Internet of Things in the Cloud

Theory and Practice

Philip Wright and Andrea Manieri

Researchers, Engineering Ingegneria Informatica S.p.A., Rome, Italy

Keywords:

Cloud, Internet of Things, Smart City, Arduino, Sensors, Raspberry PI, Pachube, Cosm, Xively, Nimbits,

oVirt, NoSQL DB, CouchDB, ZABBIX.

Abstract:

The digital convergence of IT, Network and Content produced during the last years has created a new land-

scape for the deﬁnition and delivery of ICT services to citizens. The massive introduction of hypervisors and

virtualisation techniques has allowed the (self-service) provision of any resource (software, network, data) in

a seamless way to Users and Applications. The term Future Internet was coined in EU to synthesize these

concepts and refer to the novel architectures enabling the next generation of Internet applications. The digital

convergence includes also the so-called embedded systems that through Gateways connected to the Internet to

provide a technical bridge to interact with sensors and actuators. This paper introduces the initial ﬁndings from

experimental work done in the context of ClouT

, a EU-funded project aiming at deﬁning and developing a

common virtualisation layer, allowing the access and management of IoT devices, as well as Cloud Services,

in the same way as any other data.

The authors demonstrate and provide a simple way to design and implement a real infrastructure that satisﬁes

the following requirements: cheap, easy to maintain, open source based, compatible and interoperable with

different platforms and services. It is an example of how to make a public or a private Cloud capable of hosting

data that comes from different kinds of sensors. Moreover this architecture allows the interconnection of all

devices (Internet of Things, IoT) and its implementation is illustrated in detail.

ClouT: Grant agreement, for ”Collaborative project”, number 608641, FP7 (Seventh Framework Pro-

gramme). Project full title: ”ClouT: Cloud of Things for empowering the citizen clout in smart cities”.

1 INTRODUCTION

The ClouT project regarding the ”Cloud of Things”

(http://clout-project.eu/) is the ﬁrst project that

merges the IoT and the Cloud Computing concepts.

One of the main focus’ of the ClouT project, and of

this paper as well, is not only to provide a system

capable to send data coming from different sensors

or actuators into the Cloud, but that can also work

with legacy devices and different communication pro-

tocols.

In order to respond to the described objectives, a re-

search in the open market has been made, looking for

a device that can give us the possibility to collect data

coming from sensors and we have opted for a micro-

controller. Among the many offerings, we encoun-

tered one, called Arduino (http://arduino.cc), an Ital-

ian open-source device often used for different kind

of applications, easy to use and very cheap.

The microcontroller makes it possible to ﬁnd different

solutions interfacing for example a database service

like Xively

that allows the user to simply add data,

coming for example from a sensor, and store them

into a Cloud, accessible from any Cloud service (i.e.:

web-portal or other). This example is very useful

to understand how to implement a local testbed and

demonstrates the possibility to match all lead tech-

nologies to build our experiment.

1.1 The Cloud Architecture

A typical Cloud has three possible service models: In-

frastructure as a Service, Platform as a Service and

Software as a Service

, and they must be intercon-

nected satisfying both the compatibility with all the

desired services and the performance requirements in

Xively (xively.com), previously known as Cosm and

before as Pachube, is an on-line database service.

For more details we suggest to read Nist’s documenta-

tion (see reference and resources at the end of this paper).

164

Wright P. and Manieri A..

Internet of Things in the Cloud - Theory and Practice.

DOI: 10.5220/0004948201640169

In Proceedings of the 4th International Conference on Cloud Computing and Services Science (CLOSER-2014), pages 164-169

ISBN: 978-989-758-019-2

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

a big data environment.

In particular, the software that will be deployed in or

consumed from the Cloud must be compatible with

the most used operating systems and devices; it must

be user friendly and stable. The Platform tools and

middleware must be engineered to have the best per-

formance when receiving and storing huge amounts

of data. The Infrastructure must also be stable, easy

to maintain, and through with future implementations

in mind. Last, but not least, security is an important

issue for all the three mentioned service models of the

Cloud, even though it will not be thoroughly covered

in this paper.

