timely generation and access to this data. Fortunately,

useful data already exists, albeit in many forms within typical agriculture

production systems. Understanding this variety of sources is necessary to enable

specialization of an LFMIS within a farm (Welte, et al., 2013 a.).

2.5.1.1 Manual Data Entry

The simplest method of data collection is manual input. This traditionally consists

of handwritten notes made with pen and paper. Long, standardized forms can

overwhelm operators if particular attention is not paid to the user experience.

Providing simple, specialized apps for a variety of data entry tasks is crucial to

getting standardized, minable data into the cloud where it can be put to use. Most

people will not use mobile devices for data entry if such a switch entails more

work, higher learning curves, and longer entry times than their existing system.

Therefore, each data collection task should be automated to the extent possible. By

making data entry faster and simpler than pen and paper, data in the cloud will be

10
both more complete and more correct than inaccessible stacks of paper notebooks.

Examples of manual data entry include: recording field, operator, rate, and tank

number as anhydrous ammonia is applied,

recording chemical mix, field, and date that a pesticide was applied, and recording

seed variety, fertilizer, and area during planting (Welte, et al., 2013 a.).

2.5.1.2 Machinery Data

Almost all machines and implements involved in modern production agriculture

have sensors that are critical to machine operations and automation and can also

create useful data in real-time during operation (Steinberger, et al., 2009). Some

examples include: vehicle location, seed population, chemical application rates,

wheel slip, fuel usage, crop yield, crop moisture, PTO status, hydraulic remote

actions, and many others.

The proprietary, non-standard nature of these machine sensors has traditionally

limited their usefulness due to an inability for outside systems to access them. As

compliance with the international standard for controller area network (CAN)

communications on serial busses in agriculture and forestry (ISO 11783)

communications standard progresses, this hurdle is reduced but not yet eliminated.

11
Inexpensive, wireless networks of sensors using Bluetooth for communication

would enable smart phones to collect data that is not tied to a particular proprietary

source.

While Bluetooth is not the ideal communications platform for sensor networking, it

is generally inexpensive and widely implemented in smart phones. Sensors with

relatively low data rates, such as ID tags and contact sensors can be easily

retrofitted on existing machines and implements to provide information to

autogenic algorithms (Welte, et al., 2013 a.).

The following section identifies a few machinery data projects and devices.

ISOBUS Controller Area Network (CAN) Data Connections Purdue University –

ISOBlue: The project aims to create a completely open source, inexpensive means

for getting data from any ISO11783-compliant tractor to a Bluetooth-equipped

mobile device in real-time. The mobile device can then upload the data to the cloud

over its existing cellular connection. Enabling farmers and researchers to access,

analyze, and store their own data will vastly improve the ability of precision

agriculture technologies to finally reach their long- awaited potential of using

statistical data mining techniques to optimize many features of agricultural

production from yieldto environmental impact (ISOBlue, 2013).

12
Crop Ventures, Inc. – CANPLUG: The CANPLUG device was developed by Crop

Ventures, Inc. to support new and existing agricultural data software. The device

plugs into the equipment ISOBUS diagnostic connector and can forward sensor

data to web and mobile software. Similarly to ISOBlue, the CANPLUG runs on a

Linux Operating System and utilizes Bluetooth connectivity to transfer data to

smartphones and tablets (Crop Ventures, Inc., 2014).

Implement Identification Tags

John Deere – Implement Detection: Implement Detection is part of the John Deere

Farm Sight solution, and works with any ISOBUS-compatible implement. When

paired with a Green Star 3 2630 Display and the John Deere Implement Detection

Controller, it can help operators reduce errors by ensuring implements are set up

exactly the same year after year. The system remembers the last setting used and

helps the operator get to work faster (Deere & Co., 2013 b.).

2.5.1.3 Internet-Based: Weather, and Geospatial Data

Many types of useful information for LFMIS are already publicly available online.

However, accessing this data is sometimes quite difficult due to a lack of

application programming interfaces (API), and a general lack of data format

standards. LFMIS which can utilize data which does not need to be manually

13
collected will greatly facilitate adoption and increase its ability to provide useful

analysis.

Some examples of potentially useful data available within the United States

include:

o Weather Data: Provided by the National Weather Service Advanced

Hydrologic Service (National Weather Service, 2013). Daily, monthly, and

yearly precipitation amounts are available going as far back as 2005. The

data are derived from a combination of radar and rain gauge measurements.

