Chapter 3 – Requirements - School of Computing & Intelligent ...

rabidwestvirginiaNetworking and Communications

Oct 26, 2013 (3 years and 5 months ago)

197 views





RFID Lone
Worker
Tracking

University of Ulster.
BEng Electronics and
Computing.
Final Year Project.

December 2011.

Fergal Canning

1


Abstract.


RFID has many applications in everyday life. We use RFID systems

when
we purchase goods in a shop, present a smart card at a secure facility or
present a key to the ignition in a modern car.

I
n business RFID is used for
instance, to monitor stock levels, track equipment movement or monitor gas
emissions. The introduct
ion of an RFID system into a factory
has allowed
production to track material throughout the facility. This material tracking
system utilise
s

the existing WiFi network to track the lots by physical location
rather than the supposed system that the process

build demands.

There are
times when lots go missing and as a result the processing of WIP is delayed
or worst case scenario the material is scrapped off costing thousands of
dollars to the company and putting schedules at risk.

With a fully operational
sy
stem it is hoped to be able to reduce time spent looking for material and
thus reduce idle time on machines. When material is lost for 7 days it is
scrapped off the system. This scrap can be valued at millions of dollars
annually. With the assistance of

RFID it is hoped that scrap material levels
can be reduced by being able to locate lots by their physical location.

A fully
deployed system allows for opportunities to expand its uses.



The reasons for tracking a subject
are

varied
on
e

reason is
health
and
safety. There are times when a subject would work in a remote environment
and if they stop for an extended period of time they may be in difficulty and in
need of assistance. Employers face a difficult task of checking on the status
of employees in t
his kind of environment and understand that a remote
worker may not be able to reach a phone in the event of an accident.

The
aim of this project
wa
s
also
to devise a system that w
ould

track the
movements of a subject

alongside a primary
lot tracking syste
m. The overall
goal
wa
s to deploy an easily manageable, reliable and scalable system
.





2


Acknowledgements.


I would like to thank my project supervisor Dr. Kevin Curran for his help and
support in the preparation and completion of this project.

I would a
lso like to
thank my employer Seagate Technology for allowing me the considerable
amount of time over the years to continue my studies.


Finally I would like to thank my family for their support over the last 9 years of
study even when the balance between
work life, family life and student life
was sometimes too much to bear, they kept me on the path to where I am
now.


3


Declaration


I declare that this is all my own work and does not contain unreferenced
material

copied from any other source.
I have read t
he university’s policy on
plagiarism
and

understand the definition of plagiarism. If it is shown that

material has been plagiarised, or I have otherwise attempted to obtain an
unfair advantage for myself or others, I understand that I may face sanctions
i
n accordance with the policies and procedures of the university. A mark of
zero may be awarded and the reason for that mark will be recorded on my
file.








(Fergal Canning)



4


Contents

Abstract.

................................
................................
................................
.......................

1

Acknowledgements.

................................
................................
................................
.....

2

Declaration

................................
................................
................................
...................

3

Table of Figures

................................
................................
................................
...........

6

Acronyms
................................
................................
................................
......................

8

Chapter 1


Introduction
................................
................................
.............................

10

Chapter 2
-

Literature Review

................................
................................
....................

13

2.1 Commercial location Systems

................................
................................
........

18

2.2 Tracking Personnel

................................
................................
.........................

19

2.3 Chemical requirements

................................
................................
....................

20

2.4 Conclusion
................................
................................
................................
........

21

Chapter 3


Requirements

................................
................................
.........................

22

3.1

Current Issues

................................
................................
...............................

28

Chapter 4


Specifications

................................
................................
.........................

31

Chapter 5


Planning

................................
................................
................................
.

34

5.1 Project Outputs

................................
................................
................................

34

5.2 Project Plan

................................
................................
................................
......

37

5.3 Time Man
agement

................................
................................
...........................

40

5.4 Working with Project Supervisor

................................
................................
......

41

5.5 Risk Management

................................
................................
............................

42

Chapter
6
-

Design

................................
................................
................................
.....

44

Chapter 7


Implementation
................................
................................
.......................

57

Chapter 8


Evaluation
................................
................................
...............................

72

How has the system changed how the factory functions, if at all?

...............

7
2

What were the practices on the floor prior to implementation?

....................

73

Who has benefited?

................................
................................
.............................

76

What benefits have they seen?

................................
................................
..........

78

What have been the downsides to deploying such as system?

....................

79

How much of the hidden factory has been revealed?

................................
.....

80

What has the system vendor learned about the way Seagate has deployed
their system into a pre
-
existing environment?

................................
................

82

Chapter 9


Conclusion
................................
................................
..............................

83

Appendix A

................................
................................
................................
.................

86

5


Exciter Configurat
ion

................................
................................
..............................

86

Tag Configuration
................................
................................
................................
...

88

MobileView Event Configuration

................................
................................
............

88

Appendix B

................................
................................
................................
.................

90




6


Table of Figures

Figure
1
-
Wireless Access Point Infrastructure
................................
...................

12

Figure 2
-

Passive RFID Tag System (Powell & Shim, 2009)

........................

14

Figure 3
-

Door Access Scan Point

................................
................................
.....

14

Figure 4
-

Passive RFID tag
................................
................................
..................

14

Figure 5


European RFID Frequency Table Excerpt

................................
......

15

Figure 6
-

Acti ve RFID tag
................................
................................
.....................

16

Figure 7
-
Active tag operation

................................
................................
..............

16

Figure 8
-

RFID Hospital Uses
................................
................................
..............

20

Figure 9
-

MobileView tab in FactoryWorks Client

................................
............

24

Figure 10
-
RFID TagID incorporated i nto Lot At
tributes
................................
....

25

Figure 11
-

Asset as seen on MobileView client
................................
................

25

Figure
12
-

Asset location on map of office

................................
........................

25

Figure 13
-

Systems Architecture
................................
................................
.........

26

Figure 14
-

Rack showi ng material for processing

................................
............

27

Figure 15
-

Rack Mounted Exciter

................................
................................
.......

27

Figure 16
-

Acti ve RFID Tag

................................
................................
.................

27

Figure 17
-

Choke Point Exciter

................................
................................
...........

27

Figure 18
-

Rack Mounted Exciter in Place

................................
........................

28

Figure 19
-

RFID Tag Activator

................................
................................
............

28

Figure 20
-

Material Box with RFID tag fi xed to lid.
................................
...........

31

Figure 21
-

Tag Cases

................................
................................
...........................

31

Figure 22
-
Cisco wireles AP
................................
................................
..................

31

Figure 23
-

Racki ng Diagram
................................
................................
................

31

Figure 24
-

Catalyst 3560X POE Switch(courtesy Cisco.com)

.......................

33

Figure 25
-

Work breakdown

................................
................................
................

34

Figure 26
-

Alert Configuration Page, MobileView

................................
............

35

Figure 27
-
Project Plan

................................
................................
...........................

38

Figure 28
-
Time breakdown sample

................................
................................
.....

41

Figure 29
-

Process plan showi n
g Fab changes 1 to 2 on left

........................