2 STATE OF THE ART AND THE

ClouT APPROACH

The ﬁrst step approached in the ClouT project has

been to analyse all the solutions available in com-

merce that respond to the project requirements. The

second step was to build a case study based on captur-

ing information from a temperature sensor; through

the Arduino device. In fact it was tested and selected

also because there are many different Cloud providers

compatible with it, allowing to use a web page to store

and read data, and the implementation just requires

modifying few lines of code.

In building the prototype, conﬁrmation was received

that Arduino is indeed friendly and useful for the pur-

poses of the experiment.

Below an analysis of the components and the technol-

ogy that were planned to be used in the ﬁrst phase.

2.1 Arduino

Arduino is an open-source electronics prototyping

platform, available to create interactive objects or en-

vironments, easy to use (hardware and software), as

shown in a simple schema in Fig.1. The microcon-

troller on the board must be programmed using its

own language that allows the developer, for example,

to read a value coming from a sensor in a speciﬁc time

interval.

Arduino projects can be stand-alone or can com-

municate with software on a computer; in this case a

stand-alone project with a dedicated Arduino’s Ether-

net Shield, which allows networking, was opted for

making outgoing connections to a dedicated storage

for the data coming from the sensor; in this way, for

this case study, an Ethernet Shield and the chosen sen-

sor was needed to be connected to the Arduino (See

Fig.2 below).

Figure 1: Arduino (Simple schema).

Figure 2: Sensor-Arduino-Xively.

2.2 Xively

Xively is a public Cloud for the Internet of Things that

permits to connect different private or public devices

like Arduino with dedicated API keys, in order to al-

low the user to put and get the data coming from a

device, and have simple access to it from the website

with a smartphone, a tablet or a computer, as shown

in Fig.3 below.

Figure 3: Xively (an example of visualization of the data

sent).

To carry out the example mentioned before, it is

only needed to: sign in on the Xively website, add

a new project, specify the utilization of the Arduino,

save the received APIKEY and FEEDID keys, and

modify the example code on the ”File - Example -

Ethernet - Pachube Client”

Example of the Arduino program for a simple access

data to the Cloud, created on 15 March 2010, modiﬁed on 9

Apr 2012 by Tom Igoe with input from Usman Haque and

Joe Saavedra, http://arduino.cc/en/Tutorial/PachubeClient;

this code is in the public domain.

InternetofThingsintheCloud-TheoryandPractice

165

2.3 Nimbits

Like Xively there is also Nimbits made by Google;

both can be used to collect data, because they are all

based on the same structure and logic: the data com-

ing from the sensor are all processed, and then sent

to the Cloud’s services through SQL. The data are

collected into a storage, through a web page and af-

terwards they can be visualized on a thin (tablet or

smartphone) or thick (PC) client.

2.4 Raspberry Pi

Similar to Arduino, as well, it is also available an

open-source device named Raspberry Pi, that can con-

nect and use both Xively and Nimbits services. In

order to use it, an Operating System (OS)

must be

installed onto an SD card. In that sense Raspberry

Pi represents a real microcomputer while Arduino is

a microcontroller; this characteristics gives the possi-

bility to use both the devices.

2.5 Raspberry Pi and Arduino

The possibility of connecting the two devices and of

installing the Arduino software tools inside the Rasp-

berry Pi, can solve an important management issue

represented for instance by the remote update of the

distributed Arduino devices. In fact, as more then

one Arduino microcontroller can be connected to one

Raspberry Pi microcomputer, it is possible to mod-

ify its conﬁguration through internet, i.e. without any

cable connection. Others interesting solutions can be

added like, for example, storing the data locally, re-

mote monitoring of the hardware status and perfor-

mance, changing the interval of data acquisition, and

so on.

