Faculty of Science and Technology

Department of Computer Science

Nexans Tracking System

An indoor positioning system using Bluetooth Low Energy

Ragnhild Kosmo Holm

INF-3981 Masters Thesis in Computer Science - June 2015

Abstract
Nexans Rognan is a cable manufacturer located in Nordland County. Each
new cable order have a set of requirements provided by the customer. The
cable specifications change in terms of characteristics of fibre optical elements,
type and quality of cable armouring and dimension and amounts of electrical
conductors. Due to the variety in different cables produced, they also need
a wide array of different cable drums available at the plant. The drums are
stored in sporadic locations around the plant and the lack of organization of
the cable drums creates extra work for production workers. They estimate an
yearly cost of 2-3 FTEs is lost searching for drums. This have created a need
for an indoor tracking system.
The main objective of this thesis is implementing a system which should follow
a cable drum from when it enters the factory for the first time, and until it is
disposed or shipped out to the customer. It uses Bluetooth Low Energy (BLE),
NFC, Android devices and Microsoft Azure for creating fully functional system
for cable drum organization. The system heavily emphasize the use of user
roles to personalize the web page for each user for creating a more seamless
work flow.
The system consist of different modules, performing different tasks. Three
different mobile applications, a web site, a middleware and a SQL database
have been created. Two of the mobile applications provides functionality for
tracking the cable drums inside the factory hall. The NFC application provides
functionality for identifying the drum, and for updating the information about
it. The middleware does most of the work within the system. It decides whether
or not the drum has moved, checks user privileges and implements methods
for adding, updating and deleting from the database. The web site provides
a simple user interface, and creates simple methods for a user to update the
backend with correct information.
The system have been designed and implemented from requirements set by
Nexans Rognan. Multiple smaller, fault prone and time-consuming tasks have
been adopted into the system. Users at Nexans have been included in the
testing, and they have provided feedback which have been implemented in the

ii

ABSTRACT

system. Testing have shown that the system is mature enough to be used as it
is now, and the technology chosen does suit the environment at Nexans.

Acknowledgements
I would like to thank Nexans Rognan for the unique opportunity to design
my own thesis. Your positive attitude have been very motivating and made
the cooperation with you enjoyable. I would especially like to thank Steffen
Aare for being my contact person at Nexans, for showing me around the plant,
helping me with testing and answering my questions.
Next, I would like to thank Randi Karlsen and Anders Andersen for keeping
me on the right track throughout this semester. Your comments and guidance
have been much appreciated.
Furthermore, I would like to thank my fellow classmates for the awesome five
years we have had together. I wish you all the best for the future!
Special thanks goes to my family for all your support and help through the
years. Finally, I would like my thank my boyfriend, Christopher Haugen, for
being supportive, helpful and patient through the work with this thesis.

Contents
Abstract

Acknowledgements

iii

List of Figures

ix

List of Listings

xi

1 Introduction
1.1 Pre-projects . . . . .
1.2 Project statement . .
1.3 Scope and limitations
1.4 Contribution . . . . .
1.5 Structure . . . . . . .

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

1
2
3
3
4
4

2 Context
2.1 Nexans Rognan . . . . . . . . . . . . . . . . . . . . . .
2.1.1 SAP . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.2 Production . . . . . . . . . . . . . . . . . . . . .
2.1.3 Tracking cable drums - current situation at Nexans
2.1.4 Desired situation . . . . . . . . . . . . . . . . . .

.
.
.
.
.

.
.
.
.
.

7
7
8
8
9
12

3 Technical Background
3.1 Internet of Things . . . . . . . . . . .
3.2 Cloud services . . . . . . . . . . . . .
3.2.1 PaaS vs. Saas vs. Iaas . . . . .
3.2.2 Windows Azure . . . . . . . .
3.2.3 Amazon Web Services . . . . .
3.2.4 Google Cloud . . . . . . . . . .
3.2.5 Comparison of cloud providers
3.2.6 The Fog . . . . . . . . . . . . .
3.3 Positioning techniques . . . . . . . . .
3.4 Bluetooth . . . . . . . . . . . . . . . .
3.4.1 Classic Bluetooth . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

15
15
16
17
18
20
21
24
25
26
27
28

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

vi

CONTENTS

3.4.2 Bluetooth Low Energy (BLE)

3.4.3 BLE Beacons . . . . . . . . .
3.5 NFC . . . . . . . . . . . . . . . . . .
3.6 Mobile Platform . . . . . . . . . . .
3.6.1 Comparison . . . . . . . . .
3.7 Summary . . . . . . . . . . . . . . .

.
.
.
.
.
.

4 Related work
4.1 Systems using Bluetooth Low Energy .
4.1.1 BLE within the health sector . .
4.1.2 BLE within sports equipment .
4.2 Tracking systems . . . . . . . . . . . .
4.2.1 Different tracking technologies
4.2.2 Example systems . . . . . . . .
4.3 Summary . . . . . . . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

29
32
35
36
37
38

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

39
39
40
41
41
41
47
48

5 Design
5.1 Architectural Overview . . . . . . . . . .
5.2 Mobile applications . . . . . . . . . . . .
5.2.1 BLE Reference Point Application . .
5.2.2 BLE User Application . . . . . . .
5.2.3 NFC Drum Information Application
5.3 Middleware . . . . . . . . . . . . . . . .
5.4 Database . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

49
49
52
52
54
56
58
65

6 Implementation
6.1 Middleware . . . . . . . . . .
6.1.1 Access Rights . . . . . .
6.1.2 Routing . . . . . . . . .
6.2 Mobile applications . . . . . .
6.2.1 BLE Applications . . . .
6.2.2 NFC Drum Information

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

67
67
71
72
73
73
74

.
.
.
.
.
.
.
.
.
.

75
75
75
77
78
79
80
82
83
83
83

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

7 Evaluation and challenges

7.1 Testing . . . . . . . . . . . . . . . . . . . . . . . .
7.1.1 Locating the cable drums within the factory
7.1.2 BLE beacons . . . . . . . . . . . . . . . . .
7.1.3 Position history . . . . . . . . . . . . . . . .
7.1.4 Machine Learning . . . . . . . . . . . . . .
7.1.5 Battery life of smart devices . . . . . . . . .
7.1.6 User testing . . . . . . . . . . . . . . . . .
7.2 Challenges . . . . . . . . . . . . . . . . . . . . . .
7.2.1 Security . . . . . . . . . . . . . . . . . . .
7.2.2 Extensibility . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

vii

CONTENTS

7.2.3 Cloud computing . . . . . . . . . . . . . . . . . . . .

8 Conclusion and future works
8.1 Future works . . . . . . . . .
8.1.1 Interoperability . . .
8.1.2 Beacon testing . . . .
8.1.3 Mobile applications .
8.1.4 Installation at plant .
8.1.5 Expanding the system
8.1.6 Accuracy of location .
8.2 Conclusion . . . . . . . . . .
8.2.1 Beacons . . . . . . .
8.2.2 Finished product . . .

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

84
87
87
87
88
88
88
89
89
90
90
90

Appendices

93

A Layout of factory hall

95

B Database

97

Bibliography

103

List of Figures
2.1 Spreadsheet showing the status of the cable drums . . . . .
2.2 Three subfigures showing clusters of cable drums . . . . . .
2.3 Three subfigures showing different singles that are used in
testing the solution at Nexans. All three have a height of approximately 2-4 meters . . . . . . . . . . . . . . . . . . . .

11
12

3.1 The different modules available in the Google Cloud . . . . .

3.2 A price comparison of different cloud providers, and different
specifications[19] . . . . . . . . . . . . . . . . . . . . . . .
3.3 A comparison of different Bluetooth classes, with their estimated range and output power[34] . . . . . . . . . . . . . .
3.4 The Bluetooth Low Energy protocol stack[36] . . . . . . . .
3.5 Figure showing the original StickNFind (pink) and the StickNFind Pro (black) compared to a 20 NOK coin . . . . . . . .
3.6 Figure showing two Estimote beacons compared to the StickNFind Pro beacon . . . . . . . . . . . . . . . . . . . . . . .
3.7 A comparison of some of the beacons available, as well as one
beacon (Estimote Sticker) that have not been released yet . .

22

4.1 Some different uses of the BLE technology . . . . . . . . . .

4.2 A comparison of different technologies and their uses[41] . .
4.3 A figure showing how the ZigBee network is built . . . . . .

40
42
44

5.1
5.2
5.3
5.4

50
53
54

5.5
5.6
5.7
5.8
5.9
5.10

The overall design of the tracking system . . . . . . . . . . .

Example of cable drums within the grid . . . . . . . . . . . .
The BLE scanner used for scanning the area for beacons . . .
Screenshot of the BLE User Scanner, showing the beacon MACaddress and measured RSSI . . . . . . . . . . . . . . . . . .
Register an account . . . . . . . . . . . . . . . . . . . . . .
The different access levels in Nexans Tracking System . . . .
A table showing the different user roles in the system . . . .
The administrator startpage . . . . . . . . . . . . . . . . . .
The listing of the different roles in the system . . . . . . . .
User role management . . . . . . . . . . . . . . . . . . . . .
ix

13

24
29
30
33
34
34

56
58
59
59
60
61
61

L I S T O F FI G U R E S

5.11
5.12
5.13
5.14

Information found within the Cable Drums section . . . .

List of cable drums . . . . . . . . . . . . . . . . . . . . .
Position of cable drum . . . . . . . . . . . . . . . . . . .
The logic behind updating the position of the cable drums

.
.
.
.

.
.
.
.

62
63
63
65

6.1
6.2
6.3
6.4

The different controllers created . . . . . . .

The main functionality within each controller
An overall figure of the Views . . . . . . . . .
An overall figure of the Models . . . . . . . .

.
.
.
.

.
.
.
.

68
70
71
71

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

7.1 Figure showing what happens if the range in an area changes 80

7.2 A figure shoring the battery drain on different Android devices 81
A.1 Figure showing the layout within the factory. The figure is
unclear unless it is shown in a format larger than A3 . . . . .

96

List of Listings
6.1
6.2
6.3
6.4
6.5
6.6

Routes within the Cable Drums controller . . . . . . . . . .

Routes within the Devices controller . . . . . . . . . . . . .
Routes within the Localization controller . . . . . . . . . . .
Routes within the Roles controller . . . . . . . . . . . . . . .
Routes within the Account controller . . . . . . . . . . . . .
Route for passing information from "BLE Reference point application" to the middleware . . . . . . . . . . . . . . . . . .
6.7 Route for getting the web site containing information about
the cable drum . . . . . . . . . . . . . . . . . . . . . . . . .

68
68
69
69
69

B.1
B.2
B.3
B.4
B.5

97
98
98
98
99

Database Schema for the table "Users" . . . .

Database Schema for table "User Roles" . . . .
Database Schema for table "CableDrums" . . .
Database Schema for the table "Localizations"
Database Schema for the table "Devices" . . .

xi

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

72
72

1
Introduction
This thesis is done in cooperation with Nexans Rognan. Nexans produces
different types of cables, ranging from fibre cables, copper cables and Remotely
Operated Vehicles (ROV) cables. Their customers are mostly located in the
industry, petroleum, construction and infrastructure markets. Nexans Rognan
mostly have customers within the petroleum industry.
Nexans Rognan produce mostly specially made cables for different customers.
Due to the huge variety in cables produced, they need many different cable
drums available at the plant. The cable drums used vary in terms of material
(steel or wooden), weight limit, material and diameter.
The current situation at Nexans Rognan is that they do not organize the cable
drums in a manner that makes it simple to keep track of the status of the
drums as well as locating drums within the plant. They have a system, SAP,
where orders are placed, and a list of drums at the plant is registered. As the
production have grown, the way Nexans Rognan organize cable drums have not
been amended in line with the other developments. It have become increasingly
more time consuming for the production workers to locate the drums and to
find suitable storage location. The amount of time spent on locating drums are
estimated to be around 2-3 FTEs, and a more efficient system for organizing
cable drums is now needed.
Since GPS was introduced in the 1970s, more and more services that rely
on positions of different objects have emerged. GPS have become a de facto
1

CHAPTER

INTRODUCTION

standard for outdoor positioning system. There are no such standard in indoor
positioning systems. Over the last two decades increasingly more research
have been done on indoor tracking systems, but neither of the technologies
presented have gotten the same results as GPS have for outside tracking.
Indoor tracking systems are implemented using radio waves, acoustic signals,
information from mobile devices or magnetic fields. Many of these are inaccurate or expensive, as we will present in later chapters. Bluetooth Low
Energy (BLE) have made it possible to create an affordable indoor tracking
solution. This is a radio frequency-based (RF) technique using Bluetooth and
signal strength measurements to position the object inside for example a house,
factory or hospital.
By using BLE, the tracking system is simple to implement, the BLE beacons are
reasonably priced, ordinary mobile devices can be used to locate the objects
and the range of the beacons are flexible, from only a few meters up to
approximately 50 meters.

1.1

Pre-projects

This project is done in three parts, one feasibility study, one pre-project and
this thesis.

First study - feasibility study The first study researched different technologies, for example Near Field Communication (NFC), Bluetooth Low Energy
(BLE) and Radio frequency Identification (RFID). It presented the advantages
and disadvantages with each technology, and concluded with BLE as the most
promising technology to use at Nexans Rognan. This feasibility study also
looked at some other, ready-to-use, tracking systems, and researched how
these would fit into the environment at Nexans Rognan. These technologies
and systems are presented as related work in this thesis.

Second study - pre-project The second study started testing on the BLE
and NFC technology. This testing included how the behaviour of BLE beacons
were in the harsh environment at Nexans Rognan and what range that can be
expected. A few different beacon types, from different providers, was tested.
The two most promising beacons were StickNFind (SNF) Pro and Estimote, with
SNF Pro slightly better because of its size and price. The testing done in this
pre-project did show that the BLE technology is mature enough for use in an
indoor tracking system at Nexans Rognan.

1.2

P R O J E C T S TAT E M E N T

Because of limitations with Apple iOS and Windows Phone, as presented in

section 3.6, Android was chosen as the platform to implement the tracking
system application on. This project included an unfinished BLE Reference Point
Scanner. It was not used as a tracking tool in this project, but instead as a tool
for testing different specifications of the beacons, such as range and advertising
interval.

1.2

Project statement

Two pre-projects, presented in section 1.1, are done prior to this thesis, and
this thesis will be built upon the results from these.
The system is created from requirements provided by the customer,
Nexans Rognan. The goal of this thesis is to create a functional
tracking system using BLE beacons, Android devices and Azure as
a cloud provider. The system implemented emphasizes the use of
access rights to customize the web site for each user.

1.3

Scope and limitations

The motivation for creating this system was to give Nexans Rognan, both users
and the company itself, better overview and control over the production at the
plant. This thesis will design and implement a prototype system with the functionality needed to support the goal presented in the project statement.
The existing production system, SAP, is not integrated with this system. Integrating Nexans Tracking System with SAP is extensive work, and is out of scope
of this thesis.
Due to financial restrictions a full scale system have not been implemented
and tested at Nexans. An operational system will include up to a few thousand
beacons. It will also and a large amount of reference points (smart phones) to
create a grid system for plotting the beacons around the plant. A smaller amount
of beacons and smart devices have been bought and tested, and therefore this
system is a downscaled version of the full version. Beacon testing have not
been completed. This is mostly regarding testing of battery life. Each beacon
is expected to last approximately two to five years, and testing have only been
done for one year.

1.4

CHAPTER

INTRODUCTION

Contribution

The successful use of Bluetooth Low Energy beacons shows that it is possible
to implement an indoor tracking system using this technology. Testing of the
beacons shows that, even though they are exposed to much interference from
machines at the plant, they have approximately 20 meters range.
Different modules have been implemented, together creating a fully operational
indoor tracking system for use at Nexans Rognan. The system brings together
many smaller, time-consuming and fault prone tasks at Nexans into one system
for better overview and control over production. By providing methods for
more accurate planning of the production of a cable drum, this system will
contribute to a more efficient work flow at Nexans Rognan.
The middleware implemented, contains most of the functionality in this system
and the majority of design choices are done here. Different functionalities have
been implemented, where an important part is emphasizing user roles. Users
of different roles will have customized access to different parts of the system.
The Administrator in the system have the rights to add users to different role
groups, the purchasing departments is able to add new drums to the database,
while some workers only be able to see information about the cable drums.
Three mobile applications have been implemented, utilizing the Bluetooth Low
Energy and NFC technology.

1.5

Structure

The rest of this thesis is structured as follows:

Chapter 2 presents the context of this thesis. It will present Nexans Rognan,
the challenges and limitations they have regarding tracking of cable drums at
the plant as well as presenting the motivation for creating this system.
Chapter 3 presents relevant technical background information for this thesis. This includes for example a comparison of different cloud providers and
different Bluetooth Low Energy (BLE) beacon providers.
Chapter 4 presents related work to this system. This will include information
about tracking systems using other technologies and other uses of BLE.
Chapter 5 describes the design of the system implemented.

1.5

STRUCTURE

Chapter 6 describes the implementation details of the system.

Chapter 7 presents challenges that have and can occur using this system.
It will also present an evaluation of the system implemented. This includes
results from beacon testing, security issues and details on other design choices
than the ones used in the system.
Chapter 8 will summarize the thesis. Future work will also be presented in
this chapter.

2
Context
This chapter will give an introduction to Nexans Rognan. It presents the
challenges at Nexans that can benefit from the system implemented in this
thesis.
Previous projects are presented in section 1.1, and this chapter is based on the
conclusions from them.