44

Figure 30
-

WIP waiting showi ng In Transit between stages
...........................

45

Figure 31
-

Lot properties showing tag id and MobileView link

.......................

45

Figure 32
-

Association between T
ag and Lot

................................
...................

46

Figure 33
-

Lots showing in the dispatch list

................................
......................

46

Figure
34
-

The make tag blink now activates when lot moves to the
'Selected Lots' column

................................
................................
...........................

47

Figure 35
-

Custom bulk 'Make Tag Bli nk' tab created in FW client

...............

47

Figure 36
-

Exciter configuration where range etc can be set.
........................

48

F
igure 37
-

Rack Setup 1 exciter and 2 exciters

................................
...............

49

Figure 38
-

Raw CAD map of one the fab areas

................................
...............

49

Figure 39
-

Raw CAD output
................................
................................
.................

50

Figure 40
-

3D factory image

................................
................................
................

50

Figure 41
-

Screenshot of the initial Fab survey

................................
................

51

Figure 42
-

Sample of the Oracle database schema

................................
........

52

Figure 43
-

Sample of the Exciter table
................................
...............................

53

Figure 44
-

Sample of dat
a from the asset table

................................
...............

53

Figure 45
-

Link between tag and lot in the system

................................
..........

54

Figure 46
-

Data modeller relationship diagram

................................
................

54

Figure 47
-

Location history for a tag
................................
................................
....

55

7


Figure 48
-

Clamshell for single wafer movement

................................
.............

57

Figure 49
-

Single wafer lot boxes, no tag

................................
..........................

57

Figure 50
-

Start of wafer bu
ild

................................
................................
.............

58

Figure 51
-

Pathfi nder showi ng stage and step

................................
................

58

Figure 52
-

Lot boxes with a tag on

................................
................................
.....

59

Figure 53
-

Operator dispatchi ng WIP
................................
................................
.

59

Figure 54
-

Eligible WIP gui

................................
................................
..................

60

Figure 55
-

Rack in fab with copex and wiri ng ready to connect to

network
.

60

Figure 56
-

Floor box
................................
................................
..............................

60

Figure 57
-

Network points i n flo
or box

................................
...............................

61

Figure 58
-

New rack layout i n CAD format
................................
........................

61

Figure 59
-

Han
d drawn diagrams of how rack wiring will be done

................

61

Figure 60
-

Master, Slave, Slave, Floater setup

................................
................

62

Figure 61
-

Floater rack wiri ng setup

................................
................................
...

62

Figure 62
-

Hold area fab1, under floor wiri ng

................................
...................

63

Figure 63
-

The splitter close up
................................
................................
...........

63

Figure 64
-

The design in place on the
test rack

................................
...............

64

Figure 65
-

Internal wiri ngs for new design

................................
........................

64

Figure 66

-

Rack in place i n fab

................................
................................
...........

64

Figure 67
-

Cables connected to permanent rack

................................
.............

64

Figure 68
-

Custom machined exciter bracket

................................
...................

65

Figure 69
-

Choke and exit points

................................
................................
........

66

Figure 70
-

Alert is raised in the system when a tag passes the exit choke
point

................................
................................
................................
..........................

66

Figure 71
-

An email that is sent out
................................
................................
....

67

Figure 72
-

Email configuration screen

................................
...............................

67

Figure 73
-

Monitor and lot box i n close proximity
................................
.............

68

Figure 74
-

Exciter on top shelf

................................
................................
............

68

Figure 75
-

Level of spread of the tag bouncing around off APs

....................

68

Figure 76
-

Position for Tag 00
0CCC51EAC2

................................
...................

69

Figure 77
-

4 days movement data for one tag i n one fab

...............................

70

Figure 78
-

Lot has completed its previous step
................................
................

74

Figure 79
-

Lot now goes on HOLD.
................................
................................
....

74

Figure 80
-

R&D operator checks why it has gone on HOLD
..........................

74

Figure 81
-

Operator checks its location in the factory

................................
.....

75

Figure 82
-

He can check the losts previous locations for a defined period.

75

Figure 84
-

R
&D Office WIP
................................
................................
..................

78

Figure 85
-

Mobileview realtime tracking mode

................................
.................

79

Fig
ure 86
-

ToolView showing tools IDLE
-
NO_WIP

................................
.........

81

Figure 87
-

Interfab trolley

................................
................................
.....................

81

Figure 88
-

Tagged lots on trolley

................................
................................
........

81

Figure 89
-

Factory Scoreboard 60inch monitor i n each room
........................

82

Figure 90
-
Alert Configuration GUI

................................
................................
......

89




8


Acronyms

RFID
-

Radio Frequency Identification

OCR
-

Optical Character Recognition

MES
-

Manufacturing Execution System

AMA
-

Activity Manager for Automation (copyright Applied
Materials)

AIDC
-

Automatic Identification and Data Capture devices

HR


Human Resources

RTLS


Real Time Location System

GUI


Graphical Front End

LBT


Listen Before Talk

ERP


Effective Radiated Power

COO


Cost Of Ownership

ROI


Return On Investment

R&D


Research and Development

POE


Power Over Ethernet

AP


Access Point

KPIV


Key Process Input Variable

RHO


Recording Heads Operation

WIP


Work In Process

FW


FactoryWorks (copyright Applied Materials)

DL
-

Direct Labour

OOE


Overall Equipment Ef
ficiency

TTM


Time To Market

DGR


Daily Going Rate

CT


Cycle Time

RD (R&D)


Research and Development

FY


Fiscal Year

MDA


Material Deviation Authorisation

IP


Intellectual Property

CAD


Computer Aided Design


9


FMEA


Failure Mode AND Effects Analysis



10


Chapter 1



Introduction

In the world of industry there are a number of reasons why any employer
should want to the whereabouts of their employees. One such reason, and
the subject of this paper, is for health and safety monitoring.

When we
consider th
e number of options that are available today that can facilitate in
the tracking of someone or something it becomes a matter of which of the
options can best fulfil the needs of the business.

GPS tracking can be a
useful method of tracking someone but this

is really only useful in an
outdoors system

where satellite visibility is critical to the systems success
.


Bluetooth can be used to track people or assets in an indoor environment up
with accuracy up to a range of 2
-
3 metres. This can be accomplished wi
th
either Bluetooth tags, mobile phone or PDAs that are Bluetooth enabled

(Tadlys Product list, 2004)
.
RFID passive tags are useful low frequency
devices and relatively cheap but the range is limited, less than 1.5 metres.

RFID

active tags use high frequencies and have very good range. These
tags have the ability to interact with the reader. They are more expensive
than passive tags.


ZigBee is a system that uses the mesh network system of communication
with each device intera
cting with every other device within a certain range.

The US military has just begun implementing a ZigBee system utilising the
IEEE 802.15.4 standard in order to track the logistics of returning military
equipment back to the US from the Gulf

(Swedberg, 2011)
.


Mobile phone networks are also used to track location. Each time a device is
turned on the phone must connect to the cellular network. The cellular
network is made up of numerous masts that will overlap their bro
adcast
range. A device will normally connect to the mast with the strongest signal
and only pass to another mast when a preset handoff level is reached.

