See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/304287793

Design and implementation of a connected farm for smart farming system

Conference Paper · November 2015

DOI: 10.1109/ICSENS.2015.7370624

CITATIONS READS
113 20,238

6 authors, including:

Minwoo Ryu Sungchan Choi

Korea University Korea Electronics Technology Institute
10 PUBLICATIONS 267 CITATIONS 23 PUBLICATIONS 424 CITATIONS

SEE PROFILE SEE PROFILE

Jaeho Kim
Samsung
42 PUBLICATIONS 1,230 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

oneM2M View project

Social Interaction Design with Shared Media Systems for Domestic Environment View project

All content following this page was uploaded by Ting Miao on 09 January 2018.

The user has requested enhancement of the downloaded file.

Design and Implementation of a Connected Farm for
Smart Farming System

Minwoo Ryu, Jaeseok Yun, Ting Miao, Il-Yeup Ahn, Sung-Chan Choi, Jaeho Kim
Embedded Software Convergence Research Center
Korea Electronics Technology Institute
Seongnam, S. Korea 13509
Email: {minu, jaeseok, novamartin, iyahn, csc, jhkim}@keti.re.kr

Abstract—Agriculture has been one of the most important world will be able to be ‘connected’ one another, and share
industries in human history since it provides humans with their status and surroundings, finally providing new innovative
absolutely indispensable resources such as food, fiber, and energy. services to humans (even without human intervention) [6]. This
The agriculture industry could be further developed by employing is mainly due to the advances in ICT such as identification sys-
new technologies, in particular, the Internet of Things (IoT). In tems (RFID), short-range wireless communication (ZigBee [7],
this paper, we present a connected farm based on IoT systems,
Bluetooth, WiFi), and cellular networks [8] and widespread use
which aims to provide smart farming systems for end users.
A detailed design and implementation for connected farms are of smartphones.
illustrated, and its advantages are explained with service scenarios
compared to previous smart farms. We hope this work will show Recently, Techworld pointed out the promise of growing
the power of IoT as a disruptive technology helping across multi agricultural productivity by adopting IoT-related technologies
industries including agriculture. [9]. With the literature survey and our experience on devel-
oping IoT systems, previously existing smart farming systems
I. I NTRODUCTION can be enhanced by employing IoT systems in two aspects: 1)
make it easy to extend systems by allowing new type of devices
In human history, the agriculture has been one of the most to be efficiently and effectively integrated into smart farms; 2)
important industries for humans’ living since it is responsible facilitate horizontal smart farm platforms, which enables all
for producing indispensable resources such as food, medicine, smart farms to be connected and take advantage of experts’
energy, fiber. Thus, even after the industrial revolution, most (i.e., experienced farmers) farming knowledge.
countries (even developed ones) have been emphasizing the
essential roles of the agriculture industry and related tech- The smart farm, embedded with IoT systems, could be
nologies affecting agricultural production. For example, heavy called a connected farm, which can support a wide range of
equipments and agricultural planes deployed were able to devices from diverse agricultural device manufacturers. Also,
increase the work efficiency in agriculture (while saving human connected farms could provide more intelligent agricultural
resources), dramatically increasing agricultural productivity. services based on shared expert knowledge. For example,
As with advances in other industries, the agriculture indus- people having even little experience on farming could grow
try has been also accelerating to develop by deeply employing plants or crops for profits. Infectious disease prevention could
information and communication technologies (ICT). In particu- also be another benefit of developing the connected farms by
lar, automated farm systems, built with diverse wireless sensor detecting influenza virus in a specific pig farm, and proactively
devices and actuators, are able to monitor the environmental isolating that one from others.
conditions and control the deployed devices according to the
collected data through wireline and wireless access networks. In this paper, we demonstrate a connected farm, which aims
Lin and Liu presents a remotely controlled farm farmers can to provide suitable environment for growing crops based on the
monitor and control using smartphones or tablets without vis- IoT systems developed in the previous research project [10].
iting [1]. Akshay et. al also introduced a similar work [2]. Lee All sensors and actuators for monitoring and growing crops
and Yeo proposed a pig farm monitoring system, which can are connected with a gateway installed with a device software
effectively manage the farm by monitoring the environmental platform for IoT systems, called &Cube. The gateway is in
data using temperature/humidity sensors and video cameras, turn communicated with an IoT service server, called Mobius.
and accordingly control farm facilities such as humidifiers and Accordingly, the Mobius not only monitors the environmental
air conditioners [3]. Kaewmard et. al introduced an automation condition of the connected farm by communicating with the
system based on wireless sensor network techniques, designed gateway installed into the connected farm, but also talks with
to monitor the agriculture environment [4]. They also devel- expert farming knowledge systems and controls actuators in
oped an irrigation system to collect environmental data and order to make the farm suitable to grow crops.
support remote control of the operation via mobile devices.
The rest of this paper is organized as follows. The design of
At present, we are on the verge of a technological revo- connected farms is introduced in Section 2. Section 3 illustrates
lution, called the Internet of Things (IoT). The IoT term was implementation of our proposed connected farm, and Section
coined by Ashton in 1999 [5], and represents the future of 4 describes the possible service scenarios using the connected
computing and communication, where all the objects in the farm. Finally we conclude our remarks in Section 5.