2.1

Nexans Rognan

Nexans is one of the worlds leading cable manufacturers, and they employ
over 25 000 people world wide. Their headquarters are located in Paris, France,
but they have factories all over the world. They mostly produce cables for
infrastructure, construction and petroleum.
Four of the Nexans factories resides in Norway, under the name Nexans Norway.
These factories are located in Halden, Karmy, Namsos and Rognan. They
also have an office location in Helsfyr, Oslo. Together about 1100 people is
employed by Nexans Norway. The factories in Norway is the leading supplier
of telecommunication and power cables in Norway. They also produce large
amount of offshore cables .
1. http://www.nexans.no/eservice/Norway-no_NO/navigate_217577_2371_40_6045/Overview.html

CHAPTER

CONTEXT

This thesis is done in collaboration with Nexans Rognan. The Rognan factory
is located in Saltdal municipality in Nordland County. They mostly produce
specially made cables for the petroleum industry and fibre optic cables. They
also produce ROV (Remotely Operated Vehicles) cables, which is used between
the mother ship and a submarine. What distinguish the Rognan factory from
the other plants in Nexans Norway, is that they mostly produce cables with
special requirements, while the other factories produce standard cables. More
about this is presented in section 2.1.2.

2.1.1 SAP
SAP is the production system used for planning and executing production
of a cable order. The planning includes, among other things, how much and
what kind of raw materials to use, which machines are required, in what order
machines should be used and how many work hours are needed to complete
the order within the given time.
The system also includes information about how many cable drums of each
type is available at the plant. Even though they know how many of each cable
drum type there is, this information is not completely trustworthy. SAP is rarely
completely up to date. Drums can be broken, in use by others, or containing
raw material or products.
SAP is currently frozen, waiting for a new version to be implemented. This,
the limited time span of this thesis, and the dependency on other developers,
makes it impossible to integrate Nexans Tracking System with SAP. When a new
version of SAP is implemented it is possible to add an extension to SAP, and
this extension can include the system created in this thesis.

2.1.2

Production

The Rognan plant has a wide array of customers. When a customer places an
order, they have given a set of specifications that Nexans have to follow. These
specification vary, and Nexans can not produce the cable before an order is
received. Examples of specifications that can change is the characteristics of
fibre optical elements, type and quality of cable armouring and dimension and
amounts of electrical conductors.
When a customer places an order, the specifications have to be reviewed.
Depending on these specifications they will know which machines are needed,
which tests that should be performed and how many work hours are needed.
This planning is done in the existing production system, SAP.

2.1

NEXANS ROGNAN

Due to the production being non-repetitive which cables that are needed vary
from order to order. Cable drums are either made from steel or wood and have
different properties. Wooden drums are cheaper to buy, and can therefore be
disposed. They are often used for cables that have light reinforcement/protection and they are not used in water nor for voltage tests. Wooden drums are
often the preferred drum to ship to customers.
Steel drums can withstand more and are more expensive than wooden drums.
They are not the preferred type to ship out to customers because of the cost.
They are used for voltage tests and are soaked in water. Steel drums, of type
LS11, LT10 and LT20, are used for testing in this thesis.
SAP includes a list of all cable drums at the plant. However, much of the
information needed is not included in SAP. This information includes whether
the drum is in use, if it has material spun on it, if it is being performed
maintenance on or if its broken. There are often halts in production due to the
lack of overview.

2.1.3

Tracking cable drums - current situation at Nexans

The current situation at Nexans is that they lack control over the cable drums.
This includes not knowing where the drum is, if it is in use or is damaged or
broken. Production, as it is done now, have three major challenges that can be
made more efficient by implementing a tracking system for cable drums.
The SAP production system lacks the functionality to add remarks to the
cable drum. The production workers have spreadsheets to take notes in if
they notice some faults on the drum. This spreadsheet is shown in figure
2.1. Other workers use post-it notes to write down any remarks. The
drums in use can be old, since they do not add the production date, and
they can also be crooked or rusty due to lack of maintenance. Nexans
Rognan wishes to enforce regular maintenance every two years, but this
is not done at the current moment.
The second challenge is that they do not always know whether or not
the drum is in use by other parts of production. Those who accept new
orders only sees the amount of drums at the plant, but does not know
the exact number of available drums. Lack of drums available creates a
halt in production, and they might not be able to finish the production
within the agreed time.
The last, and major, challenge is that they lack overview over the locations
of the drums. The plant covers a large area, both inside and outside, and

10

CHAPTER

CONTEXT

drums are mostly placed where there is room. Nexans Rognan have
expanded their production, including new and larger machines, and this
have created an even more shortage of storage for cable drums. They
do not have designated areas for storing drums that are currently not in
use. The drums are placed in clusters around the plant, as can be seen
in figure 2.2a, 2.2b and 2.2c. Larger drums, as the one seen in figure 2.3a,
are placed in different locations where there is available space.
Not knowing the location of the drums at all time creates a challenge for
the workers that should transport the drums from one location to another.
They rely on their own and other peoples memory when searching for
drums that are scheduled to be used. Nexans Rognan have estimated
that approximately 2-3 FTEs of 700 000 NOK each year searching for
drums.

2. Full-time equivalent. A unit that defines approximate cost of one production worker

2.1

NEXANS ROGNAN

Figure 2.1: Spreadsheet showing the status of the cable drums

11

12

CHAPTER

(a)

CONTEXT

(b)

(c)
Figure 2.2: Three subfigures showing clusters of cable drums

2.1.4

Desired situation

Through meetings with Nexans, both production workers and managers, a set
of specifications were formed. The system should include the following features
and information:
It should be easy to update and handle by the users.
Full overview of the drums at Nexans should be available. This should
include the location of the cable drums.
The users should be able to search for a particular drum or drum type.
The last maintenance of each drum should be shown. This will include
when the last maintenance was, who did it, what was done and when the
next maintenance is scheduled. They wish to enforce regular maintenance
every two years.
They also wish to add a link between the existing system SAP, and the
tracking system. Due to the extent of this task, it will not be implemented
in this thesis.

2.1

13

NEXANS ROGNAN

(a)

(b)

(c)
Figure 2.3: Three subfigures showing different singles that are used in testing the
solution at Nexans. All three have a height of approximately 2-4 meters

3
Technical Background
This chapter will present relevant technologies, concepts, frameworks used and
different development platforms.

3.1

Internet of Things

In 2010 there were approximately 12.5 billion devices connected to the Internet,
and Cisco IBSG estimates that in 2015 there will be twice this amount, 25
billion devices [1]. IoTs represents the ability to gather, analyse and distribute
data. This data is turned into information and knowledge. IoT is traced back to
1999 and MIT [2] when they started working within the field of RFID (Radio
Frequency Identification) and sensing technologies. From here the technology
has evolved into a network of networks [1], which collects and analyses data
from many different sources.
By adding more and more devices, the term ubiquitous computing emerged,
where computing appears everywhere, using any device in any format [3]. The
computing unit within the term Internet of Things is called a thing. A thing can
be a heart monitor implant, an auto mobile or a cable drum. The things have
sensors which gives information about the status of the object [4]. In Nexans
case this will be location of the cable drums, and in other cases this can be
battery life or temperature.

15

16

CHAPTER

TECHNICAL BACKGROUND

Sensor energy is one of the more relevant challenges for this thesis. In general,
for IoT to real its full potential, the sensors have to be self-sustaining. By this,
it means that they have to produce energy from light, airflow or vibrations to
keep itself going. As we know, most sensors needs batteries, rechargeable or
not rechargeable, to function, so this technology is still in the making for most
sensors [1].

3.2

Cloud services
Cloud computing is a model for enabling ubiquitous, convenient,
on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and
services) that can be rapidly provisioned and released with minimal
management effort or service provider interaction [5].

Within Cloud computing, instead of storing and accessing data and programs on
a local computer, the data is retrieved from over the Internet. Cloud computing
refers to both the software that deliver the service, and the hardware in the
data centres, which the software runs on. When a cloud is open to the public,
and is a pay-as-you-go service, it is called a public cloud [6]. A private cloud
refers to internal data centres of a business. A private cloud is mostly used by
large businesses that can benefit from cloud computing, and which have the
resources to develop their own cloud.
There are several advantages for companies to use clouds. An infinite amount
of resources are available on demand. Companies can start small and increase
the the amount of resources needed as the business increases its production.
This will decrease the need for large investments before knowing how much
resources they need [6].
One of the most important obstacles of using a cloud provider is the security
aspect. Why should the business trust a cloud provider with their data? There
are both internal and external security threats when using a cloud. The external
are in the hands of the cloud provider, and this includes the physical security.
This also includes external firewall policies [6]. The cloud user is responsible
for application-level security.
One of the more common approaches for cloud storage services is to encrypt
the data at-rest using a 256-bit AES encryption. Google Drive encrypt the data
using AES 128, while Microsoft OneDrive does not encrypt the data [7]. When
1. The cloud providers storage servers

3.2

CLOUD SERVICES

17

discussing in-transit security, many cloud providers uses SSL /TLS to establish
a secure connection between client and server. Recent research have shown
that the system is still vulnerable even though SSL/TLS is used [8].
The cloud providers does provide different levels of resources. If the user is
provided with a lower level of abstraction, the user is responsible for more of the
security challenges. The users does not need to handle these security challenges
themselves, because this can be outsourced to a third party service [6].
A challenge for many businesses is data transfer bottlenecks. If the software
running is data-intensive, the costs of transfers can become very expensive. In
Nexans Tracking system, the data sent between the cloud provider and Nexans
is not data-intensive, and only smaller amount of text is sent and received from
applications.
By renting resources from a cloud provider it is possible to cut cost. This
includes the need of hardware, software electricity and bandwidth. There is
also limited need for in-house IT personnel. By putting data in the cloud the
work regarding security, maintenance, backup and support is simplified for the
user. By using a cloud it is simple to scale up a system when in need of more
resources. With using a local server, this process would have been more time
consuming since hardware would have to be ordered and set up, and software
installed. This is handled by the cloud storage provider, which in this thesis is
Windows Azure.

3.2.1 PaaS vs. Saas vs. Iaas

When implementing a new system or adding functionality to an already existing
system the businesses have to analyse the requirements they have for. These
requirements can regard how much resources, in manpower and money, they
have available. It can also regard what security level they wish to pursue, and
how much control they want over the applications and data they store. By
having everything locally, both hardware and software, they businesses have
to think ahead and see if this is something that they will have to expand. If this
is the case, they might decide to add more hardware from the start. By using a
cloud provider they can start smaller, and expand the system when their needs
increase.
In this thesis the choice was between establishing a server locally at Nexans or
using a cloud provider. Nexans does currently have a local instance of a SAP
2. Secure Socket Layer
3. Transport Layer Security

18

CHAPTER

TECHNICAL BACKGROUND

system, however this system is currently frozen and will not be expanded in
the near future. This thesis does not include an integration with SAP. However,
if this system were to be implemented in SAP, the system could not have been
put in Azure, and would have to be run locally. The reason why choosing a
cloud provider in this theses, is because Nexans does not need to invest in new
hardware or software to run the prototype system created, it will also simplify
future maintenance. This section will present what services that are available
for Nexans if they wish to proceed with keeping the system in the cloud.
Software as a Service, or SaaS, represents the largest cloud marked. It uses the
web to deliver services through an interface that is accessed on the clients side.
These services are provided by a third-party [9]. Some of the characteristics of
a SaaS service, is that many of these services can be accessed through the web
browser, often just needing some additional plugins to function. For a business
using SaaS they can outsource the managing of applications, middleware,
storage and servers [9]. The user does not need to think about hosting, updating,
managing or developing applications or storing data [10].
Platform as a Service, or PaaS, is what Microsoft is betting on when developing
Azure as a long time project [11]. Google also presents their cloud solution as
a fully-managed PaaS [12]. One of the main reasons to use PaaS is to increase
the developers productivity. It is a framework where developers can build their
own applications, while also making it more cost-efficient and simpler to publish
applications [9]. In PaaS the system provider makes most of the decisions when
determining which operating system to use, which programming language is
supported and the API used [10].
Amazon Web Services (AWS), Azure and Google Compute Engine can also go
under the last service, the Infrastructure as a Service, or IaaS. These providers
provide access to raw computing infrastructure and the middleware software [13]. This service provides self-service models for accessing and managing
remote data centres [9]. The resources can be storage, computation power
or networking. The advantage with IaaS as opposed to having the hardware
locally is that they can pay for what they use (and not for what they have),
limited start-up cost and less risk of data loss because of globally replicated
data. They do not have to have IT staff to manage the hardware.

3.2.2 Windows Azure

Windows Azure Storage (WAS) is a cloud storage system that
provides customers with the ability to store seemingly limitless
amounts of data for any duration of time. WAS customers have
access to their data from anywhere at any time and only pay for

3.2

CLOUD SERVICES

19

what they use [14].

WAS is a cloud storage that store the data in Blobs, Tables and Queues. The
Queues are the message delivery, Tables are structured storage and Blobs are
the user files. The common usage pattern is that the data enters and leaves the
system as Blobs, Queues processes the Blobs and the service state and results
are stored in Tables or Blobs [14].
Before implementing Azure they received feedback from many potential customers, and this resulted in a set of specifications.
Strong consistency - Many of the customers wanted strong consistency.
Many enterprise customers are moving their data to the cloud, and
they want to be able to perform conditional reads, writes an deletes for
optimistic concurrency on the strong consistent data.
Global and scalable storage - WAS implements a global namespace that
allows data to be stored and retrieved consistently from any location in
the world.
Disaster recovery - Locations are always in risk of disasters, like earthquakes, floods or hurricanes. Therefore WAS stores the data across multiple data centres around the world. In case one datacenter is faulty, the
other data centres will be available.
Cost of storage - To lower the cost of storing data in WAS, multiple
customers share the same storage infrastructure.
To provide a single global namespace they leverage on DNS as a part of the
storage namespace. The namespace is divided into three parts, the AccountName, the PartitionName and an ObjectName. The AccountName is a customer
selected account name. In this thesis it is nexanstrackingsystem. This name
is used to locate the primary storage cluster and data center. It is possible
for an application to use multiple AccountNames to store the data in different
locations. The PartitionName locates the data once the request have reached
the storage center. The PartitionName holds many objects, and the ObjectName
identifies the different objects inside the partition [14].
WAS provides string consistency, global partitioned namespace and disaster
recovery. Multiple users will run on the same hardware, and this reduces the
costs per user.
Windows Azure started up as a PaaS in the form as Microsoft Windows Azure,
but later have evolved into also supporting IaaS features. This also included sup-

20

CHAPTER

TECHNICAL BACKGROUND

port for Linux servers, and not just Windows operating systems. For Windowscentric developments, Azure offers better options for this. Visual Studio, .NET
and Windows programming languages are fully supported [15], and users can
create and deploy applications straight from Visual Studio. They also provide
SDKs for many other languages, like PHP, Java and Python.

3.2.3 Amazon Web Services

Amazon Web Services (AWS) provides on-demand delivery of IT
resources via the Internet with pay-as-you-go pricing [16].
Amazon is one of the most long-living companies using a decentralized IT
infrastructure. In 2005 they have spent a decade building the IT infrastructure
that supports one of the worlds largest online retain stores. From their retail
platform, AWS emerged in 2006 . AWS was developed so that other organizations could rent and utilize their cloud resources. Data is written and read using
for example SOAP and REST, and is available through web browsers.
AWS consists of a variety of web services, from which the user can chose what or
which resources that suits the organization best. Below is a short presentation
of the two different resources most relevant for this thesis.
Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides compute capacity in the cloud. The amount of resources for the
user is not fixed, and can be scaled up or down as the need for resources
change .
Amazon Simple Storage Service (Amazon S3) is a simple web service that
can be used to store and retrieve data in the cloud .
In addition to these two, there are many services that different users can
utilize, like building their own online store upon Amazons infrastructure or
retrieving historical pricing for different items of interest [16]. Other examples
are SimpleDB and Simple Queue Service. The SimpleDB service provides a
database service that allows users to store structured data. This service provides
the facilities of a relational database as well as a web interface for this system.
The Simple Queue Service is a reliable queueing service between instances in
EC2 [17].
4. http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=irolnewsArticle&ID=1216597&highlight=
5. http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=irol-corporateTimeline
6. http://aws.amazon.com/ec2/
7. http://aws.amazon.com/s3/

3.2

CLOUD SERVICES

21

AWS is a flexible platform, making it possible for the organization to decide

what programming platform or model they wish to use to solve their problem.
The user will also chose what resources are needed, from compute power to
storage. When a user puts their data in the Amazon EC2, they pay relatively
low prices for storage. When they run the application, they pay for the amount
of compute nodes they need, and is done at an hourly rate. While having access
to the rented nodes, the user will have root access to the system. This control
implies that the user can configure the node as the user desires and can then
chose what system software is needed [17].

3.2.4 Google Cloud

This section will present some of the details from Google Cloud that is relevant
for Nexans when deciding which cloud provider to use, if they decide to have
the data in the cloud.
Google Cloud Platform is a set of modular cloud-based services
that allow you to create anything from simple websites to complex
applications
Google Cloud Service consist of modules, as seen in figure 3.1. The platform is
based on Python and it provides the user with the possibilities of storing data,
hosting applications and networking [18]. In addition to Python, they support
other programming languages like Java, PHP and Go. Google Cloud Service
consists of three parts within the compute section of the cloud, the Compute
Engine, the App Engine and the Container Engine.
App Engine is Googles PaaS, and makes it simple for users to develop
and publish applications . It is possible for users to try Google App Engine
for free. There are some restrictions to this free trial, for example it does
only include 1 GB of data storage .