The
triangulation of the cellular network masts can help pinpoint the location of a
mobile phone

(Google Latitude, 2011)
.


11


When one mentions the word ‘industry’ we normally conjure up images of a
factory floor with machinery and maybe a production line. This is only one
aspect of industry. To compliment the operations on the

factory floor we will
need additional personnel in stores, offices, canteen

and

cleaning. All of
these ancillary staff can be classed as indirect labour, i.e. do not directly
handle the goods on the production line but every bit as important. One of
the
se ancillary jobs could require a lone person having to work in a remote
environment such as plant engineering. In the event that the plant personnel
have to conduct repairs in a remote area of the facility it would be desirable
for two or more colleagues

to be in attendance, this is called ‘the buddy
system’. Unfortunately restrictions on resources do not always allow the
buddy system to be put into practice. In 2010 the number of accidents in the
work place that can be attributed to lone person working

is in excess of
14000
(Lone Worker)
. The overwhelming desire to have the additional
resources available to assist in these circumstances is countered by the
desire to reduce the overall cost of manufacturing.


W
here does this

leave the unfortunate employer that must balance cost of
manufacture versus the health and safety of the employees? Should
t
he

employer

i
ncrease the workforce to allow for a buddy system and face the
threat of cheaper foreign competition or

i
nvest in new technology and try to
streamline the production system?


In an ideal world you would choose

to

increase the workforce to allow for a
buddy system
and provide some gainful

employment where it

i
s needed,
however
in the current economic environ
ment
, this will not happen.



Therefore, i
t is intended to select a device that will provide some feedback to
a system and give the end user a location based on triangulation updated at
a predefined time. It is hoped
that
the system will use the followin
g
components to provide this information.

Some form of active tag to relay
signal s
trength triangulated from at le
ast three wireless AP's, a datab
a
se
back end to store informa
tion, a
GUI

front end, a server

and the system
software itself.

12




Figure
1
-
Wireless Access Point Infrastructure

At Seagate there are a number of reasons why an RFID system has been
proposed. RFID implementation in Seagate Technology

is an integral part of
the RHO Wafer Factory Control strategy and an
evolutionary improvement in
RHO WIP tracking systems
, which to date has involved a paper tracking
system. RFID

will enable Time to Market speed of development, increase
factory efficiency and bring Seagate in line with

Best in Class


Semiconductor WIP tr
acking systems.


The steps taken to research such a vast project have included forming
mutually exclusive
IP
partnerships with other semi
-
conductor manufacturers
and exchanging information regarding the benefits of such a system. One
such manufacturer is
Freescale. Their return on investment was used as a
prime example of how RFID has contributed to increasing
productivity and
shorter cycle time.







13


Chapter

2

-

Literature Review


Location detection systems have been in existence for a long time in one
form or another. From smoke signals to sonar ping, radar is just one of the
location detection systems that we and use today. Originally developed in
order to identify friendly ai
rcraft during WWII, radar is one of the earliest
location detection systems that
are

still in use today.


Radar stands for radio detection and ranging and works on the principle of
reflected signal. A radio signal is generated and transmitted, the signal
will
bounce off a reflected object, and the receiver being close to the transmitter
will receive, amplify and process the signal

(Simon Kingsley, 1999)
. In this
regard our modern day RFID system only has minimal similarities t
o the
original 1940’s design. RFID is one of the families of devices known as
Automatic Identification and Data Capture devices (AIDC). Others in this
field include low tech devices such as paper tagging, optical character
recognition and Infrared
(V. Daniel Hunt, 2007)
.

RFID has its origins in the
1940's when it was used by the allied forces to identify incoming aircraft as
either friend or foe. All allied aircraft carried a transponder that carried a
unique code that identi
fied them as friendly. This identification was called
"Identify Friend or Foe", IFF. Modern day aircraft still use the same method
for identification, passing a unique code to air traffic controllers known as
"squawk"

(Shepard, 2
005)
.


The first US patent for an active RFID tag with rewritable memory was
secured by Mario Cardullo in 1973.

A California entrepreneur, Charles
Walton, also patented a passive transponder that would be used to unlock a
door without the use of a key

(Brown, 2006)
.

The Los Alamos National
Laboratory had developed a number of applications that make use of RFID
technology including cow tracking.

One application in which passive RFID
tag were used was for the transportation

nuclear material from the many
secure facilities. The truck had an RFID transponder on the side and the
reader was placed on the gate of the facility.

Passive tags work on a
14


proximity basis. The tag must be within a certain range of the reader before
it can be detected.



Figure
2

-

Passive RFID Tag System

(Powell &

Shim,


2009
)


The many applications for this type of system include biometric passport
system where data about the passport holder is embedded into the cover of
the passport and read by passport control upon presentation at their card
readers. Passive RFID is also
use
in

building access control

(
Figure 3
).








Figure
3

-

Door Access Scan
Point

Figure
4

-

Passive RFID tag



Most major companies use an RFID enabled employee identification badge
to allow access to b
oth the main access points to their buildings but also to
control access within buildings, HR rooms for instance and where business
sensitive information is held. Passive tags
(
Figure 4
)

do not have a battery
and will lie dormant until they are within the range of the tag reader. When in
15


range the reader excites the tag and the data on the tag is passed to the
reader. The passive tag is not capable of passing information about it
whereab
outs without the assistance of the reader.

The main limitation with
passive RFID is its lack of distance. Passive RFID has limits that can be as
far as a few feet such as in store door ways to prevent the theft of goods, or
as little as a few centimetres,

as in id card readers.

The passive tag operates
a low frequency. In Europe, and indeed almost the entire world, this
frequency is in the range, 125
-
134 Khz

(Lahiri, 2005)

(figure 5)
.

As passive
tags work on low frequency it i
s deemed very safe and works well in
environments where metals, liquids and dirt are present.


Region

LF

HF

UHF

Microwave

Europe

125
-
134Khz

13.56Mhz

865
-
865.5Mhz,0.1 watts ERP,
Listen Before Talk(LBT). 865.6
-
867.6Mhz, 2 watts ERP, LBT.
876.6
-
868Mhz, 0.5
watts ERP,
LBT

2.5Ghz


Figure
5



European RFID Frequency Table Excerpt


The operating f
requency is
the most important attribute of the RFID system.

Now we the limitations of the passive RFID tag we can now look at the
alternative to it, the active tag.

Active RFID tags

(
Figure
6
)

have a far ranging
and varied array of uses. Active tags can hold and transmit more than just
their location. They can also report back environmental data such as
temper
ature and humidity. They can report back velocity so they can be
attached to a fragile device and if the device falls to the ground it can report
back to the system with an alert to say an item has been dropped. Active
tags can also work in liquid so the
y can be used to track artefacts and
underwater equipment. As it’s a lot more complex than the passive tag and
has the ability to interact with a wireless system it requires some battery
power.


16



Figure
6

-

Active RFID tag


In th
e mid
20
00’s (circa 2006) the estimate market for active RFID RTLS use
for 2010 was approximated at 1.6 billion

(Bing, 2008)
.