The Arduino and the Raspberry Pi devices can eas-

ily communicate with each other in different ways:

they can communicate through Ethernet, Bluetooth or

Wireless communication, but it is also possible to the

Raspberry Pi to use an Arduino Shield, or just convert

the I/O pin tension

2.6 Raspberry Pi as a Data Storage

An interesting aspect could be to make the Raspberry

Pi device such as to behave like a server and data stor-

age, installing on it a NoSQL DB, like CouchDB, and

connect it to the Arduino on internet; in this way one

Raspberry Pi can be operated by an optimised Linux

OS chosen among list.

Arduino needs 5V instead of Raspberry Pi 3.3V, the

respective Pins must be connected.

obtains a real ”Internet of Things” communication.

However we have to consider that such conﬁguration

does not provide a real Cloud service, like, for exam-

ple, an Infrastructure as a Service model may. There-

fore we have discarded this hypothesis, because the

main object of the project is ”inﬁnity processing and

storage capacity of data from trillions of things and

people that are integrated via virtual services in the

Cloud”; for this reason, we will employ a Virtual Ma-

chine (VM) running inside a private Cloud. The men-

tioned ability of the Raspberry Pi to store a limited

amount of data could be useful for backup functions

in case of a short interruption of the communication

between the device and the VM.

3 MANAGING DATA STORAGE

IN A NoSQL DB

This chapter described a practical example of how to

make a real Infrastructure as a Service, and how to

use it.

Among the different open-source solutions, it has

been chosen and tested one that appeared to be one the

most interesting: oVirt. It is an open-source Infras-

tructure as a Service solution that allows the client, us-

ing an interface (the oVirt engine, see Fig.4), to man-

age hardware nodes, storage and network resources,

and to deploy and monitor virtual machines running

on the data center.

Given that the main strength of the Cloud approach is

in the optimisation of the hardware resources through

virtualisation, enabling a meaningful cut on total costs

of ownership, we decided to choose a VM, instead of

a simple dedicated machine, as a server. Another re-

ally useful and powerful Cloud oriented virtualisation

solution that we considered was OpenStack, but in our

test case oVirt was ﬁt for the task.

It is possible to manage your own VMs using oVirt,

simply accessing a web interface known as User Por-

tal (see Fig.4).

oVirt provides a high availability mechanism that per-

mits the automatic migration of a VM to another hard-

ware node in case of failure on the current hosting

node. That happens without losing any data or infor-

mation, and no manual action on the VM is needed.

In our test case we employed a VM with CentOS

6.4 as OS and CouchDB installed as database to store

data; in this way it is possible to send a string through

TCP/IP and store data on the NoSQL DB (in this case

CouchDB). For performance reasons it is highly rec-

ommended to store data this way, and not, for exam-

ple, to employ a mySql DB, especially since a ”ter-

abyte torrent” coming from trillions of devices is ex-

CLOSER2014-4thInternationalConferenceonCloudComputingandServicesScience

166

Figure 4: oVirt Engine.

pected. Once CouchDB is installed it is possible to

have an idea on how to read and manage the data ac-

cessing an URL like: http://VM ADDRESS:5984/

utils/index.html, where WM ADDRESS is the IP

ADDRESS of the VM previously made. To provide

a uniﬁed way to access and store data we choose the

CDMI standard, that offers a standard data interface

installed in the Cloud. In our case we prefer to use a

software called CDMI-Proxy.

Whatever the solution chosen, CouchDB or CDMI-

Proxy, it is only needed to modify the sam-

ple code created for the Arduino to communi-

cate with Xively, previously created

, changing the

server name, instead of ”api.pachube.com”, insert

”your.VM.host.name”, and specify port number 5984

if using CouchDB; otherwise, if you use CDMI-

Proxy, insert port 8080 or 2365

. In addition, few

code lines must be modiﬁed in the main sendData

function, because CoachDB follows the JSON format

(see ofﬁcial website guide for the format to use)

At this point all the data, that has been automatically

stored in the chosen DB, must be read. In order to

do this, we have evaluated a well-known and standard

procedure based on a web server package like httpd.