8. https://cloud.google.com/
9. https://cloud.google.com/products/
10. https://cloud.google.com/appengine/docs/whatisgoogleappengine

22

CHAPTER

TECHNICAL BACKGROUND

Figure 3.1: The different modules available in the Google Cloud

The App Engine offers three distinct sets of features for the user. First is
PaaS, which gives the user the opportunity to create applications that
support transactions and scalable and available applications. Second is
that they provide SaaS, which is finished applications that can be used by
the user straight away. Thirdly they provide the user with the availability
to integrate systems and applications from other service platforms, into
the App Engine [18]. With Google App Engine it is easy to deploy new
applications. They support persistent storage, scaling and load balancing
and integration with other Google Cloud Services and APIs.
The Compute Engine is their IaaS. They run large-scale workloads on
virtual machines. These machines are hosted on Googles infrastructure,
and the user can chose what virtual machine, from micro-VMs to large
VMs, that would fit their needs . By using Googles fibre network the
users can gain the performance and consistency required. Depending on
the users need, Googles Compute Engine supports global load-balancing,
strong and consistent bandwidth across machines, encryption of data,
scalable and reliable machines and their VMs support several types of
distros [?]. As the other two cloud providers presented in this thesis, the
user will pay for what he uses and nothing more.
11. https://cloud.google.com/compute/

3.2

CLOUD SERVICES

23

Within the module Storage, Google also provides three more sections, the Cloud
Datastore, the Cloud SQL and the Cloud Storage .
Cloud Datastore is a schemaless database for storing non-relational
data . It automatically handles sharding and replications, and this creates an highly available and consistent database. It also provides ACID
transactions, and does this using optimistic concurrency control. By using this module the user does not need to worry about the underlying
structure. It allows the user to search through the data and sort the data
as needed .
Cloud SQL provides a relational MySQL database. The data stored in the
database is replicated at many locations, which means the data is stored
durable and it is both available and secure. Google also handles database
management and patch management to ensure the performance of the
database .
Cloud Storage is a durable and highly available object storage service .
Within this section Google provides different services, where two of them
are presented below.
The Standard Storage provides highest level of performance, durability and availability. Data that is stored here needs low latency
access. It is enabled at the bucket level . Each bucket contains objects that can be accessed by their own methods. Standard Storage
is also the default storage, so that if the user does not chose any
other storage option, Standard Storage is chosen.
The Durable Reduced Availability (DRA) Storage stores data at lower
cost. This lower cost comes with the disadvantage of less available
data. DRA is also enabled at the bucket level .
In addition to this Google provides a service so that the user can gain insight
on his application, and how it performs and the availability. They also have
12. https://cloud.google.com/products/
13. https://cloud.google.com/products/
14. Database partitioning that is used to separate a large database into several smaller and
faster shards
15. Atomic, consistent, isolated and durable
16. https://cloud.google.com/datastore/
17. https://cloud.google.com/sql/
18. https://cloud.google.com/products/
19. https://cloud.google.com/products/
20. https://cloud.google.com/storage/docs/standard-storage
21. https://cloud.google.com/storage/docs/durable-reduced-availability

24

CHAPTER

TECHNICAL BACKGROUND

algorithms to analyse data to predict the behaviour of an application.

3.2.5 Comparison of cloud providers

These three cloud providers does offer much of the same features and services,
but this section will provide some of the differences between them. For this
thesis which cloud provider chosen would not have a significant impact on the
results. Because of platforms used and knowledge of the Windows platform at
Nexans Rognan, Azure was chosen. In addition, Nexans runs Windows servers,
and it would be simple to port the data to one of their servers if needed. This
choice is not a fixed choice, and Nexans can easily change cloud provider, or
decide to have the data locally later on.
A normal PaaS service have included databases, message queues and caching,
but Azure, Google and Amazon have also included services belonging to the
IaaS core services like virtual machines. Therefore it is difficult to place each
provider into only one of the categories.
Google File System (GFS) is the system closest to Azure. One of the major
differences between the systems is that GFS allows relaxed consistency across
replicas and does not guarantee that the two replicas are bitwise the same.
WAS does provide the guarantee that two (or more) replicas are bitwise the
same [14].

Figure 3.2: A price comparison of different cloud providers, and different

specifications[19]

For many businesses it is important to keep some of the data close to home,

3.2

CLOUD SERVICES

25

and not in the cloud. For achieving this, the businesses is in need of a hybrid
solution [20]. Microsoft have tried to build their cloud solution around this
hybrid concept, and hopes to attract businesses that requires this. Amazon does
not, at the current moment, support a hybrid solution, and most of the Amazon
Web Services (AWS) resides in Amazons public cloud [20].
Figure 3.2 does show an example of prices from the three different cloud
providers. This is a very general comparison, but it is too many variables to
consider, so a conclusion cannot be made deciding which is pricier than the
other. When a solution becomes complex, with many different services running,
it will be difficult to estimate the total cost of the solution. All three providers
does have a pricing calculator, so that the user can compare them to each other,
and the result will depend on which services the user selects .
When looking at how the user should administrate the data and code in the
cloud, there are some major differences. Which of the solutions that are best,
will depend on the users needs. Azure does provide an online management
and portal service, where the user easily can organize and manage the code and
database. This solution will be the best choice for many people, but some would
feel some lack of control over the operations. Google does not have the same,
simple administration features. One example is running Hadoop [21] on the
Google Compute engine requires the user to download a Hadoop package, along
with some other packages, to deploy the cluster. Amazon does provide some
assistance with deploying applications on the AWS platform, but in addition
to this the user can be certified within different parts of AWS. For example
they provide tutorials and exams to be a certified SysOps Administrator or
Developer .

3.2.6 The Fog

Fog computing resembles cloud computing in many ways. It provides storage,
data and computational power. The services it provides is located between
the end dives, for example smart phones and tablets, and the traditional cloud
computing data centres. What distinguishes it from cloud computing is the
distance from the users. Data can be processed mostly in locally, in smart
devices, instead of the computations being done in the cloud[22]. Because of
this we can say that a fog is a cloud close to the ground [23]. It is placed at
the network edge, close to the user, and it has a dense geo-distribution. This
22. http://calculator.s3.amazonaws.com/index.html
23. https://cloud.google.com/products/calculator/
24. https://azure.microsoft.com/en-us/pricing/calculator/
25. http://aws.amazon.com/certification/

26

CHAPTER

TECHNICAL BACKGROUND

reduces the latency for the user and improves QoS [23][24].

3.3

Positioning techniques

This section will present some techniques that are most commonly used in
indoor positioning systems. In radio-frequency (RF) based systems, there are
six categories of approaches that can be used, and they are Time of Flight (ToF),
triangulation, trilateration, fingerprinting, cell-based and filter-based.
Triangulation uses angles to determine the position of an object. The
position is based on the Angle of Arrival (AoA) of wireless signals. To
estimate the position of an object atleast three reference points are
needed, and the angles between these points are known [25]. From
mathematical equations the position of the object (cable drum) can be
estimated.
Trilateration calculates the position by measuring the distance from
atleast three reference points [25]. These reference points have a fixed
distance between them. The distance between the reference points and
the cable drums are estimated based on the wireless signal parameters.
When using a fingerprint-method the area is divided into a grid. It consists
of two phases, a learning (offline) phase, and an online phase. In the
offline phase the fingerprints are created. These prints are unique for
each node in the grid, and often consists of RSSI measurements. During
the online phase, the moving objects (cable drums) receive the RSSI
measurements from the reference points. From the database the moving
objects is able to retrieve the fingerprints and these are compared to the
new RSSI measurements. The moving objects positions is registered after
analysing the difference in RSSI measurements [26]. One advantage
with using fingerprinting is the accuracy that can be achieved, but the
disadvantage is that the offline phase is very time-consuming and difficult
to maintain.
The cell-based positioning method is quite similar to the fingerprint
method. It has an offline/learning phase where (in this case) the smart
phones are placed at strategic locations at the factory, and they are registered in a database. In the online phase these smart devices read any
signals from the surrounding beacons and stores the MAC-address of the
26. Quality of Service, or QoS, is the overall quality of service seen by the users of the computer
network

3.4

BLUETOOTH

27

beacons. The set of stored beacon addresses are compared to previous

entries, and the best match is selected.
Within the filter technology there are two filters that are commonly used,
the Kalman filter [27] and Monte Carlo filter (particle filters) [28]. The
Monte Carlo filter uses an algorithm to estimate the position of an object
as it moves. By having a map over the environment and the algorithm it
uses a particle filter to represent the most likely positions, where each
particle is one possible position [29]. This method is often used in robots
when they move around inside an area. For the robot to move around a
space it needs to know the surroundings and its orientation [30].
Time of Flight (ToF) works by estimating the distance between two objects
by counting the time it takes for a signal to go from one object to the
other. This method requires the objects clocks to be synchronized with
high precision. Because a small difference between the two clocks means
a significant distance between the two objects, this method is often used
when the distance between objects are hundred meters or more .
Research have mostly focussed on the use of RSSI measurements, triangulation
and triateration. RSSI measures are used in the BLE User Scanner where it
represents a number which tells the user if he is getting closer or further away
from the beacon. This thesis also uses the cell-based method to create a grid
system within the plant and to plot the cable drums within this grid.

3.4

Bluetooth
Bluetooth technology is a wireless communications technology that
is simple, secure and everywhere .

The first section will give an intro to the Classic Bluetooth technology, while
the second section will present Bluetooth Low Energy (BLE) and its specifications

27. http://www.rp-photonics.com/time_of_flight_measurements.html
28. http://www.bluetooth.com/Pages/Basics.aspx

28

CHAPTER

TECHNICAL BACKGROUND

3.4.1 Classic Bluetooth

Bluetooth was created by Ericsson in 1994. The technology operates on frequencies between 2400 and 2485 MHz. It uses FHSS, which have a hopping carrier
that will prevent interference between devices. The Bluetooth transmission
will only remain on one channel for a certain amount of time before jumping
to another channel. If there is interference between two or more devices, the
message from one device will retransmit later when the device have switched
to another channel [31]. Bluetooth, FM radio and television all send data over
wireless. One of the specifications that distinguish Bluetooth from other technologies that uses radio waves, is that the range of Bluetooth is only within a
devices Personal Area Network (PAN).
Millions of COTS devices support Bluetooth, ranging from computers, smart
phones, medical equipment and television controls. It was conceived with the
intention of removing the use for cables between devices .

Range and positioning

According to Bluetooth Special Interest Group (BSIG), Bluetooth is not about
getting the exact location, but more about proximity[32]. And therefore the
more correct term would to call it a proximity solution, and not a positioning
solution.
Bluetooth was designed to be a short range technology, with a range of only
a few meters. However, todays Bluetooth have achieved a range of up to 100
meters [33]. Depending on the range the different devices has, they are divided
into three classes, as shown in figure 3.3. This advertised range is done where
the signals from devices are not hindered by other objects. There are many
different factors that can give the devices a much lower effective range. This
can be obstacles, like cable drums, machines and walls, interference from other
devices and machines, and signal reflection.

29. Frequency Hopping Spread Spectrum

30. Commercial Off-The-Shelf
31. http://www.bluetooth.com/Pages/Fast-Facts.aspx

3.4

BLUETOOTH

29

Figure 3.3: A comparison of different Bluetooth classes, with their estimated range
and output power[34]

Discovery and pairing of devices

For two devices to be able to send and receive data between each other a bond,
also called pairing, have to be set up first. There are two different roles that
the devices can take, and those are master and slave. Each master can have 7
simultaneous connections to different slaves. To initiate this bond between the
master and its slaves the master will emit a inquiry message. The slaves within
proximity, and which are discoverable, will respond to the inquiry process with
its name, and a connection can be established [34][35].

3.4.2 Bluetooth Low Energy (BLE)

Bluetooth Low Energy (BLE) was developed by Bluetooth Special Interest Group,
and was presented in 2011 as Bluetooth 4.0. The intention of BLE was short
range communication with low power consumption. It built upon the Classic
Bluetooth, but includes some new features, which is presented below .
Figure 3.4a shows the protocol stack of BLE. The figure 3.4b shows the structure
of the different fields of each layer. It consists of two main parts, the controller
and the host. Within the controller lies the link layer and the physical layer.
The host includes upper layer functionality and and it runs on an application
processor [36].

Link layer The link layer of BLE is a state machine. The five states of this
state machine is advertising, scanning, initiating, passive and connective. The
connective state includes the roles master and slave, and the scanning part
includes the active and passive substates [36].
Physical layer BLE works in the 2.4GHz ISM band, it has 40 channels and
three of these channels are advertising channels. The remaining channels
32. http://www.bluetooth.com/Pages/low-energy-tech-info.aspx

30

CHAPTER

TECHNICAL BACKGROUND

are data channels. The packets between the channels are either advertising
or connection events. The advertising channels are used to discover other
devices, establish a connection between devices and broadcast transmission.
The data channels are used for bidirectional communication between paired
devices [36].

L2CAP, ATT and GATT The L2CAP, or logical link control application protocol, is a protocol based upon the classic Bluetooth L2CAP. Its main goal is to
multiplex data of the three higher layer protocols [36].
The ATT is what defines the communication between the server and client.
GATT defines the framework that uses ATT to discover other devices. GATT
also controls the exchange of device characteristics from one device to another [36].

Figure 3.4: The Bluetooth Low Energy protocol stack[36]

Even though the controller inherits from the classic Bluetooth Controller, they
are incompatible. This means that an application which only implements classic
Bluetooth support, cannot communicate with a device that only supports BLE,
and vice versa. A device that supports both the classic controller and BLE

3.4

BLUETOOTH

31

controller is called dual-mode devices [36][37].

BLE Modes
BLE has four modes: technology, central, peripheral, broadcaster and observer.
The observer mode compliments the broadcaster mode, and the central role
is similar to the master role in Classic Bluetooth. The central role is more
advanced and designed for devices like smart phones, and it can manage more
connections from multiple peripheral devices [36]. Previously, two devices had
to be paired to exchange information, in the new mode the devices can be set to
broadcasting mode, and data can be sent through the advertisement channels.
For this thesis the broadcasting mode of the BLE beacons is used. The smart
phones used as reference points will pick up the signal and data (MAC-address)
from the different devices and this is added in the database.

Range BLE Beacons

The range of BLE devices are, as Classic Bluetooth, dependent on the broadcasting power, obstacles and interference from other devices and machines.
Most BLE beacons operate with a 0 dBm transmitting power, and this have
an expected range of 50 meters. Different beacon providers advertises with
different range and expected battery consumption. These numbers are showed
in figure 3.7, and we can see that most providers expect a range of about 50
meters.

BLE security
Communication over BLE is secure and protected, and the Bluetooth Core
Specification (BCS) does provide multiple features to cover the encryption,
privacy of users data and data integrity .
To ensure secure communication between two Bluetooth devices, pairing is
used. When two devices pair up, their identity set up the encryption keys
and the devices are ready to exchange information. They key generation is
performed by the host, and Bluetooth uses AES-CCM cryptography. In BCS
version 4.2 security got enhanced by using the Elliptical Curve Hellman-Diffie
(ECDH) algorithm for key exchange between devices. BLE also includes a
privacy feature, which reduces the ability to track a single BLE devices over
33. https://developer.bluetooth.org/TechnologyOverview/Pages/LE-Security.aspx
34. https://asecuritysite.com/encryption/ccmaes

32

CHAPTER

TECHNICAL BACKGROUND

longer period of time. This feature will hinder outsiders to track one particular
cable drum, and possible steal it when it is located in a less protected area.

3.4.3 BLE Beacons

From testing and research done in the pre-projects leading up to this theses,
two beacon providers stood out as the most promising. Those are the Estimote
and StickNFind Pro beacons.

StickNFind Beacons
StickNFind (SNF) have several different types of beacons for sale, as well as
some that have not been released yet. In this project, which includes the two
pre-projects and this thesis, two beacons from SNF have been tested. These
are the SNF Consumer edition and the SNF Pro . StickNFind have also
included a SDK to help developers in creating new applications.
These two beacon types differ somewhat in specifications, and they both have
pros and cons.

StickNFind Consumer Edition The biggest advantage with the SNF Consumer edition is that the battery can easily be changed while stuck on a cable
drum. When the beacon runs out of battery Nexans does not need to invest
in a new beacon and configure the beacon, they only need to change the
battery.
The disadvantages does however prevent this beacon from being used at the
plant. Firstly, the expected battery life is not sufficient. The expected battery
life is approximately one year, with a daily use of 30 minutes. With up to 2000
beacons at Nexans, this short battery life would create significant maintenance
on the beacons.
Another disadvantage is the lack of waterproof case. Many of the drums are
soaked in water for testing, and it would not be possible to remove the beacons
before every soak to prevent it from being ruined.

StickNFind Pro The StickNFind pro have an expected battery life of approximately three years. Since Nexans wishes to have regular maintenance on the
35. https://www.sticknfind.com/sticknfind.aspx
36. https://www.sticknfind.com/enterprise.aspx
37. Software Developing Kit

3.4

BLUETOOTH

33

drums every two years, it would be possible to include beacon management

as a part of the regular management.
This beacon, as opposed to SNF Consumer Edition, cannot be changed battery
on without compromising the seal that keeps it waterproof. This means that
they have to change the whole beacon every two-three years.

Figure 3.5: Figure showing the original StickNFind (pink) and the StickNFind Pro
(black) compared to a 20 NOK coin

Estimote Beacons
Estimote has the second beacon type tested in this thesis. They provide
waterproof beacons with an average battery life of three years. Estimote have
also created an SDK to help developers create applications. The beacons are
described as a mote, and as figure 3.6 shows, it is a large beacon compared to
the SNF beacons.