The advances in
technology and the associated leaps in scale have allowed the size of an
RFID tag to
be greatly reduced and with that also the size of the battery
required to feed the needs of the tag. Unlike the passive tag the active tag
gets its power from its internal battery. Depending on the system setup this
battery may last for up to 5 years bef
ore it needs to be replaced. This time
frame is dependent on the number of times it ‘checks in’ with the system to
report its location, which is after all the purpose of the active tag. If a tag
moves then it must update the system using its wireless sig
nal, this
consumes power.



Figure
7

-
Active tag
operation


As the need for tags increases, for instance in a warehouse environment, so
then does the number tags that are reporting back their position. In a
heavy
populated
wirele
ss environment the amou
nt of traffic that is present at any
point in time can be quite high. With such a high volume of traffic how will
the system cope?
A warehouse may have thousands of tags active at any
17


one time but the system can only communicate
with one at a time. Although
the system may appear to be talking with more than one at a time it is not
actually doing this. Should two or more tags attempt to make contact with a
reader at any one time this scenario is called ‘tag collision’.

When this

event
occurs the reader communicates with conflicting tags w
ith what’s called a
‘Singulation P
rotocol’. The specific algorithm used to undertake the task is
called an ‘
Anti
-
C
ollision’ algorithm. These are ALOHA when HF is used and
Tree Walking for UHF s
ystems

(Lahiri, 2005)



There have been many other milita
ry and commercial uses for RFID. Just
some of these applications include tracking equipment for use on fire and
rescue vehicles. When a vehicle is to be readied for use

in a particular type
of rescue a
n RFID system is used to

identify all equipment on the vehicle
and the location of equipment that needs to be acquired

(Lockwood
Technology Corporation, 2011)
. In hospitals the location of defi
brillators and
other specialist equipment can be determined very quickly.
It is the case that
sometimes equipment can be hidden until it is required means that only a
select few may know where it is

(Jakki J. Mohr, 2009)
.
In
some military
situations the use of RFID

by suppliers has reduced discrepancies in
packaging. As well as reducing the packaging errors the military receivers
can track receipts better resulting in faster turnaround in payments

(Swedberg, 2011)
.



There are many more examples of this type of use in military and commercial
use, t
oo many and varied in
their
use
s

to list in this one paper.

RFID uses a
range of components in their construction. The basis of

the active RFID t
ag

design consists of
a chip, an antenna and a battery. The system must also
contain a receiver, a transmitter, a database and a middle layer to interpret
the data and report certain details.

The antenna is the physical entity that
intercepts and converts

the electro
-
magnetic energy into free space electro
-
magnetic waves. The antenna is vital to system success. Depending on the
layout of the system and configuration it may be necessary to have more
than one antenna per reader. The antenna will have prop
erties that
influence its efficiency such as polarisation, gain, beamwidth and direction.
18


Where more than one antenna is required it would be appropriate to angle
the antenna to optimise area coverage

(Karmakar, 2010)
.


2
.1
Commercial location Systems


Some of the leading vendors of RFID systems are Aeroscout, Trapeze,
Placelab, Ubisense

and
Ekahau
.

Aeroscout

Aeroscout is one of the leading manufacturers of RTLS systems. The
company invented the first WiFi based RFID tag.

The company utilise
standard WiFi networks to track the location, condition and status of assets
both personnel and equipment. The use of standard WiFi networks helps to
reduce installation costs. Aeroscout has attracted many Fortune 500
companies with

its use of standard WiFi system meaning a lower overall
COO and a higher ROI

(Aeroscout)
.

Juniper Networks(formerly Trapeze)

Juniper is one of the leading providers of wireless infrastructures and boast
96 of the Global Fortun
e 100 companies among their clients. Juniper
networks offer secure, scalable and open network solutions. This company
can provide routers, switches and complete infrastructure design and
implementation solutions

(Juniper Networks
)
.

PlaceLab

Intel Research Labs provides location determination solutions for use with
location based software applications. One of these solutions is PlaceLab.
Place Lab is designed to be used both indoors and outdoors. The system
allows wireless
devices such as notebooks, Smart Phones and PDA’s to
locate themselves by listening to wireless signals such as those from fixed
bluetooth devices, cellular network towers and wireless 802.11x access
points

(Intel Labs)

UbiSens
e

19


Similar to other RTLS applications, Ubisense uses UWB to locate assets.
Ubisense uses time of arrival and angle of arrival to determine the location of
the tracked asset. This angle of arrival is unique to Ubisense and in certain
cases can be a big saf
ety bonus. Ubisense claim accuracy of 1 metre or
better in certain environments

(Ubisense)

Ekahau

Provides the world’s first visual WiFi site survey tool. Ekahau also claims to
be the first commercially available WiFi enabled

RTLS
(Ekahau)

Ekahau
claim to offer VOIP, PDA, Mobile Phone, WiFi tags and other 802.11 devices

(Ekahau)

Unlike ZigBee devices Ekahau can initiate two
-
way data
communication and send SMS messages

(Ekahau)


2
.
2


Tracking
Personnel


Personnel tracking is

just one of the uses that RFID ha
s being used for over
the last number of years.

From hospital patient tracking/monitoring to miner
tracking it has proved a very versatile resource.

In hospital situations the
RFID tag can be used to monitor patient movement
.

This can be done w
hile
waiting for surgery,
in a ward, in a

waiting room or

outside having a cigarette.
When a patient enters the hospital all their immediate information such as
allergies, current medication etc can be recorded
.


W
hen it comes to
medication the dispensing nurse can select the electronic record
to

determine if t
he patient has already had their daily medication or not.


One valuable use of RFID is its historical records. If

at some point someone
enters the hospital or a ward
, which

is subsequently found to have had a
communicable disease
,

the location records c
an be accessed and all
patients that may have been in the vicinity can be traced

(Banks, 2007)
.


20



Figure
8

-

RFID Hospital Uses


In the mining industry
the use of RFID tags has increased in
popularity over
the last few years.

In conditions where a maze of shafts and seams run
underground the possibility of getting lost or an incident arising can happen
very easily. The use of RFID has enabled rescuers to reduce the search
time and therefore

bring casualties to the surface quicker and enable
s

treatment in what is termed ‘the golden hour’
.

T
he quicker the casualt
y

gets
treatment
, the better the chance of survival

(L. K. Bandyopadhyay, 2009)
.


2
.
3

Chemical requirements


At the moment in a typical safety conscious environment where the likelihood
of a serious incident occurring could increase as hazardous chemicals are
moved around additional precautions would be taken. To help minimise the
serious
ness of these risks the best system to use would be the buddy
system. In using the buddy system there are also inherent risks, two people
are injured or worse. In order to mitigate these types of scenarios
,

the use of
RFID along with the correct type of
tag
,

t
here is the ability to tell if the person
is still alive using t
emperature and heart monitoring. It is also possible to te
ll
how fast the person has fallen using a velocity sensing tag or if they
are in a
hazardous gas environment using a gas sensin
g tag

(L. K. Bandyopadhyay,
21


2009)
. For the requirements of this project the ability to track the individual is
sufficient so a tag that reports location will be sufficient. The backend
software does have the ability to interfa
ce with other tags types so if in the
future the greater degree of tag interrogation is required the system can
handle it.


