Frank Saffery - Final Thesis
University of Huddersfield Repository
Saffery, Frank
The Asovi System: Towards a solution for indoor orientation and wayfinding for the visually
Original Citation
Saffery, Frank (2012) The Asovi System: Towards a solution for indoor orientation and wayfinding
for the visually impaired. Masters thesis, University of Huddersfield.
This version is available at
The University Repository is a digital collection of the research output of the
University, available on Open Access. Copyright and Moral Rights for the items
on this site are retained by the individual author and/or other copyright owners.
Users may access full items free of charge; copies of full text items generally
can be reproduced, displayed or performed and given to third parties in any
format or medium for personal research or study, educational or not-for-profit
purposes without prior permission or charge, provided:
The authors, title and full bibliographic details is credited in any copy;
A hyperlink and/or URL is included for the original metadata page; and
The content is not changed in any way.
For more information, including our policy and submission procedure, please
contact the Repository Team at:
A thesis submitted to the University of Huddersfield in partial fulfilment of the
requirements for the degree of Master of Science by Research
The University of Huddersfield
January 2012
Copyright statement
The author of this thesis (including any appendices and/or schedules
to this thesis) owns any copyright in it (the “Copyright”) and s/he has
given The University of Huddersfield the right to use such copyright
for any administrative, promotional, educational and/or teaching
Copies of this thesis, either in full or in extracts, may be made only in
accordance with the regulations of the University Library. Details of
these regulations may be obtained from the Librarian. This page must
form part of any such copies made.
The ownership of any patents, designs, trademarks and any and all
other intellectual property rights except for the Copyright (the
“Intellectual Property Rights”) and any reproductions of copyright
works, for example graphs and tables (“Reproductions”), which may
be described in this thesis, may not be owned by the author and may
be owned by third parties. Such Intellectual Property Rights and
Reproductions cannot and must not be made available for use
without the prior written permission of the owner(s) of the relevant
Intellectual Property Rights and/or Reproductions
Wireless communication technology is currently an expanding resource from
which solutions into indoor orientation and wayfinding for the visually impaired
can be explored. However, as a technology in its infancy a prevalent system in
the field has yet to be established. Further to this, the potential of combining
wireless communication technology with a commercially viable interface
capable of providing feedback for the end user is as yet unexplored. Research
in current wireless and mobile technology combined with acquired knowledge
of wayfinding for the visually impaired has culminated in the development of a
new system seeking to maximise the potential of technology in the field.
Testing with visually impaired participants and subsequent focus group
discussion suggests that the ASOVI (Audio Based Spatial Orientation for the
Visually Impaired) system is a viable solution to indoor orientation and
wayfinding for the visually impaired capable of providing feedback to the end
user via a commercially viable platform.
Table of Contents
List of Figures
Dedications and Acknowledgements
List of abbreviations
Chapter 1 - Introduction
1.1 – Background & Context
1.2 – Research Objectives 11
1.3 – Thesis Methodology 12
Chapter 2 – Wayfinding & Cognitive Maps 13
2.1 – Cognitive Maps 13
2.2 – Spatial Coding 14
2.2.1 Small-Scale Space 15
2.2.2 Large-Scale Space 16
2.3 – Wayfinding
2.3.1 Path Integration & Dead Reckoning
2.3.2 Obstacle Avoidance 19
2.3.3 Navigation Aids
2.4 – Summary 21
Chapter 3 – Indoor Positioning & Orientation
3.1 – Wireless indoor positioning & Orientation
3.1.1 Triangulation 23
3.1.2 Scene Analysis
3.1.3 Proximity
3.1.4 Short Range Wireless (Bluetooth + Competitors) 26
3.1.5 Ultrasound
3.1.6 Infrared 30
3.1.7 GPS & GPS with Indoor Capabilities
3.1.8 Radio Frequency Identification (RFID) 35
3.1.9 Ultra-Wideband (UWB)
3.1.10 Wireless Local Area Network (WLAN) 36
3.2 – Visually Impaired Orientation & Wayfinding Systems
3.2.2 UCSB Personal Guidance System (UCSB PGS)
3.2.3 MoBIC 40
3.2.4 SpotON 41
3.2.4 Drishti 42
3.2.5 Cyber Crumbs 43
3.3 – Commercial Products 45
3.3.1 Loadstone GPS
3.3.2 Trekker Breeze
3.3.3 BrailleNote GPS
3.3.4 RNIB React
3.3.5 Talking Signs 48
3.3 – Summary 50
Chapter 4 – Mobile Technology
4.1 – Samsung Galaxy S II 52
4.2 – HTC Sensation 52
4.3 – BlackBerry Torch
4.4 – Apple iPhone 4 53
4.5 – Summary 55
Chapter 5 – Evaluation of Technology
5.1 – iPhone Indoor GPS Performance 57
5.2 – RFID Receiver Read Range 60
Chapter 6 – Design of a Peripheral Device
6.1 – Technology Used
6.2 – Design of the Electronics
6.3 – Design of the Application
6.4 – Physical Design 68
Chapter 7 – Evaluation 71
7.1 – ASOVI’s Ability to Aid in the Navigation of Indoor Environments and its
Reception with the Target Audience
7.1.1 Information feedback from the device 76
7.1.2 Alternate delivery methods and mobile devices
7.1.3 ASOVI Marketability 77
7.1.4 Requested additional features 78
7.2 – ASOVI’s Ability to Aid in the Creation of Cognitive Maps 79
Chapter 8 – Conclusion and Further Research 82
8.1 – Summary of Findings 82
8.2 – Future Research
8.2 – Conclusion
Appendices 86
Appendix A
Appendix B
Appendix C
Appendix D
List of Figures
Figure 2.1 - Points in traversing large-scale space.