38. http://estimote.com/

34

CHAPTER

TECHNICAL BACKGROUND

Figure 3.6: Figure showing two Estimote beacons compared to the StickNFind Pro
beacon

Comparison of BLE beacons

In addition to the two beacons presented here, the pre-projects also tested two
other beacon types. This was the Easibeacons and RedBear beacons. These two
did not meet the waterproof requirement and were not as robust as SNF Pro and
Estimote. Therefore they have not been installed and tested at Nexans.
Both Estimote and SNF Pro provide waterproof beacons with long battery life.
The main difference in the beacons are the size. The SNF Pro beacons are
considerably smaller than the Estimote beacons. This feature will prevent the
SNF Pro beacons from being as easily destroyed by trucks and other machines.
Both Estimote and SNF Pro is installed at Nexans, and both are still working
now, one year after they installation. A figure showing the differences between
the beacons are presented in figure 3.7.

Figure 3.7: A comparison of some of the beacons available, as well as one beacon
(Estimote Sticker) that have not been released yet

3.5

3.5

NFC

35

NFC

Near Field Communication (NFC) is a technology based on RFID, and is used for
contactless communication between two devices. The motivation for creating
NFC was to make it easier to transfer content, connect devices and make
transactions. As opposed to other existing communication technologies, the
distance between the devices can not exceed 10 centimetres [38].
Most devices support NFC today. Apple does have a restriction on their NFC
solution, but this is explained more in section 3.6.
NFC involves a target and an initiator. The initiator generates the RF signal,
initiates contact and controls the data exchanged. The target will answer this
request. It also differs between two communication modes, active and passive,
where active is where both the target and initiator creates their own electrical
fields [39].
NFC operates in three modes; Card Emulation, Peer-to-Peer and Reader/Writer.
In Reader/Writer mode, the NFC device can read and write data to a tag.
This mode is used in the NFC application implemented in this thesis. In Card
Emulation, the phone acts as a physical object. Examples are house keys and
bus cards. In the last mode, Peer-to-Peer, data can be transferred between two
NFC devices [38].
When trying to create a technology that stores personal data, like payment
card information, how secure the technology is often the most discussed and
researched topic. Limiting the range can reduce the security challenge as
the complexity of eavesdropping increases. Since NFC per default is backwards compatible with RFID, the NFC is not encrypted. Encryption may be
implemented as an extra feature for certain applications. An example is the
AES encryption, which is implemented in many applications to prevent eavesdropping, man-in-the-middle attacks and data manipulation, corruption and
insertion.
The NFC application have to be stored on either the SIM card, the smart card
or on the smart phone itself. By storing the application only on the SIM card,
this will be the secure element for the authentication protocol. This will also
create simpler portability between devices [39].
An example where NFC is used is a payment solution created in collaboration

39. Where the two devices believes that they are communication directly with the other, but
in reality a third-party device is listening on the data sent and received

36

CHAPTER

TECHNICAL BACKGROUND

with Telenor and DNB . The payment terminals contains support for NFC,
and the users smart phone act as a VISA card. The smart phone and SIM card
has to support NFC, the user needs a VISA mobile card and the smart phones
needs an application, Valyou, installed.
In this thesis NFC is used to identify the cable drum. This is to discard the
piece of paper stuck on the cable drum which gives the user an indication of
the status of the drum. Due to most of the cable drums tested is made of steel,
special NFC tags have been used. These are designed to be placed on metal
objects.

3.6

Mobile Platform

There are three platforms mobile that are able to connect to BLE beacons, and
these are Android, iOS and Windows Phone. Previous works have tested two
of these platforms, Windows Phone and Android, and Android is the preferred
platform. This section will present a summary of the tests and research results
from the pre-projects.

Apple iOS Apple iOS was, alongside Android, early with supporting Bluetooth 4.0. In addition to supporting Bluetooth 4.0 they created a framework
which they call iBeacon. The iBeacon framework extends the location services
in iOS. Apple devices newer than the iPhone 4s, iPad 3. Generation, iPad mini
and iPod Touch 5. Generation supports Bluetooth 4.0. By using an iBeacon, the
iOS unit can notify the user when he leaves or enters an area. In addition to
this, it can estimate the distance from the user and an location, for example a
particular shelf in a store, and this is estimated using RSSI measurements .
Not all beacon providers are approved as iBeacon providers. Apple have a set
of requirements for the providers, which they have to pass, to become a licensee. These requirements will provide the user with a consistent and reliable
experience. Both StickNFind and Estimote are iBeacon approved.
Previously Apple devices did not support NFC (Near Field Communication), but
the new iPhones (iPhone 6 and iPhone 6 Plus) that was released in September
2014, does support this technology. In addition to NFC support, they also
introduced Apple Pay, which is a contact-less payment solution. NFC on the
new iPhones are currently on lockdown. This means that they cannot be used
for anything more than Apple pay. In this thesis it would not have been possible
to read the NFC tags placed on the cable drums because of the lockdown.
40. http://www.online.no/trender/nfc-betal-med-mobilen-i-butikken.jsp
41. https://support.apple.com/no-no/HT202880

3.6

M O B I L E P L AT F O R M

37

Windows Phone Microsoft have not been as early as Apple iOS and Android
in supporting BLE, but this support was implemented in Windows Phone 8.1.
In previous works a proof of concept using Windows Phone and BLE researched
to see if an implementation on this platform was possible. Research showed
that there were some challenges with using Windows Phone. An application
running on Windows Phone is only able to see and list beacons that have
previously been paired with the device. As Nexans will have more than one
thousand beacons at the plant, and due to the pre-pairing requirements, an
application running on Windows Phone is not a viable solution.

Android Android was early in supporting BLE, and it was a built-in platform
support for Bluetooth 4.0 in Android 4.3. It provides APIs that can be used by
applications to discover devices and query for services.

3.6.1 Comparison
Apple and Android have both been early in supporting Bluetooth 4.0 on their
devices. Due to restrictions on Nexans, Apple iOS is not an option in this thesis.
For systems that does not have this restriction iOS might be just as good a
choice as Android.
As presented, Windows Phone have the disadvantage of the phone having to
pair with each beacon to be able to see them in an application. This excludes
this platform from using it at Nexans.
Since neither iOS nor Windows Phone is possible, further research on which
platform to use have not been done. Further requirements from Nexans is that
the smart device should, if possible, have a certification of IP67. There are two
certificates that are common for robust smart phones, and they are IP54 and
IP67. The first number indicate how dust proof the phone is, and the second
number is how waterproof it is. The higher number, the more sustainable the
phone is. A phone with a certification of IP54 does not guarantee that it is dust
proof, and it only supports some splatter of water, for example rain. A phone
with the IP67 certificate is dust proof, and it should not take damage of falling
into freshwater, and it can be left in water for approximately 30 minutes. Three
examples of smart phones that does have the IP67 certificate is the Huawei
Ascend Y530, Galaxy Cover 3 and Motorola MOTO G. For this thesis the Huawei
Ascend Y530 and Samsung Galaxy S3 is used for testing.

38

3.7

CHAPTER

TECHNICAL BACKGROUND

Summary

This chapter have presented technologies used to create the system implemented in this thesis. It shows advantages and disadvantages with the different
technologies, and presents the reasons behind choosing the technology that is
used. Each parts of the system are good standalone technologies, but combined
they provide specifications that can support the technical requirements from
Nexans.
The hardware used, which includes Android smart phones supporting Bluetooth
4.0 and NFC, NFC tags for use on the steel drums and BLE beacons for use in
a harsh environment at Nexans are easy to acquire by Nexans. They are also
easy to set up, install and use for the workers at Nexans.
By using a cloud provider, and not a local server, the need for hardware and
software while testing the system is not needed. By using Azure, the IT personnel does not need to acquire new knowledge about a new platform. The user
interface in Azure is also simple to use, and have the required functionality for
administrating the system.

4
Related work
This chapter is divided into two sections. The first section presents work done
within the field of Bluetooth Low Energy. The other section presents other
indoor tracking systems, and the technologies used in these.

4.1

Systems using Bluetooth Low Energy

The BLE technology is a technology not only meant for indoor tracking, and
this section will present other user cases for BLE. Many different services use
BLE as a tool for collecting data, where some of these are presented in figure
4.1. The health care and sports industry, presented in this section, are two of
the largest markets for BLE today.

39

40

CHAPTER

Sensor

Device

R E L AT E D W O R K

Service

Sport
- Heart rate monitor
- Foot sensor

Sport
- Fitness coaching
- Weight watchers

Healthcare
- Glucose meter
- Blood pressure meter

Home
- Remote control
- Home sensors

Stand-alone devices:
- Watches
BLE

Doal-mode devices:
- Smart phones
- Computers

Automobile
- Parking assistant
- Tire pressure monitor

3G, WLAN

Healthcare
- Elderly monitoring
- Glucose monitoring
Home
- Smart home application

Automobile
- Car repair service

Figure 4.1: Some different uses of the BLE technology

4.1.1 BLE within the health sector

Many services within the health care industry have started using the BLE
technology as a tool to improve their work flow. A major challenge in the
health care industry is security. The medical data is often confidential, and the
data should not be accessed by outsiders. One advantage with BLE is that it
adopts the security features of classic Bluetooth. In addition it is possible to
add features to prevent outsiders to read and alter the data. This is done by
limiting the ability to track a transmitting device by using a random device
address that is changed frequently [40].

Blood Glucose Monitoring

More and more people are developing diabetes, and diabetes causes different
health care issues like blindness, arterial disease and amputation of limbs [40].
To prevent these health care issues good monitoring and management of the
blood glucose (BG) levels are needed. The BG levels needs to be kept at a
reasonable level, and this is measured by a BG meter.
By using BLE buildt into a BG meter these measurements will be made simpler
to use for the user and the health care providers. The user will measure the BG
levels as he normally does, and the BG meter will send the results to the users
cell phone and/or the users medical physician. By storing this data the user
and physician can see trends and levels over time, and can take any precautions

4.2

TRACKING SYSTEMS

41

to prevent a spike in BG levels [40].

4.1.2

BLE within sports equipment

There are a vast number of exercise equipment that uses BLE. Due to the long
lasting battery, the user does not need to charge the device as often as when
using the classic Bluetooth technology. Older models of sports devices that were
to communicate digitally as well as having long battery life have been using
mostly ANT+ or WIND. The issue with using these protocols have been that
almost no devices, like tablets or smart phones, have ANT+ chips built in, and
therefore a connection between a smart phone and a heart rate monitor have
been impossible without an extra adapter .
By using BLE different devices can be paired and are able to work together. The
information stored from foot pods, heart rate monitors and location services
can be stored and analysed together for a more personalized exercise log.

4.2

Tracking systems

Tracking, both indoor and outdoor, systems have become increasingly more
popular. The process of tracking systems is called geolocation, position location,
radiolocation or location sensing.

4.2.1 Different tracking technologies

This section will present different tracking solutions available. A comparison
between them and BLE will be done, as well as present how they would work
indoors in the environment at Nexans Rognan.
In figure 4.2 a comparison of different technologies are listed with regards to
their ability to send voice, data, audio, video or state.

1. The AND+ protocol supports interoperability between devices like heart rate monitors,
bike computers and power meters
2. WIND is Polars equivalent to the ANT+ protocol
3. http://www.dcrainmaker.com/2012/07/the-current-state-of-bluetooth-smartlow.html

42

CHAPTER

R E L AT E D W O R K

Figure 4.2: A comparison of different technologies and their uses[41]

GPS
The Global Positioning System (GPS) is one of the best known tracking tools
today. It is an outdoor tracking system, which can be used to track moving
objects, such as cars, planes or people [42]. The targeted objective, whether
it is a car or person, have to be carrying a GPS receiver for receiving radio
waves. These waves are received from multiple satellites and the position is
determined based on the data received using triangulation etc..
Due to the need of receiving signals from satellites, GPS does have limitations
in certain environments. One of the limitations are GPS signal reception. For
the GPS to function properly it needs an undisturbed signal reception from a
minimum of four satellites. Known objects that obscure the signals are water,
soils and walls [43]. The second limitation are signal integrity and accuracy.
The measurements from the different satellites are unique receiver solutions,
but some of these measurements can be wrong. When a satellite position is
wrong, the calculation of the position will also be wrong. The GPS receiver will
measure the time for a signal to travel from the satellite to the receiver. These
measurements are corrupted by different errors, and the effect of the errors
are calculated and removed from the result. The errors can be satellite errors,
signal propagation errors, which is that the earths atmosphere changes the
speed of the signal, and receiver errors [43][44]. Since Nexans Rognan mostly
keep their cable drums inside a factory building, with walls and machines that
will disturb a GPS signal, this technology is impossible to use and an indoor
tracking solution have to be implemented using other technologies.
Indoor tracking systems is used for many different things, such as tracking
items in a warehouse, detecting different personnel or equipment or tracking
elderly people. The primary progress of indoor tracking have been done the

4.2

TRACKING SYSTEMS

43

last 10-15 years, and both the industry and academia are involved in developing
new ways tracking items and people.
Both GPS and BLE can be used to determine the position a device, but there are
some major differences between them. Since GPS uses satellites to determine
the location, it will work better in an outside area, but loses accuracy if it is
exposed to interference from other objects. BLE does not have the same range
as GPS, but it can be used as a proximity tracking system for example if the user
enters the area of a particular store, and the user will get exclusive offers on
his mobile device. Another difference is that GPS uses much power compared
to BLE, and the user will often turn of GPS when it is not in use. BLE on the
other hand goes into sleep mode when it is not in use, and therefore does not
consume as much power [45].

ZigBee
From the ZigBee Alliance, which is an alliance of commercial companies of for
example Philips, Texas Instruments and Silicon Labs , ZigBee was designed.
It is a low-power technology which, most often, runs in the 2.4 GHz band
and can be powered by coin-cell-type batteries. The range can extend up to
approximately 200 meters, but the typical range is 50 meters. For many systems,
the main disadvantage with ZigBee is the low bandwidth of only 250 kbps. For
the system at Nexans this would not have been an issue because of the small
amount of data transferred.
It provides a low-cost system with reliable data transfer, short range operation
and appropriate levels of security [46]. A ZigBee network does require a full
function device to operate as a network coordinator, but most of the devices can
have reduced functionality, which means that they for example only emit the
MAC-address of the device. Figure 4.3 shows how an example of how a ZigBee
network can be built up. This figure is a star where all data goes through the
coordinator. Other designs can be that the fully functional devices can receive
and send data as well.
The Network Coordinator sets up the network, often operates in the receive
state, transmits network beacons and manages and stores information
about the nodes.
The Nodes in the network are designed for low power usage and will
sleep for periods when it is unused and it will transfer data when needed.

4. http://www.zigbee.org/zigbeealliance/our-members/

44

CHAPTER

R E L AT E D W O R K

To provide the security needed ZigBee uses MAC layer security [47]. The MAC
layer uses AES as the core cryptographic algorithm.

Full function device

Coordinator
Reduced function device

Figure 4.3: A figure showing how the ZigBee network is built

ZigBee could have been used to track cable drums within Nexans, but was not
chosen for a couple of reasons. Bluetooth is a more commonly known protocol,
where a lot of support, guides and research can be found when developing
a Bluetooth application. Bluetooth is supported in most of the COTS units
today, and this gives better opportunities to continue to develop the system
and change units as newer, more powerful, and less power consuming devices
enters the market.

Wi-Fi
Wi-Fi have been a proven solution for connecting multiple computers. As
opposed to many outdoor solutions, a Wi-Fi solution does not rely on time
measurements to position the object. Wi-Fi will calculate the approximate
position using the received signal strength from each object. The computers
are connected through wireless local area networks. The disadvantages of a
Wi-Fi solution is that it is relatively expensive to use. It also have high power
consumption, and it does require frequent (depending on the battery capacity)
recharges [48].
There are many differences between BLE and Wi-Fi. Range is one of them.
Depending on the Wi-Fi protocol used, Wi-Fi can offer up to 650 meters range,
5. Advanced Encryption Standard
6. Commercial Off-The-Shelf

4.2

TRACKING SYSTEMS

45

while BLE advertise the range to be approximately 50 meters. At Nexans the

range of the BLE beacons have shown to be around 20 meters because of the
amount of obstacles and interference from other machines.
Data transfer rates is another difference. Some of the latest specifications of
BLE is that it can offer up to 25 mbps transfer rate, while Wi-Fi can transfer at
a rate of 250 mbps. For other systems, that have to transfer large amounts of
data, this would be a significant difference. But for the system implemented in
this thesis, it does not have a significant impact.
Bluetooth does not offer much security, and is just limited to key matching [45].
This means that information that should be kept secret is not meant to be
transferred via Bluetooth. But Bluetooth does support AES-128 encryption of
the data. Wi-Fi on the other hand is secured by WPA2 [45, 49].