2
.
4
Conclusion


The RFID system is a good use of technology and ‘big brother’ helping the
safety aspect of industry.

With the degree

of flexibility that active RFID tags
provide within a broad spectrum of industry
,

such as that in
a hospital,

mining
scenarios and the ability to track personnel and equipment
. This
means that
the return on investment can be achieved quite quickly
.

It
a
lso
improves the
quality of care provided in

hospitals and enhance
s

the safety record of a
mine.



22


Chapter 3



Requirements

The
requirements of the system are not complex
.


However

the integration of
the system into
the
current
factory environment
makes it

more complex than
a standalone system.

As with all decisions, the weight of argument, final
price comparisons and support contracts determine which of the above
companies will end up providing the solution to the company. At a corporate
level the decisio
n to spend somewhere in the region of 2.5 million dollars will
go to the C.E.O. of the company.

A
ny
expenditure

over $5000
must
go up
the administrative chain. At 2.5 million dollars the justification for such
expenditure

must be backed by an impressive
argument. Corporations will
want the ROI on such a project to be documented in the very finest detail.


RFID Benefits

Some of the benefits that it is hoped an RFID system will bring to the factory
are, i
mprove
d

Wafer Speed of Development (Springtown TTM C
harter)
.
R&D is a big chapter in every industry and Seagate is no different. It is
hoped that
the ability to track lots more accurately will result in a faster
turnaround

in R&D lots and
enable

faster design configuration changes
.
Constant improvements
are being made in the processes and these are
implemented at certain steps in the build

and being able to locate these lots
more accurately will result in reduced search times
.


With real time location services available it is hoped that this will eliminat
e
l
ost
l
ots. We should be able to f
ind
l
ost

wafers in FAB or Office areas using
the RFID system.

With the rack mounted exciters in place the solution will
allow the system to make tag blink when lots are primed for dispatch to the
tool. With the system
making the tag on the lot box it should be clearly
visible where the lots are on the racks. This will help to m
inimise

l
ot retrieval
time
.
We hope to be able to m
inimise
u
ser involvement in
f
inding
l
ots
.
Currently we have a lot of material that goes miss
ing
for extended periods of
time.

There would normally be a number of operators/managers trying to
locate lots that are missing for any extended period of time. We hope to be
23


able to reduce this search time and utilise the personnel in a more productive
role.


RFID Active Roll

RFID will i
mprove lot retrieval time

from both an electronic
(
FactoryWorks
system) and a physical one.

AMA can watch the RFID database. If a lot is
moved to a location and not scanned onto a rack but this location is the
correct on
e then AMA can execute the move
-
to
-
rack function in
FactoryWorks. This means the compliance, i.e. the WIP physically on the
rack is the same as the electronic inventory, will be correct therefore the
dispatch logic will always pull the correct lots for th
e next run.

Occasionally
WIP is placed on the incorrect rack. With the help of AMA this WIP once
static for more than
one hour on a rack can be placed on hold. By placing
WIP on hold there will be greater visibility of incorrect movement of WIP.



Lost

Wafer Statistics

As mentioned earlier, occasionally WIP is placed on the wrong rack or
removed from the FAB altogether. When WIP is lost for 72 hours then it is
classed as scrapped for being lost. Over a 12 month period to July 2010
there were a total o
f
381 wafers.

For the 381 wafers lost there may be as
much as $323million worth of revenue loss if the wafers were finished and
ready to ship
.


Potential revenue loss ~ 323m dollars


(
381

wafers

* 80000 heads

per wafer



15% yield loss=25908000 heads @ 4 per drive = 6477000 drives
@ $50 per drive = 3238500000 million dollars
)


Productivity Loss

With wafers being lost for a variety of reasons there are a number of people
from each of the functional groups looking for the
m

The
Yield Team spend
15hrs per week searching for lost and mixed up
production
wafers
. The
individual shifts comprise of

seven

areas
. For each of the areas there is

one
person who would spend up to

3hrs per shift

searching for lost production
lots.

24


R&
D
will
average
five

people
each
spend
ing 3hrs per shift.
If

R&D wafers
are not found then they are handed off to the Yield team to search.



On this particular proje
ct the justification is twofold
, the h
ealth and safety of
the chemical handler (lone
worker) and a swift response in case of an
incident
.


T
he
re is also the

knowledge that there is a chemical handler on
-
site and available for the required task. This also ties into what is termed
‘Lean Manufacturing’. Lean manufacturing is the
practice of
looking at the

production system
as a whole including the KPIV’s
and
the
out
put

from

the
system. In this case there are input variables to the system, just one of
these variables is the chemicals required to etch metal or strip photo
lithographic resist.

If we take the chemical handler scenario, the production
manager wants to be sure the handler is available to refill a system before
the machine is taken offline and the old chemical
is
dumped. There may be
value running one more product in the machine b
efore calling for a change or
refill.



Figure
9

-

MobileView tab in FactoryWorks Client


There are ‘business rules’ that are custom scripts created to perform

a task
when executed. One of these rules is the assign / unassign rul
e. The
FactoryWorks system
,
a system
used to track WIP step by step, (figure 10)

25


has a new tab for MobileView.


This tab allows a lot and
a tag to be
associated within the gui

using the FactoryWorks custom business rule
‘assign lot’
. The business rule t
hen passes these attributes to MobileView
where the tag id and the asset are associated

and displayed in the custom
attributes of the lot properties (figure 11)
.

We can now the association
between tag and lot both in FactoryWorks and MobileView.



Figure
10
-
RFID TagID incorporated into Lot Attributes




Figure
11

-

Asset as
see
n on MobileView client



Figure
12

-

Asset location on map of office



The tag movement
s

can be n

in the Aeroscout solution, MobileView. The
MobileView system will display the assets
current
location and then plot this
on a map of that location (figures 12 & 13).

A historical timeline can be
26


displayed on the map. The timeline shows location at ten
minute intervals for
a given period
.



At a high level
figure 14 shows how

the
Aeroscout
RFID system integrates
into the current business architecture.


Figure
13

-

Systems Architecture



As there is a pre
-
existing Cisco network
on site and there is no need to
purchase additional hardware we can rule out the Juniper Networks system
as the factory is already WiFi capable. One can look at the four remaining
system vendors and that each can utilise the existing WiFi network and onl
y
install the
ir system software on site. N
ow it comes to the support package
that each vendor can provide. As the factory already intends to utilise the
location systems for tracking material through the factory using the
Aeroscout system of WiFi enabled

tags, rack mounte
d exciters and choke
points

the decision to use Aeroscout is
a

more cost effective one.

The use of
the rack mounted exciters is critical to the success of the system. Finding a
lot box on a packed rack makes it more effective when looki
ng for material

(
Figure
14
).


27



Figure
14

-

Rack showing material for processing



There are two type
s

of exciters in the system, the rack mounted exciter
(
Figure
15
) and
the choke point exciter. The rack mounted exciter will
provide definitive location of material and the ability to make the LED in the
tag (
Figure
16
) flash to determine its location in a heavily populated area.
The
rack mounted exciter will poll for new tags every 120 seconds. The
choke point (
Figure
17
) will provide the feedback on the material movements
through certain locations in the factory. The

choke point is set to scan
for
tags passing through every 200milliseconds.