In this case it requires a workload to carry out ”manu-

ally”, therefore there is a risk of making mistakes and

reduces the security. Therefore, we examined other

possibilities, with the aim of ﬁnding a solution that

is still based on an open-source system, but is easier

to implement and maintain, and, lastly, is more sta-

ble and efﬁcient. ZABBIX, seems to respond to the

above mentioned requirements.

See paragraph on Xively

Using one of these two ports (8080/2365), instead of

the standard CouchDB port is useful for security reasons,

but it is not enough to be safe from packet snifﬁng. In this

case using SSL protocol to encrypt data is recommended.

(These conclusions, regarding the infrastructure side of

how to use a NoSQL DB, need additional work on the plat-

form side of the Cloud. Such work will be one of our next

tasks.)

4 MONITORING WITH ZABBIX

All signals, coming from the variety of sensors in-

stalled in many different locations, must be captured,

analysed, stored and represented in a quite under-

standable form and possibly, in case of alarms, in

real time. It would be interesting, and quite useful

for our purposes, if there is the possibility to control

and monitor in real time all the items previously men-

tioned using a software available in the open source

market. Therefore we have tested some solutions and

ﬁnally found one, named ZABBIX, that can carry out

the needed controls. It consists of two different el-

ements in connection: the agent and the server. If,

for example, we want to control all the sensors with

a device like Arduino, and we want to connect more

of these devices to only one microcomputer, that in

our test is a Raspberry Pi, it sufﬁcient to install ZAB-

BIX agent on the Raspberry Pi and ZABBIX server

on a Virtual Machine. The agent sends all the data to

the ZABBIX server so that the data can be not only

stored, but also shown on a map and monitored in a

very easy way. In a test we made, the information

shown was an alarm for an exceptional high tempera-

ture captured by some sensors connected to one of the

devices.

The monitoring options can be added during the ﬁrst

installation, but they can be also modiﬁed later, re-

motely, by an administrator who can also add more

users having each one different powers of access to

the system. This feature allows not only the sim-

ple monitoring of the sensors, but also of the devices

used, the Virtual Machine, the Raspberry Pi and all

the network connected, as it is shown in the enclosed

ﬁg.5.

Figure 5: A screen of a web page interface of ZABBIX

server running on Engineering’s VM.

In Fig.5 there is also an actuator, which will be

studied in the next ClouT project phases.

Another example of the data coming from a sensor

is in Fig.6, where no ﬁlter was applied, and for this

reason there are different range of values in spite of a

InternetofThingsintheCloud-TheoryandPractice

167

small time interval. The continue line is because the

Raspberry Pi and the Arduino were connected, but,

after some time, we manually stopped the code run-

ning on the Raspberry Pi.

Figure 6: An example of a temperature graph of data com-

ing from a sensor.

5 POSSIBLE APPLICATIONS

FOR SMART CITIES

The ClouT project aims at empowering the citizen

clout in smart cities.

What has been designed so far in the course of the

project, and here brieﬂy described, can cover many of

the different features that are now included in what is

called a Smart City. Reference is made to the possi-

bility of collecting data, producing statistics, monitor-

ing services, etc. and, for example, two applications

aimed at improving the citizen’s quality of life. All

this is obtained by capturing data from many sensors

and devices positioned in different places and each

one devoted to a speciﬁc function. One application

drafted regarded an intelligent ﬁre alarm system, in-

volving temperature sensors and web cams automati-

cally switching on, in case of rapidly rising tempera-

tures, and sending images of the scene to a command

and control center or to supervisor, on his smartphone.

Another example of a possible and not expensive

applications using the devices described in this paper,

we can cite for instance the monitoring of people in-

ﬂows and outﬂows from a room (i.e.: for security pur-

poses or marketing), or, the presence of persons close

to a particular area of a hall and the duration of their

presence (i.e.: for marketing purposes; cultural pur-

poses; etc.). Such data, can be easily provided also in

time series format, with statistics and graphs.

The two applications have required an in-depth study

and implementation, to simulate different situations,

to detect all the problems that may occur, and, to make

the application completely automatic.