978-1-4799-8203-5/15/$31.00 ©2015 IEEE

 farm send collected environmental data from monitored areas
to the &Cube, and then the &Cube transmits the data to the
Mobius, as shown in Figure 1b. Finally, end users can monitor
and control their connected farm using IoT applications on
their smartphones or tablets by accessing virtual representation
of the devices resided in the Mobius, as shown Figure 1c.
Here, it should be noted that all interfaces provided by the
Mobius are representational state transfer (REST) application
programming interfaces (APIs). That is, the interfaces between
the Mobius and &Cubes could be standardized and thus easily
extended by integrating with other IoT gateways installed with
the &Cube, regardless of the type of devices connected into
the gateways and access networks. Also, from the point of
view of IoT application developers, they can implement their
applications using REST APIs provided by the Mobius in order
to offer their specialized farming services to customers, even
though they have little knowledge about the devices of the
connected farm. As a conclusion, our proposed design for
connected farms enables seamless deployment of IoT devices
into connected farms, but also fosters third party developers
to develop their innovative farming services and applications
using REST APIs of the Mobius platforms.

III. I MPLEMENTATION
In this section, we introduce our connected farm implemen-
tation including its three main components, Mobius, &Cube,
and physical devices.

A. Mobius
The Mobius is an IoT service platform complying with
Fig. 1. Design of connected farms based on IoT systems, (a) register devices oneM2M standards, which supports to create virtual repre-
as virtual representations in the Mobius, (b) transmit data collected from the sentations of physical IoT devices. The oneM2M initiative
registered devices via &Cubes, (c) monitor and control the connected farm is an international partnership project established in 2012,
through REST API-based IoT applications. and its aim is to standardize machine-to-machine (M2M) and
IoT service layer specifications for globally-applicable and
access-independent M2M/IoT services [11]. The Mobius is
II. D ESIGN OF C ONNECTED FARMS designed to comply with the oneM2M specifications, and sup-
The connected farm is a new type of automated farming ports common M2M/IoT service functions, including device
system developed using IoT infrastructure. As shown in Figure registration, data repository and management, subscription and
1, the connected farm system has three main components, notification, security, etc. Thus, the Mobius can support virtual
i.e., connected IoT devices including monitoring sensors and representations of the data collected from physical devices
controllers, IoT gateway (called &Cube), and IoT service plat- installed into the connected farms. Furthermore, as mentioned
form (called Mobius). In connected farms, there are physical in Section 2, the Mobius provides REST APIs to access data
sensors (e.g., temperature, humidity, CO2 , illumination) and resources resided in the Mobius, and control IoT devices.
controllers (e.g., sprinkler, LED lights, air conditioner, heater)
for monitoring and controlling the environmental conditions of B. &Cube
the farm. All the sensors and controllers are connected with the
IoT gateways in turn connected with the IoT service platform. The &Cube is a device software platform (i.e., a mid-
End users (i.e., farmers) can interact with the connected farm dleware), which can be installed into IoT gateways. As in
for monitoring its environmental conditions or triggering some the Mobius, the &Cube is also designed to comply with the
farming utilities. oneM2M standards, and thus send the collected data from
physical devices to the Mobius via standardized REST APIs.
The overall operational procedure is as follows. First, The &Cube has been implemented as a Java program so that
each device (i.e., sensors and controllers) deployed in the it could be easily ported on any type of embedded machines
connected farm has to be registered into the Mobius using installed with the Java virtual machine (JVM). In addition, the
&Cube, as shown in Figure 1a. Once registered with the &Cube supports various protocol bindings including HTTP,
Mobius successfully, every IoT device can have its virtual MQTT, and CoAP. For our connected farm, we have installed
representation in the Mobius in the form of a resource type. the &Cube in a Raspberry-Pi, which is a single-board Linux-
When running the connected farm, sensors for monitoring the installed computer.
Fig. 2. Our connected farm implementation.