Figure
15

-

Rack Mounted
Exciter

Figure
16

-

Active RFID Tag

Figure
17

-

Choke Point
Exciter

28



Newly acquired tags are activated using the
tag
activator system
(figure 20)

prior to use in the factory and assigned to the asset using Mobile View.

During the activation of the tag a predefined configuration file is saved to the
tag. This configuration
file defines

attributes for the tag such as

the
channel
the tags broadcasts on, the frequency of the tag blinking and the
transmission interva
l
.



The exciter
(
EX
-
3210 model
)
(figure 19)
, is used to locate the material lot box
when placed on a rack in the
fab. The exciter can, with the aid of a
command sent from the mobileview system, make the LED in the tag flash t
o
show its location on the rack, particularly a specific location within a group of
racks.

The choke point enables large amounts of data to be
acquired in a
sho
r
t space of time. It is particularly effective at doorways and in corridors.




Figure
1
8

-

Rack Mounted Exciter in
Place





Figure
19

-

RFID Tag Activator







3
.1

Current Issues


Overall there are a number of issues that will be addressed by installing the
RFID system. The
biggest ROI from this system is the ability to track in real
time the location of over five thousand production lots. The ability to track
these lots in real
time will help to reduce to cycle time. By being able to find
wafers by their physical location rather than the supposed location, it is
hoped that the cycle time will be reduced.

29



As
this is such a large investment the need to justify such a spend must b
e
looked at from more than one view point. The ability to track production lots
through the factory is a high level statement. On the floor the ramifications of
losing lots can be very costly. When an operator attempts to dispatch WIP to
a tool the disp
atch system will look at the apparent WIP on the rack and
suggest certain lots based on a set criterion. Once the system produces the
list the operator goes in search of the lots. There may be as many as 300
lots in one area. The operator must now find
6 lots among 300. If one of
these lots is not on the correct rack then the operator will continue to search
until he feels he has done all he can to find it. This may take 10 minutes or
so. Meanwhile the machine is sitting IDLE and there is the potentia
l for lost
production time. This all affects tool utilisation which results in a decrease in
OEE and it will take longer to recoup the capital investment on site.


The ultimate goal of the company as a whole is Time To Market. R&D is the
life blood of th
e company and
is a major factor in the success of the
Springtown operation. R&D lots generally take priority over all other lots in
the factory a
nd

are given priority one status. These lots are pushed through
production and out to the slider operation in

Penang
, Thailand,

where the
wafers are sliced and mounted onto the gimble ready for testing. Upon
successful evaluation of the R&D product the new specs may be
incorporated into the existing product. If there are changes made to the
product or the R&D b
ecomes a ship product then the WIP in the line running
under the R&D code will be converted to a new ship code. These R&D lots
still in the line need to be tracked and retrieved from the system, relabelled
and recoded then placed back into the system from

where they came out.
This is a major headache when trying to find every lot in the line of a certain
type. There may be WIP in the line that was placed on hold pending a re
-
measure or gone off to have a sample taken for evaluation. This is where
RFID t
racking is invaluable. The ability to track the lots from the web
interface and pin point them on a map will reduce search time considerably.

As mentioned earlier in the document the reasons for using the RFID system
to track the chemical handler have ari
sen from a need to have a buddy
30


system in order for a chemical handler to be able to acquire chemical from
the storage area and fill a machine. It is not always practical to have two
people on this
duty;

one may be on break or may have to cover another
op
erators break. The cost implications of waiting for two people to be freed
up at the same time also come into play. If a machine is what’s called a
priority one then the time it spends idle must be reduced to a minimum as
per lean manufacturing practices
.


The current practice is not conducive to a lean manufacturing facility where
all redundant processes and unnecessary handling of material are negated.
This is the point at which the RFID system can become an integral part of the
manufacturing process.

One man to fetch and fill a machine will help to
reduce the wait time at the chemical store.



31


C
hapter
4



Specifications

The system must be able to handle the high volume of RFID tags that are
required to track material within the current production fac
ility in Springtown.





Figure
20

-

Material Box with
RFID tag fixed to lid.

Figure
21

-

Tag Cases

Figure
22

-
Cisco wireles AP



At the moment production requirements means
that current stock levels sit at
8000 lots of which approximately sixty percent would be production lots that
will ship
. T
he remainder
are
R&D lots.

The factory already has an extensive
network infrastructure but as this is almost 15 years old is badly in
need of
updating. The network switches for the new system run POE, this allows the
installation of exciters without the need for additional power containment.

The manufacturing facilities must be networked with a series of POE cat5
cables and what are cal
led ‘Daisy Chain’ cat5 cables
(
Figure
23
).



Figure
23

-

Racking Diagram

32



Additional Hardware requirements


It may m
a little ‘over the top’ but in order to protect the tags a custom case
(figure 23)
has been
commissioned. The case was designed by a company
called Accutool. The first mould was presented to Seagate for evaluation.
The mould was made of a composite material

for evaluation
. The model was
tested to if it was fit for purpose prior to ordering t
en thousand pieces. The
cases prevent the tag from getting damaged or stolen.

The final cases will
be made from poly
-
carbonate material.

There are two colours of case
produced, clear and red. The red will be used for R&D lots, clear for regular
product
ion.

The tag fits very tightly into the case so there is nt room for
movement(figure 22). A special release tool has also been designed to
release the two sides of the case in order to remove the tag.
The wireless
access points have been purchased and in
stalled by the IT department. As
per the request from the system vendor the AP have been purchased to
meet the requirements of the system. The AP’s are Cisco Aironet 1260
series (AIR
-
LAP
-
1262N
-
E
-
K9)
(figure 24)
. The 1260 (1262) is a dual band
controller
offering 2.4 Ghz and 5 Ghz frequencies on 802.11a/
g/n standards.

The AP also has
internal RAM, so it’s not just the common wireless router
. It
dr
aws 12.95w over a POE network.



LAN Switching


POE Switches are Catalyst
WS
-
3560
X48PF
-
S

series
.
They have
an i
nput
power
of
1100Ws
, of which

800W

is available for use by the network. It can
support

48 ports @ 10/100/1000 mbps
.
POE
of

30W power on all ports

is
available which when fully utilised will push the unit to the limits.

Dual
redundant power supplies
with fans
, which provides some power outage
protection.

IPV4 and IPV6 routing

but as its on an internal network only IPV4
is required.

33



Figure
24

-

Catalyst 3560X POE

Switch(courtesy Cisco.com)



Aeroscout Equipment Specs.

Tags
that will be used with the system
are Aeroscout
T
AG
-
2000
.

These tags have an outdoor range of

600m

and indoors range of
180m
.

The tags operate on

802.11b/g
@ 2.4ghz and have clear channel sensing to
avoid

interference with wireless networks.

The transmiss
ion interval can be
set between 128msec to 3.
5 hours
.