Figure 2.2 - Boats predicted path using dead reckoning.
Figure 2.3 - Urban area with distances and angles needed to traverse it.
Figure 2.4 - Roadwork’s
Note: this is one example of a route that a visually impaired user may take.
Figure 3.1 - Triangulation with three reference points. (Liu, Darabi, Banerjee, &
Liu, 2007) 24
Figure 3.2 - Angle of arrival demonstated (Liu, Darabi, Banerjee, & Liu, 2007).
Figure 3.3 - Triangulation using RSS to locate a mobile device (Rodriguez,
Pece, & Escudero, 2010).
Figure 3.4 - Sound scale identifying Ultrasound (Raferty), 2010) 28
Figure 3.5 - Different examples of Ultrasound (Pitman, 1994).
Figure 3.6 - Ultrasound transmitter within an indoor environment.
Figure 3.7 – A diagram of light spectrum (The Modern Green, 2008). 31
Figure 3.8 - An example of DGPS (Note signals from the satellites go to both
the mobile device and the base station). 32
Figure 3.9 - Shows A-GPS assisting a mobile device that is indoors (Diggelen,
Figure 3.10 - European radio spectrum allocations. 36
Figure 3.11 - LANDMARC network diagram.
Figure 3.12 – A diagram of UCSB PGS (Golledge, Schematic of the PGS
System). 39
Figure 3.13 - Reginald Golledge demonstrating UCSB PGS(Golledge, Schematic
of the PGS System).
Figure 3.14 - SpotON compared to a standard biro pen.
Figure 3.15 - Shows the Drishti wearable system(Helal, Moore, &
Ramachandran, 2001). 42
Figure 3.16 - RNIB React Speaker (RNIB Business Development Team / SFX
Technologies, 2007). 47
Figure 3.17 - A town centre scene demonstrating how Talking Signs
works(TalkingSigns). 49
Figure 3.18 – Shows a transmission beam hitting pedestrian thus rendering the
system non functional. (Pedestrians from (TalkingSigns)). 49
Figure 4.1 - Bruno Fosi’s silicone case for the iPhone that allows users to feel
the screens layout.
Figure 4.2 - ColourID identifying colours from a photograph taken with the
iPhone (Esquirol, 2011).
Figure 4.3 - LookTel identifying American dollar bills (Esquirol, 2011). 55
Figure 5.2/5.3 – Showing the use of GoogleMaps to work out straight line
distance between points (Google, 2011). 59
Figure 5.4 – Showing the use BROADCOM BCM4750 Chipset in the iPhone4
(Chipworks, 2011).
Figure 5.5 – Diagram of experiment configuration.
Figure 5.6 – Diagram of experiment angles and positions. 61
Figure 5.7 – A graph to show read ranges of RFID chips through various
materials at a number of specified angles.
Figure 6.1 – A diagram of an iPhone cable 30 pin connector (Pinouts.RU,
Figure 6.2 - A diagram of the circuit used for the peripheral device.
Figure 6.3 - The design of the modified RFID device. 68
Figure 6.4 - RFID device heading towards an obstacle.
Figure 6.5 - Arcs of tags showing points of interest. 70
Figure 7.3 A graph to show the number of correct and incorrect answers by
each participant. 80
Figure 7.4 A graph to show the number of correct answers on each question.
Dedications and Acknowledgements
This thesis is dedicated in loving memory of my mother PC Lisa Jane Saffery.
I am truly indebted to my supervisors James McDowell and Dr Rupert Ward for
all their help and encouragement throughout my research. I am certain that
without it this thesis would not have come to fruition. I owe a great deal to my
family for their financial and moral support throughout this project and without
it this project would not have been completed. I would like to show my
gratitude to the Batley Blind Association for help with providing participants for
my investigations. Finally I would like to thank Sam Horseman who boosted
me morally and provided me with a huge amount of support.