Fig. 3. Three types of sensors to monitor internal environment of the

connected farm, (a) a compound sensor for monitoring temperature, humidity, Fig. 4. Six types of controllers to control the environmental conditions of
CO2 , (b) PPFD sensor, (c) soil moisture sensor. the connected farm.

C. Implementation of the Connected Farm IoT applications developed with REST APIs provided by the
Mobius.
Our connected farm is built in the rooftop of one building
in the Korea Electronic Technology Institute. Figure 2 shows
D. Mobile Application
the photo of the implemented connected farm.
We have developed a smartphone application to enable
We deployed three types of sensors in our connected farm
users to remotely monitor and control the connected farm,
in order to monitor the environmental conditions, including
as shown in Figure 5. Currently, the application is developed
a compound sensor (i.e., temperature, humidity, and CO2 ),
for the Android-based smartphones. End users can monitor
a photosynthetic photon flux density (PPFD) sensor, and a
environmental conditions of the connected farm and control
soil moisture sensor, as shown in Figure 3. All sensors are
the controller deployed in the connected farm in real-time.
wirelessly connected with the Raspberry-Pi installed with
the &Cube through a ZigBee-based network, and the virtual For smart farming services, the application provides two
representations of the physical devices are created as a cor- main menu. Figure 5a shows the monitoring menu. The
responding resource type stored in the Mobius. Each sensor monitoring menu allows end users to check the environmental
uploads the data collected from its surrounding environment conditions of the connected farm including temperature, hu-
to the Mobius through the &Cube at a 20-second time interval. midity, CO2 , and illumination. All the environmental variables
collected at a 20-second interval can be retrieved from the
In order to build an automatic connected farming system, Mobius, and plotted in a graph. Figure 5b shows the control
we deployed six types of controllers in our connected farm, menu. By pushing buttons in the control menu, end users can
as shown Figure 4. The controllers include intake and exhaust remotely control the controller deployed in the connected farm.
fans (a-b), an air conditioner with heating and cooling (c),
six sprinklers (d), LED lights (e), a cover controller for As described in Section 2, all the interfaces used in the
preventing the light from reaching the plants in the connected smartphone application are REST APIs provided the Mobius.
farm (f), an irrigation and nutrient management system (g). This implies that any application developers who are familiar
As with sensors, all the controllers are connected with the with the REST APIs can develop more user-friendly, easy-to-
Raspberry-Pi installed with the &Cube. Each controller can use GUI applications than our prototype one, which explains
be operated remotely through the &Cube and Mobius, that is, exactly our design principle for IoT-based connected farms.
 V. C ONCLUSION
In this paper, we have presented a connected farm based
on IoT systems for smart farming systems. To provide Internet
connectivity for the sensors and controllers of the connected
farm, we have deployed the &Cube, a standardized (i.e., com-
plying with oneM2M specifications) device software platform
for IoT devices. Also, we have used the Mobius, an IoT service
platform (also oneM2M-compliant) that provides REST APIs
with which the data collected from sensors (e.g., CO2 sensor)
can be retrieved, but also the control commands can be sent to
controllers (e.g., air conditioner). We have also implemented
a smartphone application that allows end users to remotely
monitor and control their connected farm, e.g., turn on air
conditioner by pushing a button on the smartphone. Our service
scenario with the connected farm shows a promise of building
a farm knowledge expert system that encourages people to
work in the agricultural industry.