The tags have 3.6volts lithium ½ AA
battery which is replaceable. The battery may last u
p to 4 years dependant
on use.


The r
ack mounted ex
c
iters are EX
-
3210.

These exciters

have a working
range of

20cm to 3m
. The
p
ower

input

is 12volts over

POE (802.3af)

hence
the need for the POE switches.

Choke point exciters are EX
-
2000B.

These
have a range of
50cm to 6m
.
They have an input of
24VDC

POE (802.3af)
,

again, as with the EX3210 these must be powere
d by the POE switches.




34


Chapter 5



Planning

Over the duration of this project from inception to the final completed project
there will be certain deliverables and milestones that will be achieved. In this
assignment the aim is to cover the scope of
work for this project and it is
hoped the required delivery times are achieved. Taking into consideration
home and work commitments there are still quite a few deliverables that
must be attained. As a part
-
time student this means a lot of home time spent

on literature review and documentation. Certain items have been marked as
milestones in the project. All of these are detailed in the plan of work

(figure
27)
, the schedules and Gantt chart.



Figure
25

-

Work breakdown


5
.1 Project Outputs

5
.1.1 Milest
ones

In order to get to a fully deployed test system all the hardware must be on
site and installed. Once the hardware is in

place

the software must be
installed and configured.

35


The system will be a h
igh availability system meaning that there will be
two server instances. There will be a primary server that will carry the
load and a fail
-
over system
that will be constantly updated. The data will
resi
de on an Oracle database and

all data will be repli
cated to a second
data repository. All system
s

have data duplication enabled with the
health of the

systems monitored by Wily intro
scope.
Wily reports back
statistics such as end
-
to
-
end response times.
Should a system fail then a
load balancer called
Big
-
IP will manage the data requests and direct the
requests to the second server instance or data repository.

The plan is to have the test system in place for a number of weeks to fully
debug the system. All aspects of the system must be fully tested. I
n
order to fully te
st the system it is planned to run a pilot of 20 lots with
tags
. This pilot will involve
placing tags on a batch of WIP,
placing into
the production line and monitoring their movements.
The test system
must be as robust as the producti
on system and will, for all intents and
purposes, be a fully deployable system.

The test system will
be
configured to send alerts via email in the event of
tags leaving the building or
a lot not reporting to the system for more than
one hour
(figure 28)
.

I
f this happens then a system of tracking will have to
be created starting from the last known point and working out where the
lot should go next.


Figure
26

-

Alert Configuration Page, MobileView


36


Should a lot fail to report then an investigation of circumstance must be
completed to find the cause.
There may be issues with the refresh rates
on the WiFi network or perhaps the tag has failed. There are a lot of
variables in the system
that cause som
e form of failure. Some of these
failure points include, tag failure, WiFi failure, server failure or network
failure. One main area for concern is the
levels of interference from
machinery in the FAB areas. When an exciter is attached to a rack the
fie
ld has to be custom set to allow for interference from the lcd display
attached to the rack and from interference from the adjacent machine. All
of the exciters are custom set for individual racks.

Movement of tags
through the FABs allows the blind spots
to be identified and corrective
measures taken.

This is also done to complement the wireless site
survey that was completed some months ago.

Alerts will also be raised if a lot goings through one of the exit points at
which the EX2000 exciters have been p
laced. We anticipate that tags will
go off site either by accident or by someone trying to steal one. At $70
each they are an expensive item and cannot afford to lose them. If one of
the tags goes through the exit point then an email alert is sent to ke
y
individuals.

Part of this staging process involves the training of the IT
support staff. The IT support network will have 24/7 duty of care for all
systems deployed on site. If an event happens for which there is no
OCAP then it may put the factory at
risk.


5
.1.2
Production System


If all of the tests are completed and the results look favourable then we
can plan to deploy the production system replicating settings used for the
test system. The WiFi infrastructure will remain the same only the
servers

and associated software and data repositories will change.

Again, as with the test system some test scenarios will be formulated and
acted upon to analyse the production systems performance. If all of the
37


variables for both systems are comparable then
a production pilot will be
implemented.

Once the system has been tested and performs as required then the final
report will be created that will document the systems setup and
management. The system must be accompanied by an OCAP(Out of
Control Action P
lan) that will outline procedure for stopping, starting and
updating the system. Procedures such as data purging within the database
etc must be documented. Part of the final report will be the creation of an
FMEA(Failure Modes and Effects Analysis) whic
h will attempt to weight
failures on a system and possible reasons and solutions. Any potential
system will either stand on its own or fall according to the outcome of this.


5
.1.
3

Deliverables


It is hoped the system will deliver all of the expectations

that were predicted it
would when the first discussions occurred.

Some of these deliverables
include greater

WIP location awareness, historical data retrieval and reduced
cycle time.

A report

documenting the system and how it functions including
system
start up and shutdown

will be produced
. This report must document
how the system will perform and will serve as an action plan if necessary to
bring the system down and backup. Production of the FMEA will also
determine if the system can withstand a prod
uction environment.

Finally
p
roduction acceptance

of the system
if all of the documentation is in order
and plans for system maintenance are in order then production will sign off
on deployment. This final test is very stringent as the failure of the syst
em
could result in lost production time.

5
.2 Project Plan


Figure 29

below shows the project plan with all the major events included
with expected time estimates.


38




Figure
27
-
Project Plan


Some of the major tasks involve the
procurement of the servers that the
system will reside on. The software will be developed to interact with the
central systems API interface. When the system is deployed on the test
server it will undergo load testing to ensure the system is robust enoug
h for a
production system.

It is hoped that the majority of the output will be
generated using SQL. It is hoped that using Oracle SQL and a copy of Toad
(Quest Software) that data can be extracted from the database and used to
track WIP and personnel. In

conjunction with the site maps it should be
possible to place assets on the map to some degree of accuracy.


System Testing

After the system has been deployed, the hardware tested, the software
installed and tested a number of simulations will be conducte
d. The
simulations will include tagging a lot box and placing it on one of the racks in
the FAB. A check of the database will show which of the APs pick up the
tag
. Within the FABs at least three APs will have to register the tags
location

in order to t
riangulate the tags location.


The signal strength will determine
the distance that the tag is

from the AP.

The lot will then be

moved to another location to determine how fast the
system will refresh the tags location. Ideally this will be real time as a
39


smooth handover between APs would be desirable.

A check will be made to
how the data is stored in the database and just how eas
y it will be to extract
this data.


It is hoped that the data from the wireless APs can be transferred to a map of
the FABs and an intelligent path mapped. It is not known how the data can
be interpreted if only two access points pick up the tag. It may
map the tag
between these two points in a straight line. At the moment this would not be
good enough to fulfil the requirements of the system. The next step would
be to tag one of the handlers or a volunteer. The subject would then be
asked to perform
a

task such as walk from FAB2 gowning area to the
chemical store
. The time taken for the subject to move from one area to
another would be gathered by stopwatch and compared to the database
entries. As the subject moves passed APs and through the choke po
ints
a
time stamp
would need to be noted. The
physical timing would then be
compared to the database timestamp to if there are any differences. Having
completed one task a number of times, i.e. walk from the gowning area to the
chemical store, the timin
gs would then be compared to if there are any big
deltas between the times. In an ideal world the person would walk the same
distance in the same time with very little error margin.