Other weather data of interest could include temperature and wind speed.

o Soil data: Available from the USDA NRCS Web Soil Survey (USDA

Natural Resources Conservation Service, 2013). It consists of geo-located

polygons representing the survey map units, and tabular data with soil

attributes to which the polygons are referenced.

o Light Detection and Ranging (LiDAR) Elevations: This extremely precise,

remote sensed elevation data is available from the Open Topography project

(Open Topography, 2013). Most LiDAR data has a horizontal resolution of

1.5 meters or less, but only 28% of the United States excluding Alaska was

covered as of 2011 (National Oceanic and Atmospheric Administration

(NOAA) Coastal Services Center, 2012). Potential issues with this high

14
 resolution data include data conditioning and delivering specific data sets to

mobile devices (Noel, 2014).

o Common Land Units (CLU): “A Common Land Unit (CLU) is the smallest

unit of land that has a permanent, contiguous boundary, a common land

cover and land management, a common owner and a common producer in

agricultural land associated with USDA farm programs. CLU boundaries are

delineated from relatively permanent features such as fence lines, roads,

and/or waterways“.

o CLU borders are available to Farm Service Agency, Natural Resource

Conservation Service, and Rural Development employees through the

USDA Geospatial Data Gateway (USDA Farm Service Agency, 2013). CLU

borders for the majority of the United States are available from AgriData,

Inc. through their Surety software with a paid subscription. “Due to Section

1619 of the 2008 Farm Bill the CLU borders are in Surety and Surety Pro

are dated May of 2008“ (Chad, 2013).

o Cropland Data Layer (CDL): This remotely sensed georeferenced raster data

depicts detailed information on crop and non-crop land use to explore

landcover and land-use change in the contiguous United States (Han, et al.,

2014).

15
 o This spatial data is available through the CropScape (USDA National

Agricultural Statistics Services, 2014) web application for visualization and

download.

2.5.2 Data Transfer

The advancements in wireless data transfer technologies should enable LFMIS

providers to move away from the obsolete method of transferring data cards.

Wireless networks are great tools to help farmers automatically transfer data back

and forth between machines, employees, and the office (Pesonen, et al., 2008;

Fulton, et al., 2013). The following section identifies some data transfer

technologies common with consumer grade mobile devices.

Bluetooth

Bluetooth wireless standard is a technology for convenient and secure wireless data

transfer over short distances, up to 100 meters, using radio transmission.

Thetechnology allows paired devices to share voice, data, music, photos, videos

and more.

It has been built into billions of mobile devices and an ever expanding list of other

products including cars, medical devices, computers, and many more (Bluetooth

SIG, Inc., 2013).

16
The Bluetooth Special Interest Group (SIG) recently released the Bluetooth 4.0

specification, Bluetooth Low Energy (BLE), to simplify the Classic protocol and

eliminate some Classic features. These changes enabled battery-powered mobile

devices with BLE to achieve power savings necessary to extend the battery power

(Balmos, et al., 2013).

Wi-Fi

The Wi-Fi term is actually a certification for wireless local area network (WLAN)

devices (Wi-Fi Alliance, 2014). The Wi-Fi Alliance (2014) specifically defines

Wi-Fi as any “WLAN products that are based on the Institute of Electrical and

Electronics Engineers’ (IEEE) 802.11 standards.”

Wi-Fi technology is another extremely popular wireless data transfer technology

with consumers. It uses radio wave transmissions for medium range data transfer

and is a very common networking solution for homes, businesses, schools, airports,

cafes, and more. Nearly all consumer-grade mobile devices have a Wi-Fi modem

built in from the device manufacturer.

Cellular Data Networks

Cellular data networks allow mobile devices to connect to the internet when a Wi-

Fi connection is not available. The range of cellular data network signals can reach

beyond 8 km (five miles) from the cellular network tower in rural areas. In the

17
United States, cellular data providers include AT&T, Verizon, Sprint, and others,

and require each user purchase a data plan. The common cellular network

technologies include GSM, CDMA, and LTE. These technologies are widely

available in the United States with varying connection reliability and transfer speed

depending on the provider’s coverage in theuser’s area (Miser, 2012). Many device

manufacturers build products with cellular modems built in, and it has become a

very popular technology in consumer-grade mobile devices.

The potential for network interruptions is a major risk for a LFMIS. A network

disruption is more likely to be caused by lack of network availability than by a

network failure at the provider’s end. This type of interruption is unlikely because

the service is likely maintained by professionals. A user would be unable to access

the internet-based services in the event of network disruption regardless of its

cause. The risk can be mitigated by enabling the user to load critical data to the

mobile device when a connection is available to avoid major issues (Pesonen, et

al., 2008).

18