Radio Frequency Identication (RFID)

RFID is a passive technology that only prompts a transaction whenever the object passes near a sensor [50]. It will consist of a RFID-tag, a transceiver, which
decodes and interprets the data, and a scanning antenna that communicates
with and gives power to the RFID tag.
RFID systems can either be passive or active. The active tags requires a power
source, whilst the passive does not. The passive tags are small, and contains an
antenna and a semi-conductor chip attached to the antenna. The RFID reader
will power up the tag to read information from it. Further, RFID is divided
into near-field and far-field RFID. Far-field would be the type used at Nexans,
where tags capture electromagnetic waves from the RFID reader. The typical
reader can emit signals to readers 3 meters away [51].
A system using RDIF for indoor tracking is SpotON [52]. The concept behind
this system is having multiple base stations, each base station provides signal
strength measurements to the central station. The central station calculates
the exact position of the objects using triangulation [52].
One of the main differences between BLE and RFID is that the passive RFID-tag
does not need to contain any batteries, since it is powered by the antenna
nearby. By using RFID-tags they can last for years, whilst BLE beacons run out
of battery after a couple of years. One of the major disadvantages with RFID
is the limited range, with only a 3 meter typical range without disturbance
from objects. More powerful readers could extend the range, but compared to
7. Wi-Fi Protected Access

46

CHAPTER

R E L AT E D W O R K

the expected range of 50 meters of BLE, RFID would not work as well inside
Nexans.

Infrared (IR)
There are two kinds of IR motion systems, and those are active (AIR) and
passive (PIR). Passive means that they will measure the change in energy of the
surrounding area, and in this way detect objects moving. The passive method
will not work at the Nexans plant since we wish to detect what objects (cable
drums) that have moved, and not just that something have moved .
The active IR have two implementations, the proximity sensors and motion
sensors. The proximity sensors can be used to sense when a persons hands
are close to a hand dryer in a public bathroom or to automatically open a
garbage dispenser when a person gets closer to it. Motion sensors will emit an
IR-beam, which is received by an IR receiver. The disadvantage with IR is that
the beam can easily be hidden by other objects, so that the position will not
be registered .
The main difference between IR and BLE is how the signal is distributed. IR is
a beam, which have to be in line of sight of the receiver. With an environment
like the one at Nexans, where there are large obstacles and the drums are often
stacked in clusters like figure 2.2, a solution where the receiver and transmitter
have to me in line would be hard to maintain. When using BLE, the signal can
reach the receiver even though the beacon and receiver is not in line of sight
of each other.

LTE
LTE Direct provides device-to-device communications over cellular connections.
It can have a range of up to 500 meters, supports thousands of devices, and
is designed to consume little power. It uses radio signals, and they can set to
private that only certain devices or people are able to receive the signals, or
they can be public, so that everyone can receive them . By using LTE Direct
the developers and users do not need extra beacon hardware, as in Bluetooth,
8. http://www.facilitiesnet.com/lighting/article/Occupancy-Sensors-Passive-InfraredUltrasonic-and-Dual-Technology-Facility-Management-Lighting-Feature9608
9. http://www.digikey.com/en/articles/techzone/2012/jun/sensing-motion-with-passiveinfrared-pir-sensors
10. http://www.guardall.com/product.asp?PageID=245
11. http://arc.applause.com/cards/lte-direct-future-beacons-explained/
12. https://www.qualcomm.com/invention/research/projects/lte-direct

4.2

TRACKING SYSTEMS

47

but the cellular device becomes the beacon.

It is created as a competitor to BLE, but it supports longer range. It has somewhat
other use cases. Examples are for a user to discover a particular store, to receive
offers from a store or to notify the user if a friend is nearby. Since the LTE
supporting smart phone have to be used as the beacon, this technology would
not work in the use case at Nexans Rognan, since a cable drum will have to
have a LTE device attached to it.

4.2.2 Example systems

This section will present some working tracking systems, starting with the first
well-known system, Active Badge.
The first indoor tracking system created was Active Badge [53]. In the early
1990s, when this system was created, the way people connected and located
other people was via a pager. The pager system works by sending a signal to
a receiver, and the receiver will produce sound to alert the pagers owner that
someone wants to contact him, and the pager had a display showing who the
caller was. One drawback with the paging system was that the caller did not
know whether or not the callee received the message or not. Around this time
the way people located others were by using key cards which registered when
a user entered or left a room or building.
When Active Badge was applied the users wore a badge, and this badge emitted
an unique code every 15 seconds using infrared signals [53]. The signals are
collected and sent to a master station so that the positions registered can be
read by others. The Active badge system is quite similar to the BLE tracking
system created in this thesis, but it does have a couple of drawbacks. It has only
a range of approximately six meters, and the signals cannot pass through walls.
This would mean that the grid system at Nexans would have to have very small
masks. The system is quite costly in terms of installation and maintenance. It
also performs poorly in bright environments, for example in direct sunlight [54].
The last drawback is that when there are many badges present within the range
of a station, the signals collide and some will be lost [53].
RADAR [54] is another tracking system using radio frequency (RF) instead of
IR. It is the Wi-Fi signal-strength based indoor positioning system. This system
works by recording signals strength information at multiple base stations. These
signals are interpreted using different methods, and the location is stored for
later use. It works using a map, which is a lookup table holding collections
of packet signal strengths and where in the building these where measured.
The location is calculated using these reference points. RADAR also takes

48

CHAPTER

R E L AT E D W O R K

into account any environmental challenges in the different areas inside the
building [54]. RADAR is able to calculate the object or persons location with
an accuracy of a few meters. A system using RF instead of IR is less prone to
disturbances from different objects, such as walls or machines.
LANDMARC is a system based on active RFID [50]. The purpose of the research
done was to investigate whether or not RFID could be used in an indoor
tracking system. In LANDMARC reference tags, or reference points, are placed
at given locations. Signal strength measurements from the moving objects
(eg. cable drums) are compared to the reference points, and the location is
calculated. This will limit the cost of adding more expensive RFID readers, and
instead use more of the less expensive RFID tags. This system require signal
strength information from each tag to the readers. These signals strengths
varies depending on environmental challenges in each area, and to increase
the accuracy of the readings and measurements, more RDIF readers are needed.
So when implementing LANDMARC in an environment, for example Nexans
Rognan, the desired accuracy have to be weighed up against the costs of
expensive RFID readers.

4.3

Summary

As the first section, about other user cases using BLE, presents, the BLE technology is a versatile technology that can be used to solve many issues in different
industries. In this thesis, BLE is only used as a tool for tracking cable drums.
If requirements change from Nexans more data can be collected from the
beacons. Examples are temperature changes on the cable drums. They can also
use the technology to track and analyse data from other elements at the plant,
like tire pressure monitors on the trucks and temperature sensors at different
locations.
The second section presents other technologies for tracking, as well as some
systems using these technologies. A comparison of the different technologies,
with the Nexans environment in mind, have been done. BLE is one of two
technologies that could have been used at Nexans, and the other is ZigBee. BLE
have been chosen for its easy accessible BLE beacons and easy set-up.

5
Design
5.1

Architectural Overview

The Nexans Tracking System is a loosely coupled modular system, consisting of

three mobile applications, a RESTful middleware and a SQL database. Loosely
coupled systems has some important benefits for Nexans, including modularity
which allows for easily upgrading the modules that the system consists of.
An overall design of the system is shown in figure 5.1 and a more thorough
description is presented below

49

50

CHAPTER

DESIGN

BLE Scanner
Application

BLE User
Application

Middleware

Database

NFC
Application

Web site

Figure 5.1: The overall design of the tracking system

One of the subgoals of this thesis was to implement a system for tracking cable
drums inside the factory. This will limit the time spent searching for drums
every day, and increase efficiency by knowing where every cable drum in the
factory is located. By knowing the location of each cable drum, the down-time
of each machine can be decreased. This increases the production rate, which
will again mean that Nexans can accept more orders or are able to make shorter
deadlines for customers.
The subgoal of tracking cable drums could have been achieved with only
implementing the cell-based positioning system, including the BLE Scanner
Application described below and a simple database containing only the names
of the reference points and the cable drum number. This application was created
in the pre-project, but was only used for testing how beacons behaved inside
the factory.
However, that system would lack in areas such as user-friendliness, areas that are
important for the users at Nexans. This creates a need for another goal: to create

5.1

ARCHITECTURAL OVERVIEW

51

a user-friendly interface that Nexans can use in their daily production. This
functionality addresses other issues regarding the cable drums. For example
representing specification details or remarks to the drums.
The goal of this thesis is to be able to follow the cable drum from when it enters
the factory the first time and until it is disposed. The system should contain
functionality for tracking and maintaining all aspects of the drums life at
Nexans, from production details, to maintenance and shipping. In other words:
a fully functioning tracking system should be designed and implemented.
Each part of the system are created to expand the functionality of the system and
together to create an easy-to-use, fully functional system for the administration
and workers to use on a daily basis. A benefit of the system is the possibility to
retire older, more inefficient operations, such as adding remarks to the drum
or adding an order number to the drum ticket.

Mobile applications
The first application is the BLE Reference Point Scanner. This application
is installed on the reference points (smart phones) in the cell-based
positioning system. It periodically scans the area for beacons, and reports
any cable drums found to the middleware through a RESTful api. Most of
the functionality of this application was implemented in the pre-project.
The pre-project did not implement a tracking system for the cable drums,
and the application was only used for testing the specifications of the
beacons.
The second application, the BLE User Application, is used by the workers
at the plant. After finding the approximate location of a cable drum
using the web interface, the user can go to that area He can use the BLE
User Application application to navigate within this area, and to locate
the exact position of the drum. This application is not connected to any
of the other modules of this system and does not require an internet
connection.
The third mobile application is the NFC Drum Information Application.
There are cable drums located all around the plant, which either are
empty or containing some raw material or product. The users does not
always have control over each cable drum at the plant, and this application
allows the user with a easier way of identifying the cable drum and the
product that is coiled on the drum. When a user wishes to gain more
information about the drum, both specifications and status, he can use
1. Range, advertising interval, robustness, etc.

52

CHAPTER

DESIGN

the NFC application installed on his phone. By touching the NFC tag stuck
on the drum, the user will be redirected to the cable drums web site. He
can then read and edit information about the drum (granted that he has
the correct access rights).

Web site The web site is created to collect information from the database
and present the information in an user-friendly way. More details about the
different functionalities of the site is presented in the middleware section of
this chapter.

Middleware The middleware will handle any incoming messages from the
web site or mobile applications. It contains logic for deciding whether the user
has access to view the page, or if a cable drum have moved and the position
should be updated.

Database The database contains multiple tables with information about

the cable drums, localization, user credentials and the reference points (smart
phones) used. More details about the contents of the tables can bee viewed in
Appendix B.

5.2

Mobile applications

The mobile applications are divided into three applications. The two BLE
applications is built around the same code-base with minor modifications each.
They are built upon the BLE scanner implemented in the pre-project.

5.2.1 BLE Reference Point Application

This application is used to create a grid inside the Nexans factory. An example
of the grid is shown in figure 5.2. This application is created to implement
support one of the subgoals of this thesis; to create a indoor tracking system
for cable drums.

5.2

53

M O B I L E A P P L I C AT I O N S

Figure 5.2: Example of cable drums within the grid

This application will, every two minutes, scan for nearby beacons. The beacons
are placed on each cable drum, with a dependency between the MAC-address
of the beacon and the drum id provided by Nexans. The two minute delay
is chosen based on the behaviour of the drums, they are located at the same
spot for extended periods of time. This means that continuous updates are not
needed, nor is a more accurate time variable required by Nexans. By increasing
the delay, the power requirements and data requirements lowered. The longer
the delay is, the fewer database queries the middleware will have to perform.
This delay can easily be changed in the future if needed.

54

CHAPTER

DESIGN

The application generates a list, as seen in figure 5.3, and it is populated with
the MAC-addresses of the nearby devices. The screen of the mobile device is
never dimmed or turned of, to prevent the application of exiting and to make it
easier to debug. Each of the entries in the list is sent to the middleware, which
handles whether or not the position of the drum have changed.

Figure 5.3: The BLE scanner used for scanning the area for beacons

5.2.2 BLE User Application

This application is run on the users phone and it has no connection to the
middleware, the other applications nor the database. It is created to give an
more accurate measurement of the distance between the user and the cable
drum. The BLE Reference Point scanner gives a rough estimate of the position

5.2

M O B I L E A P P L I C AT I O N S

55

of the cable drum, while this application will create a more accurate position
since the user will receive the RSSI measurements from the beacon.
From information found on the web page (populated by the BLE Reference Point
Application), the user will know what area in the factory the cable drum is
located. This can for example be in one of the storage tents, near the WaterSoak
pool or near the laboratory. The BLE User Application is started on command
from the user and scans the surrounding area for one minute and populate the
list of drums found. If needed, for example if the users moves around, can the
scanning be restarted.
When started, the BLE User Scanner will populate a list with the MAC-addresses
of the beacons that are nearby. From the web page, shown in table one of figure
5.8, the user will know which MAC-address he is looking for. The user will
locate the beacon address in the application and may move according to the
RSSI measurements located underneath the MAC-address of the beacon in the
list. This RSSI measurement will tell the user if he is close or far away from
the beacon. If the user have an measurement of -100 dBm he is further away
than if he had had a measurement of -50 dBm.

56

CHAPTER

DESIGN

Figure 5.4: Screenshot of the BLE User Scanner, showing the beacon MAC-address and
measured RSSI

5.2.3 NFC Drum Information Application

At Nexans, each drum currently has a paper ticket as shown in figure 2.1, stuck
on the drum. This has information about the current status of the drum and
the users have to update this when the drum enters a new state. This status
includes for example how much rust the drum has and how clean the drum
is. This paper ticket can easily be lost or ruined when the drum is exposed to
different environment challenges. The ticket also has to be removed before the
drum soaked in water. This is an inefficient and fault-prone method of keeping
track of the cable drum, and it is difficult to keep track of which drums that
should be taken into maintenance.

5.2

M O B I L E A P P L I C AT I O N S

57

The NFC application created in this thesis should replace this method. By using
NFC tags and the web site, it is simpler to update and keep track of the status
of the drum. The NFC tags used in this thesis are water proof, and are specially
made to be placed on metal. The NFC tag will contain the drum id of the cable
drum, and by reading the sticker using the NFC Drum Information Application
the user will be redirected to a web page containing information about the
drum. If the user has the sufficient access rights, he will be able to edit the
data on the web site.

58

5.3

CHAPTER

DESIGN

Middleware

The middleware is what contains most of the logic in the system. It is divided
into several parts.

User Accounts The user accounts consist of two parts, the account itself and
the user roles. First the user have to register with name, email and password
as shown in figure 5.5. A new user are then created in the system, but it has
no user roles assigned to it. This means that they will not be able to see any
sensitive information on the site, and the site will present the same data as
when the user was not logged in.
When a user is assigned an user role he can see the data assigned to that role.
The administrator have to assign the required roles to the user based on the
work that the user is assigned.

Figure 5.5: Register an account

In figure 5.6 the different access levels are shown. New access levels can easily
be added, unused ones can be deleted and the existing ones can be edited. The
different levels gives an indication on what the user can access in the system.
The user role, canEdit is the administrator of the system.

5.3

M I D D L E WA R E

59

Figure 5.6: The different access levels in Nexans Tracking System

Figure 5.7 shows the different user roles in the system, and what features of
the system they can access. As of now the higher level user have the same
accesses as the lower level ones in addition to some new ones. This means that
the option of adding several user roles to an user is redundant in this thesis.
For future work functionality can be added so that one of the lower user roles
should be able to access some data, but not a higher user role. For example
could the MasterUser not be able to edit maintenance history of a drum, but
the User should be able to. This is why the functionality of adding more roles
to a user have been implemented.

Figure 5.7: A table showing the different user roles in the system

The administrator account is created at first compile. It is not possible to

create another administrator account runtime, the administrators of the system
have to be specified in the code. The administrator have access to all data
in the database, and is the only user that can add user roles to the users
registered.
When the administrator logs in on the site, he will have access to the different
functionalities implemented, but the main role of the administrator is to give
other users the correct access rights. Figure 5.8 shows the home page of the

60

CHAPTER

DESIGN

administrator. The bottom table, list of users with no role, shows the users that
still not have been assigned any user roles.

Figure 5.8: The administrator startpage

By navigating to the User role management, the administrator can see the
different user roles, add and delete roles, retrieve user roles for a person or add
and delete a user from a user role.

5.3

61

M I D D L E WA R E

Figure 5.9: The listing of the different roles in the system

Figure 5.10: User role management

Cable Drums This section contains information about the cable drums. This
will include the specifications provided by the cable drum manufacturer, the
general status of the drum and maintenance history.

62

CHAPTER

DESIGN

Each of the information fields (cable drum specifications, general status and
maintenance history) requires the user to be assigned one or more user role(s).
The users ability of interact with the fields depends on the access level that he
is given. Some roles can only inspect data, whilst others can also edit.

Cable drums
Specifications

General
status

Maintenance
history

Figure 5.11: Information found within the Cable Drums section

Cell-based positioning The localizations of the cable drums is one of the

more important parts of the system. To localize the beacons inside the plant, a
grid of reference points, using the smart phone Huawei Ascend Y530, have been
created. The technique used in this thesis is a cell-based positioning technique.
The application BLE Reference Point Application is implemented to support this
functionality. The factory is covered with reference points, where each reference
will read data from nearby beacons. To create full coverage of the factory hall
and to avoid deadspots, some of the circles overlap[55].
Each of the reference points are smart phones, which have an unique MACaddress. This address is mapped to a name, given by Nexans, in the database.
The name, depending on where in the plant it is located, can for example be
laboratory, west emergency exit , or machine name (for example WaterSoak).
These names are well known for workers at Nexans.
In figure 5.2, the reference points are a represented with a number from 1 to
6. When a user accesses the web site, he will have information about which
reference points that can see the beacon. If the user was looking for the red
drum in figure 5.2, he will know that it is only seen by reference point number
2. If the user is looking for the yellow cable drum, this is seen by both reference
point 3 and 4.
By using a grid, the possible locations of a cable drum is limited to maybe 50
square meters, instead of 18 000 square meters (total area of the factory).