(a) (b) ACKNOWLEDGMENT

Fig. 5. Smartphone application to monitor and control the connected farm,

This work was supported by Institute for Information &
(a) monitoring menu, (b) control menu. communication Technology Promotion (IITP) grant funded by
the Korea government (MSIP) (Development of IoF Linked
Open Knowledge for Improvement about Quality of Farming
and Farm Product [2014K000173, Development of Open IoF
IV. S ERVICE S CENARIO USING THE C ONNECTED FARM
Server Platform]).
Our service scenario can be described as follows:
R EFERENCES
1) All IoT devices including sensors and controllers have [1] J. Lin and C. Liu, “Monitoring system based on wireless sensor network
to be connected with the &Cube, and then the &Cube could and a SocC platform in precision agriculture, ” in Proceedings of the In-
register them into the Mobius through the REST APIs. ternational Conference on Communication Technolgy (ICCT), Hangzhou,
China, pp.101–104, 2008.
2) A farm knowledge base (like IBM Watson) will be [2] C. Akshay, N. Karnwal, K. A. Abhfeeth, R. Khandelwal, T. Govindraju,
able to be built, which aims at providing farm knowledge on D. Ezhilarasi and Y. Sujan, “Wireless sensing and control for precision
Green house management,” in Proceedings of the International Confer-
behalf of experts (i.e., experienced farmers) by sharing the ence on Sensing Technology (ICST), pp.52–56, 2012.
optimal growth conditions for diverse crops and plants. Thus, [3] H. Lee and H. Y, “Design and implementation of pig farm monitoring
the knowledge base will provide rich guidance for farmers to system for ubiquitous agriculture,” in Proceedings of the International
control their connected farm effectively and efficiently. Conference on Information and Communication Technology Convergence
(ICTC), Jeju, S. Korea, pp.557–558, 2010.
3) After all IoT devices are successfully registered into the [4] N. Kaewmard and S. Saiyod, “Sensor data collection and irrigation con-
Mobius, farmers can remotely monitor and control the IoT trol on vegetable crop using smart phone and wireless sensor networks for
smart farm,” in Proceedings of the International Conference on Wireless
devices to make their connected farm appropriate for their Sensors (ICWiSE), pp.106–112, 2014.
crops and plants. For example, if the soil moisture sensor [5] K. Ashton, “That ‘Internet of Things’ Thing,” RFiD Journal, vol. 22, no.
uploads a particular value less than a threshold suggested by 7, pp. 97–114, July 2009.
the knowledge base, farmer receives an alert message, and [6] ITU-T, “ITU Internet Reports 2005: The Internet of Things – Executive
triggers the irrigation system immediately. Summary,” 2005.
[7] P. Baronti, P. Pillai, V. W. Chook and S. Chessa, “Wireless sensor
4) Farmers can create their own connected farm control networks: A survey on the state of the art and the 802.15. 4 and ZigBee
policy, which will automate the monitoring and control proce- standards,” Comput. Commun., Vol. 30, pp. 1655–1695, 2007.
dure we described above. [8] M. K. Karakayli and G. K. Foschini, “Valenzuela, R.A. Network coordi-
nation for spectrally efficient communications in cellular systems,” IEEE
5) Finally, farmers can share their knowledge and expe- Wireless Communications, Vol. 13, pp.56–61, 2006.
rience for growing their crops and plants to make the farm [9] A. Savvas, “Farming industry must embrace the Internet of Things to
grow enough food,” Techworld, Available: http://www.techworld.com/
knowledge base smarter. news/big-data/farming-industry-must-embrace-internet-of-things-
3596905/.
The service scenario described above demonstrates the ad- [10] M. Ryu, J. Kim, and J. Yun, “Integrated semantics service platform for
vantages of our connected farm-based smart farming systems. the Internet of Things: a case study of a smart office,” Sensors, vol. 15,
The connected farm allows end users not only to remotely no. 1, pp. 2137–2160, January 2015.
monitor and control the physical devices deployed via mobile [11] J. Swetina, G. Lu, P. Jacobs, F. Ennesser, and J. Song, “Toward
applications, but also to share their knowledge and experience a standardized common M2M service layer platform: Introduction to
on farming, and finally build a farm knowledge expert system oneM2M,” IEEE Wireless Communications, vol. 21, no. 3, pp. 20–26,
June 2014.
that enables farmers with little experience to start a farm.

View publication stats