List of abbreviations
3D - Three Dimensional
A-GPS - Assisted GPS
AOA - Angle of arrival
ASCII - American Standard Coding for Information Interchange
ASOVI - Audio Based Spatial Orientation for the Visually Impaired
Cell-ID - Cell phone Identification
DGPS - Differential GPS
DLA - Disability living Allowance
GIS - Geographic Information System
GND - Ground
GPS - Global Positioning System
IR - infrared
IrDA - Infrared Data Association
LOS - line of sight
MoBIC - Mobility of Blind and Elderly People Interacting with Computers
NFC – Near Field Communication
O&M - Orientation and Mobility
PDA - Personal Digital Assistant
PRN - Pseudorandom Noise code
RFID - Radio Frequency Identification
RNIB - Royal National Institute for the Blind
RSS - Received signal strength
RTOF - Return Rime of flight
RTPI - real time passenger information
Rx – Receive
SMP - Smallest M-vertex Polygon
SVM - Support Vector Machines
TDOA- Time difference of arrival
TOA - time of Arrival
USB - Universal Serial Bus
UWB - Ultra-Wideband
Wi-Fi - Wireless Local Area Network
WLAN - Wireless Local Area Network
Chapter 1 - Introduction
This chapter introduces the project by giving a subject background and
subsequent focus. It then outlines the thesis objectives and intended
methodology proposed to achieve these objectives.
1.1 – Background & Context
Recently technological advancements regarding mobile devices incorporating
Global Positioning System (GPS) receivers such as the Apple iPhone 4 (Apple
Inc, 2011), has meant that it is now possible to estimate the location of such
devices. This information can then be cross-referenced against a geographic
information system (GIS) or map such as GoogleMaps, to aid user orientation
and wayfinding. It is possible that the advent of this technology may have
significant repercussions when utilised for its orientating and wayfinding
capabilities to aid the visually impaired.
A significant body of research exists regarding spatial orientation and
wayfinding of the visually impaired in outdoor environments and there are
several products commercially available to this end. However, there is
significantly less research into the spatial orientation & wayfinding of an indoor
environment. GPS, the standard technology used for locating a mobile device,
requires a line of sight with the satellites surrounding the earth. The use of
GPS in an indoor environment is unable to maintain this line of sight and is
consequently inaccurate and in some cases will cease to function. Furthermore,
indoor environments differ to outdoor environments because they are
potentially more complicated to navigate due to their confined spaces. In
addition to this, obstacles such as supports, pillars and wall-divides
may provide extra hazards for the visually impaired. For instance, an office
building with identical floor plans for three different floors provides an easily
navigated environment for the visually able traveller but without guidance,
may cause confusion and disorientation for the visually impaired traveller due
to difficulty in distinguishing between floors on account of their identical floor
For years the visually impaired have had to rely on aids such as canes and
guide dogs in conjunction with specialised training that can takes years to
master in order to navigate around public places such as shopping centres,
airports, bus stations, town centres and train stations. This not only detracts
from the independence of the visually impaired but can also deter some from
leaving their homes altogether (The Times, 2007). In existing research
regarding orientation, mobility, wayfinding and cognitive mapping there is
great evidence that orientation within an environment leads to a sense of
security and independence and is therefore crucial for the incorporation of the
visually impaired within our intricate society (Espinosa & Ochaita, 1998). Thus,
to increase independence of the visually impaired and enhance their traversal
of buildings in public and private places such as the aforementioned airports,
shopping centres and train stations, an effective alternative to GPS must be
found. However, this alternative must be accessible and commercially viable.
There are products on the market for aiding the navigation of outdoor
environments such as Trekker by HumanWare. The company describe the
product as, “a hand-held talking GPS” device, although on account of it
providing no other functionality and costing £550 per unit it may be considered
commercially unviable (HumanWare, 2010).
Currently, the number of registered blind and visually impaired people in the
United Kingdom is approximately 370,000 and the estimated number around
2,000,000й The RNIB currently state, “Only one-third of registered blind and
partially sighted people of working age are in employment” (RNIB Research,
2011). The government offers Disability Living Allowance (DLA) to those who
need care or have walking difficulties due to mental or physical disability. The
maximum an individual can receive per week is £125 with the lowest amount
being £39.10 (UK Government, 2011). This suggests that for a number of blind
and partially sighted individuals dependant on DLA, the disposable income will
not be available to purchase single function products such as Humanware’s
Trekker. In order to increase commercial viability, products such as these must
either reduce in cost or incorporate multiple functionality in order to
justify such expenditure. One possible solution would be to utilise a mobile
phone. A large number of mobile phones currently on the market have multiple
uses as personal organisers, handheld games consoles, mp3 players alongside
the original functionality of a portable phone. And so, if a mobile phone was
developed to offer the additional functionality of orientation and wayfinding in
an indoor environment, it may offer a solution to this problem.