5.3

63

M I D D L E WA R E

How it works Within the list of cable drums on the web site, the user can
search for one particular drum or drum type. The results are presented in a
list, like in figure 5.12. Each of the drums contains a link which redirects the
user to a page showing where the cable drums is located.
The location page includes the name of the cable drum, when it was last seen
and which reference points that are nearby. In the case in figure 5.13, the cable
drum is only seen by one of the reference points. The user can then take his
Android smart phone and go to the approximate location he found on the web
site. If the user starts up the BLE User Application the screen will be populated
with a list containing the MAC-addresses of the beacons nearby as well as the
RSSI measurements from these beacons. By moving towards a beacon, the
signal will grow stronger, but with moving from the beacon the signal will get
weaker. This way the user is able to navigate to the drum he is looking for
within the grid mask of created by the reference points.

Figure 5.12: List of cable drums

Figure 5.13: Position of cable drum

The system will only store new and up-to-date entries in the database. Old
and stale entries are deleted as needed. Each time a reference point sends data

64

CHAPTER

DESIGN

to the middleware, the middleware will check the data up against previous
entries regarding that cable drum. Depending on which reference point, time
and cable drum, the database will be updated with correct information. The
logic behind these updates are shown in figure 5.14, but an explanation is also
presented below.
It will check if the beacon exists in the database, if not the location is
added to the database.
If the beacon is already registered in the database, it will check whether
or not it is the same reference point that has the old entry that sends an
update, or if it is a new reference point that sends data.
If it is the old one, the middleware will only update the time variable of
the entry.
If it is a new reference point, which is not registered with that beacons
MAC-address before, the middleware will check if the beacon can be
registered as seen by both reference points at the same time. This is an
example of overlapping range circles within the figure 5.2. A beacon can
be seen by two reference points at the same time, and this is handled in
the middleware.
If it is possible that the beacon can be seen by multiple reference points,
a new entry with the MAC-address of the new reference point is added to
the database.
If the beacon can not be seen by the old reference point entry as well
as the new reference point entry, the old entry is deleted and the new is
added.

5.4

65

D ATA B A S E

No

No

Yes

Yes

No

Can the beacon

be at current
position as well as
the position of the
last entry?

Is the beacon
registered on this
position already?

Does the beacon

exist in the DB?

Delete entries
that are not
correct and add
new, correct entry

Yes

Add new entry

Update time
variable

Add:
BeaconID
DeviceID
Time now

Figure 5.14: The logic behind updating the position of the cable drums

5.4

Database

The database is a an Azure SQL database. This database is made up of the

schema that can been seen in appendix B. Since the Azure SQL server is built
using Microsoft SQL, and is compatible with the local version of the Microsoft
SQL server, is Nexans able to move their data of Azure and to a local database.
By using SQL and only having the middleware communicating to the database
can the SQL storage be moved to another cloud provider or other type of SQL
service if needed. This might require some changes in the schema or queries
based on differences in the SQL dialect.

6
Implementation
This section presents implementation details about the middleware and mobile
applications created. The middleware section presents some of the frameworks
used, and how these benefit the system implemented. The section mobile applications presents the different applications and any special methods used.
The main functionality and contribution of this thesis is designed and implemented in the middleware, so therefore it represents the majority of the
chapter.

6.1

Middleware

The middleware is built upon Microsofts ASP.NET MVC. It is a open source

web application framework. This framework separates the application into
three main components: the model, the view and the controller.
The model represent the core of the application, for example a list of database
records. Each of the tables in the database is represented in a model, for
example CableDrums or Devices. The Controller handles the input from the user.
When an user in the system creates a new cable drum, this insert is handled in
1. Model-View-Controller
2. http://www.asp.net/mvc

67

68

CHAPTER

I M P L E M E N TAT I O N

the controller. It contains the logic for determining of the user has the correct
access rights and if the input is in the correct format. The View displays the
data for the user.
By separating the input, business and UI logic the complexity of the application
is more manageable. This is because the developer can focus of one aspect of
the task at a time. It will also help the developers with parallel developing of
the application. One developer can work on the controller part, while the other
manages the view.

Controller The middleware consists of multiple controllers to separate the

code for each task within the system. These controllers are shown in figure 6.1.
Included in the explanation about each controller a listing of the most used
URLs is presented
CableDrums

Devices

Localizations

Roles

Account

Home

Figure 6.1: The different controllers created

The CableDrum controller includes the functionality of adding, updating

and deleting information about each cable drum at the plant. It does
only keep information about the specifications, maintenance history and
general status of the drum, as presented in design.
Listing 6.1: Routes within the Cable Drums controller

<domain>/CD
<domain>/CD/ C r e a t e
<domain>/CD/ D e l e t e /DrumID
<domain>/CD/ D e t a i l s /DrumID
<domain>/CD/ E d i t /DrumID
The Devices controller makes it possible to add new reference points to
the cell-based positioning system. It contains variable to position the
reference point within the grid. For example is position (1,1) and (1,2)
located next to each other, while (1,1) and (10,10) might be in opposite
corners within the factory.
Listing 6.2: Routes within the Devices controller

3. String, Integer, DateTime, etc.

4. https://msdn.microsoft.com/en-us/library/dd381412(v=vs.108).aspx

6.1

69

M I D D L E WA R E

<domain>/D e v i c e s
<domain>/D e v i c e s / C r e a t e
<domain>/D e v i c e s / E d i t / DeviceID
<domain>/D e v i c e s / D e l e t e / DeviceID

The Localization controller functions are used by the BLE Reference Point
Scanner to add or update the position of a beacon (cable drum). This
controller collects information about the Devices to check whether the
beacon (cable drum) can be seen by multiple reference points at the same
time. For example can a beacon be seen by both (1,1) and (1,2), since
they are neighbours, but it cannot be seen by (1,1) and (3,3) since they
are too far away from each other.
Listing 6.3: Routes within the Localization controller

<domain>/ L o c a l i z a t i o n s
<domain>/ L o c a l i z a t i o n s / C r e a t e / BeaconID / DeviceID
<domain>/ L o c a l i z a t i o n s / D e t a i l s / BeaconID

The Roles controller includes functionality for adding new roles to the
system. It also provides methods for adding and deleting roles to a user
account.
Listing 6.4: Routes within the Roles controller

<domain>/R o l e s
<domain>/R o l e s / C r e a t e
<domain>/R o l e s / E d i t ? roleName=UserRole
<domain>/R o l e s / ManageUserRoles

The Account controller contains methods for registering a new account

and updating an existing account. It is responsible for user access, ensuring that every user is an authenticated user with the correct access
rights.
Listing 6.5: Routes within the Account controller

<domain>/Account / R e g i s t e r
<domain>/Account / Login
<domain>/Manage
<domain>/Manage/ ChangePassword

70

CHAPTER

I M P L E M E N TAT I O N

The Home controller provides the user with a personalized home page. As
shown in design, in figure 5.8, the home page of the administrator shows
a list of users with no role, as well as all cable drums in the database.
Each controller contains a set of methods, where the main methods are shown
in figure 6.2. Each of the controllers also performs checks to see if the user
does have the required access rights to see the data.
The Index function, for example within the cable drums controller, lists
all cable drums in the database.
Details lists details about one particular drum, for example diameter, core
diameter and weight.
Create checks input provided by the View, and inserts the data into the
database.
Edit updates the database with input from the View, with the proviso
that the data is in the correct format.
Delete deletes the entry from the database.

Index

Details

Create

Edit

Delete

Figure 6.2: The main functionality within each controller

View The views and controllers often operate as a pair, where data shown in
the view is populated by the controller. Figure 6.3 contains the different views
of the system.
In addition to the views shown in figure 6.3, it also contains a shared view.
This view contains the main design of the web page. Views in the system
inherits the template created in the shared view. The template contains a link
to Cabledrums, Devices, Home and Localizations.

5. String, DateTime, Integer, etc.

6.1

71

M I D D L E WA R E

Shared
Account

CableDrums

Devices

Home

Localizations

Roles

Figure 6.3: An overall figure of the Views

Model A model contains the data, which is retrieved by the controller and is
displayed in the view. An overview of the models created in this thesis is shown
in figure 6.4. More details about what which model contains is provided in
Appendix B.

Account

CableDrums

Device

Localization

Figure 6.4: An overall figure of the Models

6.1.1 Access Rights

The middleware implemented is protected by SSL and the authorize attribute.
The Authorize attribute specifies that access to a controller or action is restricted
only to those who meet the User Role requirement. Users registered in the system
does contain an attribute specifying the granted access level. This attribute is
determined by the work tasks at Nexans. Only the administrator of the system
is able to add new users and then the access level attribute is set. This can be
altered if the tasks of the user changes.
For implementing different user levels in the system ASP.NET Identity is used.
The Identity framework can be used with all of the ASP.NET frameworks, such
as ASP.NET MVC[56]. The system stores the user information in a database,
and it uses Entity Framework Code First to implement the persistence mechanism.
In this implementation anonymous users are allowed to look at the home page
of the site. This is done by adding the attribute Allow Anonymous to the given
methods.

6. https://msdn.microsoft.com/en-us/library/system.web.mvc.authorizeattribute.aspx
7. http://www.asp.net/identity
8. http://blogs.msdn.com/b/rickandy/archive/2012/03/23/securing-your-asp-net-mvc-4app-and-the-new-allowanonymous-attribute.aspx

72

6.1.2

CHAPTER

I M P L E M E N TAT I O N

Routing

MVC contains methods to create routes. These routes goes to different parts of
the web site, and the developer can make routes that connects the different elements and makes it possible to redirect a request to the correct view/controller.
In this thesis, the routes have been used to connect the mobile applications
created and the middleware.
The first route created, in listing 6.6, creates an route used for sending the
MAC-address of a reference point (DID) and the MAC-address of the beacon
found (BID).
Listing 6.6: Route for passing information from "BLE Reference point application" to
the middleware

r o u t e s . MapRoute (
" Insert " ,
" { c o n t r o l l e r }/{ a c t i o n }/{ BID }/{ DID} " ,
new {
controller = " Localizations " ,
action = " Create " ,
BID = " " ,
DID = " " }
);

The second route defined is used in the NFC application. When the user reads
a NFC tag placed on the drum, he will be redirected to the correct web site.
This is done by a HTTP request containing the BeaconID (BID) written on the
NFC tag.
Listing 6.7: Route for getting the web site containing information about the cable
drum

r o u t e s . MapRoute (
" Index " ,
" { c o n t r o l l e r }/{ a c t i o n }/{ BID } " ,
new {
c o n t r o l l e r = "CD" ,
action = " Details " ,
BID = " " }
);

6.2

6.2

M O B I L E A P P L I C AT I O N S

73

Mobile applications

The BLE Reference Point Scanner did have most of its functionality implemented
in the pre-project, but some alterations have been done to it. The BLE User
Scanner is built upon the BLE Reference Point Scanner, but it is not connected to
the backend. Additional functionality of including measurements of the RSSI
from each beacon have been added.
The NFC application in the pre-project was discarded because the functionality
did not suit the design of the system. A new application have been implemented
in this thesis.

6.2.1 BLE Applications

The two BLE applications are built upon the same base, only with minor differences. They both contain a list, which is populated with the MAC-addresses of
the nearby beacons. To enable the built-in platform support for BLE Android
4.3 (API Level 18) is used.

BLE Reference Point Scanner The Stationary application is built up as a

loop, which runs every two minutes. Within a this loop, the application scans
the area for beacons, populates the list, and sends the data (MAC-addresses)
found to the middleware, which handles the input.
The data is sent as a HTTP request, which includes the MAC-address of the
reference point as well as the MAC-address of the beacons nearby. The URL
used to send data from application to middleware is shown in listing 6.6. It
will send one request to the backend for each MAC-address populated in the
list shown on the display.

BLE User Scanner The user scanner does not have a connection with the
backend, and therefore it does not need to set up a connection. It does not
require a internet connection.
The BLE User Scanner have a continuous measurement of the RSSI of each
beacon that the smart device sees. To store the RSSI measurements from
each device a HashMap is used. It contains the BluetoothDevice and an Integer
representing the RSSI-measurements.

74

CHAPTER

I M P L E M E N TAT I O N

6.2.2 NFC Drum Information

This application is started at command from the user. When the user touches
the NFC tag stuck on the cable drum an activity is started and the tag is read.
The tag contains the SuppliersID provided by Nexans. This SuppliersID is unique
for each cable drum at the plant. After a tag is read the user is redirected to
the web page containing information about that drum. The URL the user is
redirected to is shown in 6.7.

7
Evaluation and challenges
This chapter will present challenges that are relevant for this solution as well
as an evaluation of the technology and methods used.

7.1

Testing

Due to the short time span and limited resources only a limited part of the
factory hall have been used for testing. Only a few reference points (smart
phones) have been used for testing the system created. This is sufficient for
testing how cell-based positioning works, the range of the beacons, the battery
life of the smart devices used by the users and the logic implemented within
the middleware.

7.1.1 Locating the cable drums within the factory

The challenges presented in this section are time consuming tasks and have
therefore not been prioritized to perform full tests on. The tests performed
in this thesis have been enough to predict the behaviour of the system implemented and technology chosen. Some of the tests were started in previous
projects, but have been continued in this thesis, and therefore they are presented in this thesis also.

75

76

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

Range
The factory hall at Nexans spans 18 000 square meters. It also have numerous
storage tents outside the plant. Inside the factory hall, Nexans have different
sections, machines and equipment for producing and testing cables. In the
storage tents they mostly store empty cable drums, or cable drums containing
some unfinished product or raw material.
One known challenge with BLE, is the possible interference of signal by other
machines and objects. As presented before, the advertised range of BLE beacons
are 50 meters, but the signal interferences from other objects causes limitations
regarding the range. The extent of this possible issue have been tested on
numerous occasions on different locations of the plant, with different cable
drums. It shows that the expected range of the beacon inside the factory
hall is about half of the advertised beacon range. Approximate measurements
have shown that the beacons have a range of approximately 20 meters in
the most difficult environments, while some areas provide a range of 30-40
meters.
These challenges have to be addressed in future works. It demands a significant
offline or learning phase to populate the grid within the plant with regards to
the known objects that cause signal disturbances. Thorough testing have to be
performed at all locations of the plant to estimate suitable locations to place
the reference points as well as to register which of the reference points range
circles that overlap.

Accuracy
The BLE user application shows the RSSI received from each of the beacons. The
closer to zero the RSSI measures are, the closer the user is to the beacons. When
using only one type of beacon all beacons will use the same broadcasting power.
If Nexans use different beacons, the signal strength of each beacon emitted can
be different. When two beacons send signals with different strengths, but are
located at the same spot, the users scanner will detect the beacon which sends
out with highest signal strength as closer than the other beacon[44].
At Nexans, only the StickNFind Pro beacons have been tested thoroughly, and
they are all set to emit with the same broadcasting power. For the future,
if Nexans changes the beacon type, this issue might have to be addressed
again.
Testing at Nexans have shown that the accuracy of these are RSSI measurements
are low, and that the measures RSSI cannot be directly linked to a distance in

7.1

TESTING

77

meters. The BLE User Scanner can only be used to give the user an indication of
which direction the cable drum is located. The user will know if he is getting
further away or closer to the cable drum.
The reason why the accuracy of these RSSI measurements are so low, is the
amount of disturbance/interference from other objects in the factory hall. There
are large machines and walls located around the plant. These stationary objects
could have been plotted on the grid map, and the signal disturbance from these
objects could have been calculated in the offline phase of the cell-based position
technique used in this thesis. The challenge at Nexans is the many moving
objects that cause signal disturbance. These objects are often trucks and other
large steel cable drums. These moving objects cannot be plotted in the offline
phase of the cell-based position since they are often moved.

7.1.2

BLE beacons

The main technology to test out in this thesis, is how the BLE beacons behave
inside the Nexans plant. This section will present different challenges that can
or have occurred.

Battery life
The advertised battery life of the beacons are 2-5 years. The testing performed
this far can neither confirm nor deny this. The beacons placed on the cable
drums in May 2014 are still operative in May 2015. This includes both Estimote
beacons as well as StickNFind Pro beacons.
What can affect battery life significantly is temperature changes. If a battery
is exposed to cold weather the battery life may be compromised. The storage
tents are not warmed in the winter, nor are they air-conditioned in the summer.
This creates temperature swings that can affect the beacons battery life. This
have not been tested in this thesis, but is left for future work.
The StickNFind SDK provided a variable for battery measurements, and these
have been measured and registered. These measurements fluctuated, and did
not make any sense because they were neither measured in hours, days, years or
percentages. After multiple emails to the StickNFind company, they responded
with that these measurements are incorrect and does not give any indication
of the expected battery life left of the beacons. As of now there are no methods
of estimating the remaining battery power of the beacons.

78

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

Robustness
The beacons used at Nexans would have to endure a rather harsh environment,
compared to other environments like hospitals or warehouses. One possible
challenge was if the beacons were broken or damaged easily. It was also a
possibility that the larger Estimote beacons were more prone to damages than
the smaller StickNFind Pro. During the last year of testing, this have not caused
any problems. Both the StickNFind Pro and Estimote beacons have endured all
challenges, from water soak, voltage tests and temperature changes.
These tests have shown that the size of the beacons have not had such an
impact as first assumed. This might be due to the placement of the beacons on
the cable drum. The beacons are located inside the grooves of the drum, and
therefore somewhat protected from damage. For future works the StickNFind
Pro would nevertheless be less prone to damage than Estimote.