Next would be to ask a subject to walk from FAB1 to FAB2 and again do t
his
a number of times and compare the timings.

Next would be to ask one of the
subjects to exit through one of the choke points such as that at the smoke
hut and then back in through another entrance to how the system will react.
How would such a test be

presented in the database. As the reaction times
for the APs and the reaction times for the choke points differ the database
may show someone as being outside for 10 minutes when they are actually
only outside for 2 minutes depending on where in the refr
esh cycle the tag
comes back inside.

Can all the timings be accurately placed on a map and
provide true representation of a persons movement.

The tests may prove
that the tag2000 cannot fulfil the requirements of the subject tracking and
that an alternativ
e tag is required.


40



5
.3 Time Management


As a part
-
time student there is a need to balance home life, work life and
student life.


Work Life



As the work routine is days only it is almost the same every
day, with a few exceptions. It’s a busy schedule
but one that can be worked
around with 9am starts and 6:30am starts to assist with the home life. Work
life is busy and there is very little time to look at project work in the normal
work day.

Home Life
-

The home life provides some time to de
-
stress, a v
ery
important time. With the time spent at home a mix of study time, family time
and routine home care time. There are times when the best ideas can be
thought through when one is not actively focused on it. Time lines and work
strategies can be sorted
and stored at home and can be prepared for
college. This is also time that can be spent reading and researching
important topics. Ideas can noted and snippets of information can be
catalogued. Home time is very important with respect to the outcome of a

successful project.


School Time
-

Time at college can be broken down into class time and
study time. Class time is important to broaden the knowledge base. This
time is also used to provide direction with respect to the expected outcomes.
Study time is

very important both on campus and at home. The advantage
of the study at college is the software to implement changes or test new
theories is readily available. Study in college usually provides the best focus
for the student with little external interf
erence. This is a good time to switch
off the mobile phone.

An outline of a typical day can be
see
n in figure 30 below.


41



Figure
28
-
Time breakdown sample

5
.4 Working
w
ith Project Supervisor


The project supervisor can usually be
contacted by email or phone if
anything arises or if advice is required. A meeting with the supervisor is
generally scheduled for Wednesdays before class. The project supervisor
has given clear direction as to the requirements over the semester and
leadi
ng into semester two. Clear goals that have been set and being worked
towards right now and deadlines for producing written work for the project
supervisor have been set. The interim report is the main focus right now.
Research into the field of RFID tec
hnologies to gain a better insight into the
workings of the technology is on
-
going. How RFID is making an impact in
current production environments and the field of health and safety is also the
focus of research.


Alongside this work there is an API for
the main system, so the intention is to
use this to access the system. A meeting with the IT department needs to be
arranged to discuss the security of the system and how the system can be
accessed. Once this information is gained one can then look at th
e software
development and again the supervisor can help with this to provide direction.
This may require a refresher in C++ or study C# if necessary. Again the
supervisor may be able to provide advice in the best language to develop the
software on.


42


5
.5 Risk Management


Risk management is an important consideration in any project. As just one
aspect of the risk management strategy an FMEA must be completed. ALL
projects that are deployed or potentially deployed in the production
environment must have

an FMEA completed and evaluated by affected
parties, IT, MES and engineering departments.

Table
1

shows the FMEA
study conducted on site to evaluate the greatest ris
ks to the factory and
related systems.


Risk

Category

Severity

Occure
nce

Trigger

Action

Ownership

Server Failure

Business

5

1

cannot access system

Contact IT
-

Follow OCAP

Michele S

Network Failure

Business

5

1

Cannot access WCS systems

Contact IT
-

Follow OCAP

Noel H

RFID Tag failure

MES

3

2

Tags do not update

Replace tags

Fergal C

Software failure

Business

5

1

Load balancer alert

Switch to fail over server

Darren McL

Client failure

MES

3

1

Location cannot be determined
within FW client

Resport
issue to IT

Michele S

Table
1


Business FMEA data


Risk

Category

Severity

Occurrence

Trigger

Action

Ownership

Cannot write
the sof tware

Personal

High

Low

Project Falls
Behind

Extra learning required

Me

Cannot
make API
connections

Sof tware

High

Medium

WCS system
lockout

Ask WCS vendor f or
necessary settings

IT

Family
issues

Personal

Medium

Low

Cannot attend
class

Try to stay ahead of
schedule so project does
not suf f er also

Me

Supervisor
issues

School

Medium

Low

Unable to
check
project
progress

Get ahead as much as
possible so I am not
reliant on superv isor
constantly

Me /
Superv isor

Hardware
f ailure

Technical

High

Low

PC does not
start

Keep all project work
backed up in the cloud
and on USB driv e

Me

Cannot
access
Cloud data

Te
chnical

High

Medium

Cloud storage
is
inaccessible

Keep USB drive with me
at all times

Me

Cannot
complete
project

Personal

High

Low

Do not make
f inal deadline

Try to stay f ocused and
concentrate on tasks that
lie ahead, may be weeks in
adv ance

Me

Table
2

-
Personal Risks


After listing all possible failure modes the overall weighted result of the data
in the FMEA is evaluated and then the risk element can be applied to the
factory overall. If the risk is deemed to be too great the
n the highest valued
failures must be analysed to determine what if any action can be taken to
reduce this up to the point where the software/hardware vendor is contacted
for a fix or work around.


43


Looking at the personal risks
(
Table
2
)
that may affect the project there are a
number of steps that can be taken. If one can devote one or two hours per
night just on the research and write up of the project then one
should be able
to stay ahead of the deadlines. The biggest risk is that one cannot complete
the software necessary to interface with the wireless control systems. The
vendor has been contacted as yet about accessing the API. Time will be set
aside, at l
east one hour per night and more at the weekends, to work on this
project. It is intended that on a week to week basis one can achieve at least
10 hours study / write up time on this. Family and home commitments mean
that time may be lost during the week

but this can be made up at the
weekend.






44


Chapter 6

-

Design

There are a number of criteria that must be fulfilled in order for this project to
be successful. First and foremost the system must be able to integrate into
the existing factory systems.
If this first criteria is not met then the system
will fall down. Secondly it must be able to use the existing wireless
infrastructure. If it cannot use the existing Cisco system then additional
capital must be raised to purchase the required hardware.


Once all of the factory systems are in place, tested and verified then
additional projects, such as operator/chemical handler tracking can be rolled
out.

Once the system is implemented, bugs ironed out and the results of the
data analysis are available it
should show where operator compliance needs
to be addressed. Operators are required to move WIP to the next stage of
the build process

(
Figure
29
)

and ‘rack’ the wip

onto a holding shelf.



Figure
29

-

P
rocess plan showing Fab changes 1 to 2 on left


By analysing the timestamp from the RFID system and comparing to the
factory system a delta will be produced showing how much additional time
the wip has been on the rack

(
Figure
30
)

and not ‘racked in’ or can show a
disparity between the finish run time of the last stage and the wip moving to
the next stage
.

45




Figure