The interesting feature, in our opinion, is that in

using the suggested hardware, conﬁguration and con-

nections between the different technologies, it means

to create the Internet of Things. The beneﬁts of Cloud

Computing are that the end users need not worry

about the exact location of servers and they can switch

to their application by connecting to the server on

Cloud and start working without any trouble and no

installation is required: just an internet connection

and the right credentials. All the data are stored on a

dedicated database, that permits big data storage, like

the Cloud Platform allows.

Several other applications may be conceived to

empower the citizens in smart cities: to improve mo-

bility, to help social services, to save energy, etc.

6 CONCLUSION AND NEXT

STEPS

As we wrote on the ﬁrst chapter of this document,

the ﬁrst step we approached in the ClouT project was

to analyse all the things previously implemented and

existing in commerce. We started using open-source

technology, hardware and software, because they are

cheap and can be used for different purposes; but this

document wants to be useful also using other similar

technology, because it wants to be as general as pos-

sible. Our idea on using a gateway or a little device to

control all the microcontrollers that collect data from

different kind of sensors, makes it easy to control and

to monitor the system as we can see on ﬁg. 7.

Figure 7: An example on the monitoring of the hardware

used.

All the technologies used are already existing, but

all their functions are not used at all.

The prototype we have implemented, has also given

us the opportunity to understand that the infrastruc-

ture can be applied for the production of applications

in domotics, medical environment, etc. In addition,

the prototype showed us that it can manage a large

multitude of sensors, like those obtained from social

CLOSER2014-4thInternationalConferenceonCloudComputingandServicesScience

168

network. Another possibility that came out during our

experiments is that the Arduino device has demon-

strated its ability to read sensors data in real time,

while the Raspberry Pi device has greater potential-

ities because it hosts an operating system. Such fea-

tures can be used to manage and update the software

in a high number of distributed Arduino devices.

Therefore both devices can provide a useful tool for

future implementations.

In conclusion, the project is still in the ﬁrst phases,

but since now a new and ample number of applica-

tions can be deﬁned and implemented, on the basis of

the described architecture.

ACKNOWLEDGEMENTS

Special thanks to our colleagues Marco Emanuel

Palazzotti and Daniele Pavia for their contribution.

REFERENCES

Arduino, ofﬁcial website, http://www.arduino.cc.

M. Banzi (2011), ”Getting Started with Arduino”,

O’REILLY.

Nist, National Institute of Standards and Technol-

ogy, http://csrc.nist.gov/publications/ nistpubs/800-

145/SP800-145.pdf.

Snia, Cloud Data Management Interface (CDMI), http://

cdmi.sniacloud.com/.

CDMI-Proxy provides CDMI-compliant proxy

server to public cloud backends, https://

github.com/livenson/vcdm.

C. Pﬁster (2011), ”Getting Started with the Internet of

Things”, O’REILLY.

Cosm, ofﬁcial website, http://www.xively.com.

Raspberry Pi, ofﬁcial website, http://www.raspberrypi.org.

Nimbits, ofﬁcial website, http://www.nimbits.com.

oVirt, ofﬁcial website, http://www.oVirt.org.

CouchDB, ofﬁcial website, http://couchdb.apache.org/.

Zabbix, ofﬁcial website, http://www.zabbix.com/.

OpenStack, ofﬁcial website documentation, http://docs.

openstack.org.

IoT-A, Internet of Things Architecture, http://www.iot-

a.eu/public.

C. Doukas (2012), ”Building Internet of Things with the

Arduino”.

K. Ashton (2011), ”That ’Internet of Things’ Thing”. RﬁD

Journal.

ZigBee, as a 6LoWPAN, low power wireless, IEEE

802.15.4, http://www.zigbee.org/Standards/ ZigBee

NetworkDevices/Overview.aspx.

T. Igoe (2011), ”Making Things Talk”. O’REILLY.

InternetofThingsintheCloud-TheoryandPractice

169