7.1.3

Position history

In this thesis position history have not been stored to limit the data in the
database. An interesting feature could have been to store the information about
the past locations of the cable drum. This can be used to track the activity
of each drum, and to analyse how much each drum is in use. It would also
make it simpler to find faults in the cell-based positioning. Since the range of
the beacons fluctuates between 20 and 40 meters, depending on interference,
inside the hall the number of reference points in the grid that can see the
beacons will also fluctuate. The position history can be used to test the range
of the reference points, and to see which reference points that has overlapping
range areas.
The last position of each cable drum is always stored in the database, and
is never timed out. If the location entry did time out, and got deleted after
a while, the workers would not have any way of knowing where it was last
seen, or why the cable drum is registered with a location anymore. If the last
position is stored in the database, it is possible for the workers to find the drum
again. Reasons why the drum have not been seen lately is because it are being
performed tests on in the water soak, it have been moved, sold to a customer
or the beacon might have run out of battery.

1. The water soak tests are where they soak the cable drum and cable in a pool of water.
Then AC/DC testing is performed on the cable

7.1

TESTING

7.1.4

79

Machine Learning

As the system is designed and implemented, it consists of a significant offline/learning phase where the range and interference on different locations
within the factory is measured. This offline phase will estimate which reference point ranges that overlap. Where two reference point ranges overlap, the
beacon (cable drum) can be seen by both. Since the range within the plant is
not constant at for example 20 meters, maintaining which ranges overlap is
very time consuming and nearly impossible to keep correct at all times. The
reason why the range fluctuate is the difference in amount of objects blocking
the signal. Where a cable drum is located in a cluster, as seen in figure 2.2, the
signal is blocked and does not have the same range as if it was placed in an
open space, such as in figure 2.3.
To limit the manual learning phase in this system, machine learning could
have been implemented. Nexans does have a total of around 20-30 different
machines at the plant, and up to 2000 cable drums. This implies that most of
the cable drums is not in use and not moving; they are stored somewhere for
future use. Some cable drums are stored for days or weeks, and therefore the
need for continuous updates are not needed.
This property can be can be taken advantage of with implementing machine
learning. If conditions at Nexans change, and the range of a beacon (cable
drum) is extended to reach 40 meters, instead of 20 meters, it will be seen by
more reference points. Figure 7.1 shows an example what could happen if the
range of a beacon change without updating the backend, and which reference
points that see the same beacon. If the range within an area was shown to be
the old range of figure 7.1, the beacon (cable drum) would only be seen by
reference point 1 and 2.

80

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

Old range

2
New range

Figure 7.1: Figure showing what happens if the range in an area changes

If the range were to change, because conditions at Nexans change, the beacon
(cable drum) would now be seen by both reference point 1, 2 and 3. The backend
does not at the current moment handle changes in range, and this would have
to be updated manually.
To make this happen automatically machine learning can be implemented. If
the range of a beacon (cable drum) changes, the backend will jump back and
forth between two (or more) reference points, and this will include adding new
entries and deleting old entries. If this happens over a given period of time, for
example 12 or 24 hours, the backend should learn from this. The backend will
learn that the beacon in figure 7.1 can now be seen by all three reference point,
and not just by 1 and 2. This automatic learning will limit the need for manual
updates of the ranges at different locations at the plant.

7.1.5

Battery life of smart devices

The battery life of the reference points at the plant are not necessary to test
because they are always connected to power.

7.1

TESTING

81

The smart devices used by the workers at the plant will notice a drain in battery
life when having their Bluetooth feature turned on. Minor tests have been
performed testing the battery life of a Samsung Galaxy S3 in an environment
which has multiple BLE beacons. Due to the limited testing equipment and use
of an older Android device, the test results are not viable.
Because of this, some tests from Aislelabs are used as a reference for this
section. This section was discussed in the pre-project as well, but it is still just
as relevant in this thesis.
The battery life of a smart phone will depend much on the amount of surrounding BLE beacons broadcasting signals. When there are no BLE beacons around
when Bluetooth is turned on, the battery usage on the smart device is much
lower. At Nexans there will be a considerable amount of BLE devices constantly
emitting a signal, and therefore we can assume that the battery drain on smart
devices will be significant.
Since we only use Android devices in this thesis, we will only focus on these.
Table 7.2 shows the battery drain for three different Android devices. The
Bluetooth radio is ON for the same amount of time on all devices. The Baseline
is where there are no BLE beacons around.

Figure 7.2: A figure shoring the battery drain on different Android devices

The reason why the drain is so different from device to device, is the difference in efficiency of the chip within the phone. When more BLE beacons are
emitting signals, the phone will have to process these signals. Newer smart
devices does this processing more efficient, and therefore they consume less
power. The Nexus 4 is one of the older smart devices supporting BLE, and it
2. http://www.aislelabs.com/
3. http://www.aislelabs.com/reports/ibeacon-battery-phones/

82

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

contains a more inefficient chip. On the Nexus 4 there was a bug in the BLE
implementation done, and therefore Aislelabs could not get any data on 10
beacons and more.
From the table showing battery drain on the different phones, it does show
a significant decrease in battery life when the area contains multiple BLE
beacons. Workers at the Nexans plant would notice a difference in battery life
when using their mobile at work, granted that they have Bluetooth toggled on.
During an 8 hour shift they would notice a decrease in battery life of up to 40
percent more than if they have Bluetooth off.

7.1.6

User testing

The application have on several occasions been tested at Nexans Rognan. These
tests have been performed both with and without the workers and managers at
Nexans. The collaboration with Nexans Rognan have worked well throughout
the duration of this thesis. Challenges could have arisen when working with
a real customer, for example if they have difficulties to acquire hardware for
testing. This have not caused any problems in this thesis. They have always
been able to answer any questions, they have ordered beacons and smart
phones and they have been available for testing the system created.
The tests have been performed with workers from different parts of production.
They have the application installed on their smart phones, and have tested
the web interface. They have proposed different functionalities to be added to
the web site, and these have been taken into account and added to the web
site.
Some challenges during this testing have occurred. Late arriving mobile devices
limited the amount of testing done on the cell-based positioning system. The
functionality of this system does, as far as testing have shown, work as expected,
but it lacks testing in case of any special cases that have not been addressed
yet.
The second challenge is the varying internet connection within the factory hall.
The BLE Reference Point Scanner requires an internet connection for sending
data to the backend. Some areas of the plant have a limited connection, and
will cause an issue for the system. For future work the IT personnel have to
install multiple wireless routers around the factory and storage tents so that
this will not be a problem.

7.2

7.2

CHALLENGES

83

Challenges

The previous section presented challenges with the technologies used. This
section will present any challenges that are important to consider when implementing a system like this.

7.2.1 Security
In any system there are some security risks. This system will store information
about the production at Nexans. By gaining access to the information stored in
the database, the person will have access to know what materials are stored on
each cable drum. There are many materials that can be worth a lot of money, for
example copper. But also finished products can be worth millions of Norwegian
Kroner. By knowing the exact location of a drum, for example storage tent 1, the
outsider can easily find the most valuable drum and steal it. Not all locations
are as thoroughly supervised as others, and by knowing where an exact drum
is located, thieves can be quick at collecting the particular drum.
By sending the data over the internet to Azure, it can be picked up my outsiders
and used for malignant purposes. Web browsers often rely on SSL/TLS to
prevent attackers from obtaining the data, but these security measures was
bypassed in 2009. If an attacker is sniffing any packages on the host, he can
pick up any credentials that pass through. The attacker can then use these
credentials to log on to the site. When the SSL/TLS implementation was
breached in 2009, the users were now able to perform requests to the cloud
systems without these credentials [57]. This is one of the more likely breaches
that Nexans may face, and a reason for considering having the data stored
locally.

7.2.2 Extensibility
The middleware is, as presented before, built upon the principles in ASP.NET
MVC, with controllers, views and models. Each of the features like location of
cable drums specifications, cable drums and device logic is encapsulated into
different controllers to provide an extensible system as well as to decrease the
dependabilities between the parts. Each of the parts can easily be exchanged
or altered without affecting the other functionality significantly.
The modules created are easily exchangeable if specifications and requirements
at Nexans change.

84

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

7.2.3 Cloud computing

The three cloud providers presented here all advertises with a:
Flexible, cost-effective, secure, scalable and elastic and experienced
cloud service.
By having the system locally, they would have to invest in hardware and
software, and resources would have to be set aside to get the system up and
running. By choosing a cloud provider, Nexans will not have any up-front costs
in form of new hardware and software. The costs of the system would come as
a result of how much of the resources at Azure they use. In the long run, using
a cloud provider might provide some extra expenses, but there are several
advantages with outsourcing this job.
As presented, there are no up-front costs with taking an application up in the
cloud. This means that they do not have to estimate the potential resources
needed when investing in hardware and software to be used locally. By using Azure, the system is flexible, and Nexans would only pay for what they
use.
By having the system in the cloud, Nexans does not need to worry about making
the system fault tolerant regarding hardware or software failure, because this
is handled in Azure. If a node falls down, or the data becomes corrupt, this
will be handled by Azure, having multiple nodes storing the same data.

Scalability
One challenge that have not been tested in this thesis is the scalability of the
system. How will the system handle up to 2000 beacons within a relatively
small area? Nexans have not yet decided how many of the drums on the plant
that they should include in the system, but it will atleast be around 500 beacons.
Many of the drums at the plant are of the same kind and they are placed at
one location most of time. These cable drums often only store raw materials
used in production, and they are most likely not needed to be tracked in this
system.
Since this thesis intends to run in the cloud, Azure will handle any necessary
scale ups regarding computational power. To limit some of the computations
done in Azure, the application on the mobile devices can do more of the work.
As they are now they only collect and send data, and the middleware, which
runs in Azure, does all the tests on the data received. The BLE Reference Point
Scanner can for example check the new list of beacons towards the old list and

7.2

CHALLENGES

85

send the changes to the middleware. The changes will be if a new beacon have
entered or left the range of the reference point. There is limited data stored in
the Azure database, and the data stored is only text, so limited storage capacity
is needed in the cloud database.

Who owns the data in the Cloud?

More and more people, both individuals and businesses, are putting their data
in the cloud. Information, ranging from emails, pictures, financial information,
information about cable drums and peoples whereabouts are stored in the
cloud. Who owns this data, and who has access to it?
Data in the cloud is spread out over multiple datacentres that are localized
in multiple countries. Not necessarily in the country where the user is located.
Microsoft, Apple, Verizon, AT&T and Cisc been contacted by the US government,
where the US government have requested to collect data from their storage
locations outside of the US. Microsoft and its affiliates on this matter expresses
that this would harm the international business as well as they do not want to
compromise one countrys laws for another Microsoft and the other companies
have tried to prevent the government from collecting this data. They lost the
first round in court, with the reason that they would hinder the law enforcement
in gathering information that could lead to capturing of criminals.
There have been multiple questions regarding use of cloud computing in
Norway. In 2011 and 2012 Narvik and Moss municipality contacted Datatilsynet
in Norway regarding their use of Google and Microsoft Office 365. Datatilsynet
presented some prerequisites for when deciding to use a cloud. The business
should analyse what the impacts of using the cloud is, and what could possible
go wrong with storing the data in the cloud? The data processor agreement
should be in compliance with Norwegian law. These questions should be
discussed not only in public sector, but in private businesses too.
Meetings with Nexans have been done regarding any issue they might have
with having the data in the cloud. The main conclusion was that as long as
the system is not connected to the production system SAP, it does not contain
any sensitive information that can not be presented for the public. The worst

4. http://www.cbc.ca/news/business/microsoft-and-other-tech-giants-fight-u-s-right-toseize-cloud-data-1.2677688
5. https://www.datatilsynet.no/Teknologi/SkytjenesterCloudComputing/Risikovurdering-og-informasjonssikkerhet/?showContentList=true&showDetailedContentList=false&readMode=fa

86

CHAPTER

E VA L U AT I O N A N D C H A L L E N G E S

case scenario would be an leakage of the database content, this would provide
external parties information about the status and location of every cable drum
at Nexans. Nexans has considered this risk, and decided that since this is not
connected to the SAP system, and therefore does not contain any sensitive
information that would impact their ability to compete, would a leakage not
cause too much harm. For future works, when the system will be included in
SAP, it needs to be taken down from the cloud and ported to a local server.

8
Conclusion and future
works
This chapter consists of two parts. First, future work is presented. Since this is
a prototype of a system there are features that can be enhanced, new features
can be added for a more seamless system and the system can be integrated
with the production system SAP. The second part summarizes the thesis and a
conclusion is drawn from the results of testing, how the different modules of
the system works together and meetings with Nexans, both production workers
and production managers.

8.1

Future works

This section presents different new functionalities that can be added and parts
of the system that have potential to be improved on.

8.1.1

Interoperability

The system implemented in this thesis is a standalone system, not integrated

with the existing system SAP. Future work will include creating an extension in
SAP for the tracking system. It will make it possible to go through SAP, where
87

88

CHAPTER

CONCLUSION AND FUTURE WORKS

all the orders for cable drums are placed, and to this system which only handles
organization about the cable drums.

8.1.2

Beacon testing

Beacon testing have been performed from May 2014 and until May 2015. This
testing have included multiple features regarding the beacons, both range, if
the beacon is water proof and how the beacons handle the harsh environment
they are exposed to.
Testing are still being performed, and more beacons are now placed on cable
drums around Nexans. This testing is regarding battery life, how the beacon
responds to temperature changes, range and robustness.

8.1.3

Mobile applications

The BLE User Scanner is not complete. The user will have to recognize the
MAC-address from the web page when searching for a beacon. The application
will not collect the drum id from the database. For future works this should
be implemented. With performing additional requests to the database, the
application might be less responsive, but it will improve the user experience
significantly.
Users have also started using the applications at Nexans. The users have and will
continue to provide feedback regarding the functionality and user-friendliness
of the application.

8.1.4

Installation at plant

Only a minor area have been used for testing. Future work will install more
reference points (smart phones) around the plant to locate beacons. This installation will also include a longer offline/learning phase for testing range of the
beacons at different locations around the factory area, both inside and outside.
The range of the beacons in different locations will influence the frequency of
reference points and how the calculations with regards to the grid that each
reference point has.

8.1

FUTURE WORKS

89

Map of plant
As of now, the system only gives information on which reference points are near
the cable drum the user is looking for. This implies that the user have to know
where all the reference points are located. Nexans have expressed a desire for
an interactive map which shows the whole plant, the location of the reference
points, and the cable drums plotted on the map. An example of how this map
would look is presented in figure 5.2.
By including this map it will be simpler for the users to navigate to the correct
drum without knowing the name of each reference point in the plant.

8.1.5

Expanding the system

The system implemented in this thesis is a system meant for use at Nexans
Rognan. Future work may include expanding the system to work at other
locations, for example Nexans Halden, Karmy or Namsos. It is also possible
to create a solution which, with few alterations, can be used in many other
industries. For example tracking cars or real-time tracking of inventory in a
department store.
Nexans Rognan has the most potential to improve their efficiency. The other
factories in Norway does mostly produce standard cables, while Nexans Rognan
produces specially made cables. This means that the other factories does not
need such a wide range of different cable drums as Nexans Rognan. The other
factories stock many drums of each type, and therefore have several drums
available of each type and they are more often stacked in clusters including
the same type of drum. Therefore the need for tracking each individual drum
is not needed in the same extent as Nexans Rognan. There can be exceptions
where they order a few cable drums of one type, and in these cases it might
be a reason to adopt the tracking system from Rognan.

8.1.6

Accuracy of location

For future works a more accurate positioning can be implemented. The beacons
are now set to emit power to support an range of approximate 20 meters.
Without disturbances from objects and walls, these beacons would have had a
range of approximately 50 meters. For a more accurate solution, for example
10 meters, the broadcasting power can be set lower to achieve this. This also
have the positive side effect of better expected battery life.
In this thesis only a smaller area at Nexans have been used for testing, and

90

CHAPTER

CONCLUSION AND FUTURE WORKS

it seems that a 20 meter range does work well. Testing performed by several
workers at Nexans have resulted in an overall consensus that a 20 meter range
works well for their use. When the system is implemented throughout the plant,
this might be reconsidered, and a shorter range might be chosen. A longer
range would impact the battery life of the beacons badly, and they would have
the problem of beacons running out of battery too often.

8.2

Conclusion

The prototype of Nexans Tracking System implemented in this thesis does perform as expected. The system implements the specifications provided by Nexans
Rognan, and it solves the goals presented in the project statement.

8.2.1 Beacons
Testing performed in this thesis have shown that, so far, the beacons used for
this purpose does works in the challenging environment at Nexans Rognan.
Because of the size of the StickNFind (SNF) Pro vs. the Estimote beacons, the
SNF Pro did show the most promise. It was expected that the Estimote beacon
would get damaged and broken because of the size of it. This have not yet
happened, and both the SNF beacons and Estimote beacons are still running.
Neither are broken or have run out of battery yet. The only difference between
the two beacons is the price per beacon. The Estimote company sells only three
beacons in a package, and this costs 99 dollars for three beacons. This is a
price of 33 dollars per beacon. The StickNFind providers does sell beacons
in packages of 10, 20 and 30 beacons, with the 10-pack costing 25 dollars per
beacon, the 20-pack costing 24.5 dollars per beacon and the 30-pack costing 24
dollars per beacon. So price-wise, the SNF Pro is considerable cheaper than
the Estimote beacons.

8.2.2 Finished product

This section will present how the system implemented compares to the project
goals presented in chapter chapter 1. It will include what the customer, Nexans
Rognan, thinks about the system implemented, if they find it promising and if
1. http://estimote.com/#jump-to-products
2. https://www.sticknfind.com/store.aspx
3. A query have not been sent to neither Estimote nor SNF to get an offer for if Nexans were
to buy 1500 beacons

8.2

CONCLUSION

91

it is worthwhile pursuing further.

Testing done in collaboration with the workers at Nexans have shown that
the system is intuitive to use and provides enough functionality for the users
to use the system as it is now. The users are able to add, update and delete
information about the drums. The tickets, as shown in figure 2.1, and the excel
sheet used to add remarks about the drum is now redundant. The users can
easily find information about the drum, using the NFC application, and they
can also update the cable drum information when the NFC application redirects
them to the page containing data about the particular drum. The cell-based
positioning technique, using Android devices and BLE beacons, provides a good
overview over the location of each cable drum inside the plant.
If Nexans were to fully adopt this system, and continue to implement the
features presented in future work, they have estimated that they can save 2-3
FTEs each year because the workers no longer need to search for cable drums.
Also, if they at all time have enough cable drums available at all time, and they
are all in good state, they can have lower downtime on the machines they. Each
machine costs a significant amount to buy and install, and if they can have
less downtime, the cost/hour per machine will decrease. This also means that
Nexans Rognan can meet more deadlines, or they can offer shorter deadlines
to their customers.

Appendices

93

A
Layout of factory hall

95

96

APPENDIX A

L AY O U T O F F A C T O R Y H A L L

Figure A.1: Figure showing the layout within the factory. The figure is unclear unless
it is shown in a format larger than A3

B
Database
Listing B.1: Database Schema for the table "Users"

CREATE TABLE [ dbo ] . [ AspNetUsers ] (

[ Id ]
NVARCHAR
[ Email ]
NVARCHAR
[ EmailConfirmed ]
BIT
[ PasswordHash ]
NVARCHAR
[ SecurityStamp ]
NVARCHAR
[ PhoneNumber ]
NVARCHAR
[ PhoneNumberConfirmed ] BIT
[ TwoFactorEnabled ]
BIT
[ LockoutEndDateUtc ]
DATETIME
[ LockoutEnabled ]
BIT
[ AccessFailedCount ]
INT
[ UserName ]
NVARCHAR
[ Firstname ]
NVARCHAR
[ Lastname ]
NVARCHAR
CONSTRAINT [ PK_dbo . AspNetUsers ]
);

97

(128) NOT NULL ,

(256) NULL ,
NOT NULL ,
(MAX) NULL ,
(MAX) NULL ,
(MAX) NULL ,
NOT NULL ,
NOT NULL ,
NULL ,
NOT NULL ,
NOT NULL ,
(256) NOT NULL ,
(MAX) NULL ,
(MAX) NULL ,
PRIMARY KEY CLUSTERED ( [ I d ] ASC)

98

APPENDIX B

D ATA B A S E

Listing B.2: Database Schema for table "User Roles"

CREATE TABLE [ dbo ] . [ AspNetRoles ] (

[ Id ]
NVARCHAR (128) NOT NULL ,
[Name] NVARCHAR (256) NOT NULL ,
CONSTRAINT [ PK_dbo . AspNetRoles ] PRIMARY KEY CLUSTERED ( [ I d ] ASC)
);
Listing B.3: Database Schema for table "CableDrums"

CREATE TABLE [ dbo ] . [ CableDrums ] (

[ SuppliersID ]
NVARCHAR
[ NexansDrumType ]
NVARCHAR
[ NetWeightEmptyDrum ] NVARCHAR
[ MaxLoad ]
NVARCHAR
[ ProductionDate ]
NVARCHAR
[ FlangeDiameter ]
INT
[ OverallWidth ]
INT
[ InsideWidth ]
INT
[ CoreDiameter ]
INT
[ BoreDiameter ]
REAL
[ CableEntry ]
NVARCHAR
[ DrivePinSlit ]
NVARCHAR
[ SpindleSpeed ]
INT
[ TurnoverSpeed ]
INT
[ TopCoatColor ]
INT
[ CableWireDimension ] NVARCHAR
[ MinCableTension ]
INT
[ MinPayloadFullDrum ] INT
[ L a s t Ma i n t en a n c e ]
DATETIME
[ LastMaintenanceWho ] NVARCHAR
[ LastMaintenanceWhat ] NVARCHAR
[ NextMaintenance ]
DATETIME
[ GeneralStatus ]
INT
[ CleanlinessGrade ]
INT
[ RustGrade ]
INT
[ Color ]
NVARCHAR
[ DateOfRemark ]
DATETIME
[Comment]
NVARCHAR
[ BeaconID ]
NVARCHAR
CONSTRAINT [ PK_dbo . CableDrums ]
);

NOT NULL ,
NULL ,
NULL ,
NULL ,
NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
(MAX) NULL ,
(MAX) NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
(MAX) NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
(MAX) NULL ,
(MAX) NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
NOT NULL ,
(MAX) NULL ,
NOT NULL ,
(MAX) NULL ,
(MAX) NULL ,
PRIMARY KEY CLUSTERED ( [ S u p p l i e r s I D ]
(128)
(MAX)
(MAX)
(MAX)
(MAX)

Listing B.4: Database Schema for the table "Localizations"

99

CREATE TABLE [ dbo ] . [ L o c a l i z a t i o n s ] (

[ Id ]
INT
IDENTITY ( 1 , 1) NOT NULL ,
[ DeviceID ] NVARCHAR (MAX) NULL ,
[ BeaconID ] NVARCHAR (MAX) NULL ,
[ Time ]
DATETIME
NOT NULL ,
[Comment] NVARCHAR (MAX) NULL ,
CONSTRAINT [ PK_dbo . L o c a l i z a t i o n s ] PRIMARY KEY CLUSTERED ( [ I d ] ASC)
);
Listing B.5: Database Schema for the table "Devices"

CREATE TABLE [ dbo ] . [ D e v i c e s ] (

[ DeviceID ] NVARCHAR (128) NOT NULL ,
[ GridX ]
INT
NOT NULL ,
[ GridY ]
INT
NOT NULL ,
[Comment] NVARCHAR (MAX) NULL ,
CONSTRAINT [ PK_dbo . D e v i c e s ] PRIMARY KEY CLUSTERED ( [ DeviceID ] ASC)
);

Bibliography
[1] D. Evans, The internet of things, how the next evolution of the internet
is changing everything, 2011. [Online]. Available: http://www.cisco.
com/web/about/ac79/docs/innov/IoT_IBSG_0411FINAL.pdf
[2] J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, Internet of things
(iot): A vision, architectural elements, and future directions, Future
Generation Computer Systems, vol. 29, no. 7, pp. 16451660, 2013.
[3] G. Kortuem, F. Kawsar, D. Fitton, and V. Sundramoorthy, Smart objects
as building blocks for the internet of things, Internet Computing, IEEE,
vol. 14, no. 1, pp. 4451, 2010.
[4] M. Chui, M. Lffler, and R. Roberts, The internet of things, McKinsey
Quarterly, vol. 2, no. 2010, pp. 19, 2010.
[5] P. Mell and T. Grance, The nist definition of cloud computing, 2011.
[6] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. Katz, A. Konwinski,
G. Lee, D. Patterson, A. Rabkin, I. Stoica et al., A view of cloud computing,
Communications of the ACM, vol. 53, no. 4, pp. 5058, 2010.
[7] A. A. T. Hardersen and N. Schirmer, Privacy for cloud storage, in ISSE
2014 Securing Electronic Business Processes: Highlights of the Information
Security Solutions Europe 2014 Conference. Springer, 2014, p. 211.
[8] S. Vaudenay, Security flaws induced by cbc paddingapplications to ssl,
ipsec, wtls... in Advances in CryptologyEUROCRYPT 2002. Springer,
2002, pp. 534545.
[9] apprenda, Iaas, paas, saas (explained and compared), 2015. [Online].
Available: http://apprenda.com/library/paas/iaas-paas-saas-explainedcompared/
[10] G. Lawton, Developing software online with platform-as-a-service tech103

nology, Computer, vol. 41, no. 6, pp. 1315, 2008.

[11] B. Stineman, Why windows azure? February 2011. [Online].
Available: https://brentdacodemonkey.wordpress.com/2011/02/18/whywindows-azure/
[12] Google, Google cloud platform, 2015.
[Online].
Available:
https://cloud.google.com/appengine/?gclid=Cj0KEQjw_
pmoBRDu986bpISz5ZsBEiQANiuHDDDMO8meZpwnoLNie3V2vLekq75IcDeTRsQqbmlsP0aAvHp8P8HAQ
[13] R. Prodan and S. Ostermann, A survey and taxonomy of infrastructure
as a service and web hosting cloud providers, in Grid Computing, 2009
10th IEEE/ACM International Conference on. IEEE, 2009, pp. 1725.
[14] B. Calder, J. Wang, A. Ogus, N. Nilakantan, A. Skjolsvold, S. McKelvie, Y. Xu,
S. Srivastav, J. Wu, H. Simitci et al., Windows azure storage: a highly
available cloud storage service with strong consistency, pp. 143157, 2011.
[15] ScriptRock, Aws vs azure, June 2014. [Online]. Available: http:
//www.scriptrock.com/articles/aws-vs-azure
[16] A. E. C. Cloud, Amazon web services, Retrieved November, vol. 9, p. 2011,
2011.
[17] S. Hazelhurst, Scientific computing using virtual high-performance computing: a case study using the amazon elastic computing cloud, in
Proceedings of the 2008 annual research conference of the South African Institute of Computer Scientists and Information Technologists on IT research
in developing countries: riding the wave of technology. ACM, 2008, pp.
94103.
[18] E. Ciurana, Google app engine, Developing with Google App Engine, pp.
110, 2009.
[19] P. Wayner, Amazon vs. google vs. windows azure: Cloud
computing speed showdown, March 2014. [Online]. Available: http://www.computerworld.com.au/article/539633/amazon_vs_
google_vs_windows_azure_cloud_computing_speed_showdown/
[20] W. V. Winkle, Microsoft azure vs. amazon web services: Cloud
comparison, January 2015. [Online]. Available: http://www.tomsitpro.
com/articles/azure-vs-aws-cloud-comparison,2-870.html

[21] K. Shvachko, H. Kuang, S. Radia, and R. Chansler, The hadoop distributed

file system, in Mass Storage Systems and Technologies (MSST), 2010 IEEE
26th Symposium on. IEEE, 2010, pp. 110.
[22] M. Rouse, fog computing (fogging), March 2014. [Online]. Available:
http://whatis.techtarget.com/definition/fog-computing-fogging
[23] F. Bonomi, R. Milito, J. Zhu, and S. Addepalli, Fog computing and its
role in the internet of things, pp. 1316, 2012.
[24] C. reseach, Fog computing, ecosystem, architecture and applications,
2013. [Online]. Available: http://www.cisco.com/web/about/ac50/
ac207/crc_new/university/RFP/rfp13078.html
[25] H. Liu, H. Darabi, P. Banerjee, and J. Liu, Survey of wireless indoor
positioning techniques and systems, Systems, Man, and Cybernetics, Part
C: Applications and Reviews, IEEE Transactions on, vol. 37, no. 6, pp. 1067
1080, 2007.
[26] Y. Luo, O. Hoeber, and Y. Chen, Enhancing wi-fi fingerprinting for indoor
positioning using human-centric collaborative feedback, Human-Centric
Computing and Information Sciences, vol. 3, no. 1, pp. 123, 2013.
[27] S. J. Julier and J. K. Uhlmann, New extension of the kalman filter to
nonlinear systems, in AeroSense97. International Society for Optics
and Photonics, 1997, pp. 182193.
[28] T. Higuchi, Monte carlo filter using the genetic algorithm operators,
Journal of Statistical Computation and Simulation, vol. 59, no. 1, pp. 123,
1997.
[29] A. Doucet, N. De Freitas, and N. Gordon, An introduction to sequential
Monte Carlo methods. Springer, 2001.
[30] J. J. Leonard and H. F. Durrant-Whyte, Mobile robot localization by
tracking geometric beacons, Robotics and Automation, IEEE Transactions
on, vol. 7, no. 3, pp. 376382, 1991.
[31] J. Haartsen, Bluetooth-the universal radio interface for ad hoc, wireless
connectivity, Ericsson review, vol. 3, no. 1, pp. 110117, 1998.
[32] S. Hegenderfer, Everything you always wanted to know about beacons,
April 2014.

[33] P. Murphy, E. Welsh, and J. P. Frantz, Using bluetooth for short-term

ad hoc connections between moving vehicles: A feasibility study, in
Vehicular Technology Conference, 2002. VTC Spring 2002. IEEE 55th, vol. 1.
IEEE, 2002, pp. 414418.
[34] C. Bisdikian et al., An overview of the bluetooth wireless technology,
IEEE Commun Mag, vol. 39, no. 12, pp. 8694, 2001.
[35] P. McDermott-Wells, What is bluetooth? Potentials, IEEE, vol. 23, no. 5,
pp. 3335, 2004.
[36] C. Gomez, J. Oller, and J. Paradells, Overview and evaluation of bluetooth
low energy: An emerging low-power wireless technology, Sensors, vol. 12,
no. 9, pp. 11 73411 753, 2012.
[37] J.-S. Lee, Y.-W. Su, and C.-C. Shen, A comparative study of wireless
protocols: Bluetooth, uwb, zigbee, and wi-fi, in Industrial Electronics
Society, 2007. IECON 2007. 33rd Annual Conference of the IEEE. IEEE,
2007, pp. 4651.
[38] A. Andersen, R. Karlsen, and A. Munch-Ellingsen, Nfc provided user
friendliness for technologically advanced services, in Human Interface
and the Management of Information. Information and Interaction for Health,
Safety, Mobility and Complex Environments. Springer, 2013, pp. 337346.
[39] K. Curran, A. Millar, and C. Mc Garvey, Near field communication, International Journal of Electrical and Computer Engineering (IJECE), vol. 2,
no. 3, pp. 371382, 2012.
[40] A. H. Omre and S. Keeping, Bluetooth low energy: wireless connectivity
for medical monitoring, Journal of diabetes science and technology, vol. 4,
no. 2, pp. 457463, 2010.
[41] J. Decuir, Bluetooth 4.0: Low energy, Cambridge, UK: Cambridge Silicon
Radio SR plc, 2010.
[42] Y. Masumoto, Global positioning system, May 11 1993, uS Patent
5,210,540.
[43] R. Alfred Kleusberg, The limitations of gps, March 1990.
[44] Y. Zhou and Y. Fang, A location-based naming mechanism for securing
sensor networks, Wireless Communications and Mobile Computing, vol. 6,
no. 3, pp. 347355, 2006.

[45] S. C. Alliance, Bluetooth low energy (ble) 101: A technology primer with
example use cases, June 2014.
[46] P. Kinney et al., Zigbee technology: Wireless control that simply works,
in Communications design conference, vol. 2, 2003, pp. 17.
[47] V. B. Misic, J. Fang, and J. Misic, Mac layer security of 802.15. 4-compliant
networks, in Mobile Adhoc and Sensor Systems Conference, 2005. IEEE
International Conference on. IEEE, 2005, pp. 8pp.
[48] S. Woo, S. Jeong, E. Mok, L. Xia, C. Choi, M. Pyeon, and J. Heo, Application
of wifi-based indoor positioning system for labor tracking at construction
sites: A case study in guangzhou mtr, Automation in Construction, vol. 20,
no. 1, pp. 313, 2011.
[49] A. H. Lashkari, M. M. S. Danesh, and B. Samadi, A survey on wireless
security protocols (wep, wpa and wpa2/802.11 i), in Computer Science
and Information Technology, 2009. ICCSIT 2009. 2nd IEEE International
Conference on. IEEE, 2009, pp. 4852.
[50] L. M. Ni, Y. Liu, Y. C. Lau, and A. P. Patil, Landmarc: indoor location
sensing using active rfid, Wireless networks, vol. 10, no. 6, pp. 701710,
2004.
[51] R. Want, An introduction to rfid technology, Pervasive Computing, IEEE,
vol. 5, no. 1, pp. 2533, 2006.
[52] J. Hightower, R. Want, and G. Borriello, Spoton: An indoor 3d location
sensing technology based on rf signal strength, UW CSE 00-02-02, University of Washington, Department of Computer Science and Engineering,
Seattle, WA, vol. 1, 2000.
[53] R. Want, A. Hopper, V. Falcao, and J. Gibbons, The active badge location
system, ACM Transactions on Information Systems (TOIS), vol. 10, no. 1,
pp. 91102, 1992.
[54] P. Bahl and V. N. Padmanabhan, Radar: An in-building rf-based user location and tracking system, in INFOCOM 2000. Nineteenth Annual Joint
Conference of the IEEE Computer and Communications Societies. Proceedings. IEEE, vol. 2. Ieee, 2000, pp. 775784.
[55] R.-H. Jan, H.-C. Chu, and Y.-F. Lee, Improving the accuracy of cell-based
positioning for wireless networks, Computer networks, vol. 46, no. 6, pp.
817827, 2004.

[56] T. D. Pranav Rastogi, Rick Anderson and J. Galloway, Introduction to

asp.net identity, October 2013. [Online]. Available: http://www.asp.net/
identity/overview/getting-started/introduction-to-aspnet-identity
[57] D. Jamil and H. Zaki, Security issues in cloud computing and countermeasures, International Journal of Engineering Science and Technology
(IJEST), vol. 3, no. 4, pp. 26722676, 2011.