Overall, this information suggests that a device that is easily portable for the
user, commercially viable on account of its compatibility with an existing multifunctionality and capable of aiding orientation and wayfinding in an indoor
environment would have beneficial repercussions for the visually impaired. For
example, in buildings such as an airport, it could be used to inform and
orientate a user as to which check-in desk to use and how to navigate to it.
1.2 – Research Objectives
With the contextual knowledge outlined above, the following research objective
has been outlined. How might it be possible to spatially orientate visually
impaired persons within an indoor environment using a combination of mobile
phone technology and wireless communication technology?
In order to successfully create an indoor orientation and wayfinding system,
research into cognitive mapping must be conducted. This will include
investigating the methods by which humans store data using different sensory
channels and subsequently use it to traverse both indoor and outdoor
environments. This will be followed by further investigation into how the
visually impaired create cognitive mapping without the use of vision in order to
create an orientation system that can successfully bridge the missing visual
channel. Secondly, an investigation into the different technologies available for
indoor orientation and positioning will be conducted to evaluate their
advantages and disadvantages dependant on technique, accuracy and cost for
example. This will culminate in identifying suitable technology with which to
develop the system. Finally, research will be conducted into the different
mobile technologies in order to identify the most suitable multi-functional
device on the market. In doing so, the chosen device may then be developed
to incorporate the function of indoor orientation and wayfinding capabilities for
the visually impaired providing an alternative solution to standalone devices.
1.3 – Thesis Methodology
To address the objectives outlined, this exploratory investigation must begin
by conducting a pinpointed survey into the overlapping areas of wayfinding,
mobile technology and indoor positioning technology. Specifically, research into
wayfinding will highlight how humans and the visually impaired in particular
acquire, store and utilise sensory information in the form of cognitive maps for
use in orientation. Having laid a theoretical foundation of wayfinding,
orientation and cognitive mapping, an exploration of the opportunities and
limitations associated with appropriate wireless communication technologies
will be undertaken. This should highlight appropriate technologies for use with
an indoor orientation and wayfinding system while taking into account cost,
availability, precision and functionality. Following this research, a further
exploration of current wayfinding systems on the market will aim to identify
current competitor pricing, availability, infrastructure and success. This should
provide the knowledge necessary to identify suitable technologies to be used in
the design of an indoor orientation and wayfinding solution to aid the visually
impaired as outlined in the thesis objectives. Suitable investigations may then
be designed and implemented to ascertain the specifications of the proposed
technology to be used in order to facilitate the design of a successful system.
Using the information gathered in the previous review and investigation, a
peripheral system can then be designed and engineered. Once complete, an
investigation with end-user participants will be conducted. This investigation
will use a mixture of both qualitative and quantitative methodologies to extract
information pertaining to the success of the designed peripheral system. The
results of which will then be used to answer the proposed research objective
regarding the aid of cognitive mapping processes, orientation and indoor
wayfinding of visually impaired individuals.
Chapter 2 – Wayfinding & Cognitive Maps
Current research regarding cognitive mapping defines it as the methods by
which humans build up a mental map that provides information regarding the
location of objects and their relativity to each other and themselves
(Morrongiello, Timney, Humphrey, Anderson, & Skory, 1995). This chapter will
introduce and analyse wayfinding and cognitive maps as a primary background
contextualisation for the project. It will continue to dissertate different
techniques used for orientation and mobility (O&M). These techniques range
from using a cane to using audio or haptic feedback that help guide the
visually impaired through environments. Having established this, it will be
possible to investigate methods to improve cognitive mapping for the visually
2.1 – Cognitive Maps
A cognitive map can comprise of complex geometric, trigonometric and
mathematical data, quite often far more complex than a person could solve
using their own knowledge (Golledge, Klatzky, & Loomis, 1996). This
information is essential for humans to locomote and is collected using various
sensory channels such as vision, audio and touch. It is then processed with a
number of mathematical and psychological transformations and turned into a
map in the hippocampus of the brain. The hippocampus is the area of the brain
that deals with large-scale spatial orientation and navigation (Fortin, et al.,
Cognitive maps are fundamentally a mental representation of the external
environment in which an individual is situated. These maps are built up by
input from different sensory channels and allow us to navigate the world
around us. For instance, a visually able person sat at a desk can see a cup and
reach for it, pick it up, take a drink and place it back down with relative ease.
This is because their cognitive map will tell them how far away the drink is and
the proximity in relation to other objects on a desk etc.
Vision enables humans to calculate distal judgements of objects in relation to
themselves and other objects around them. Most of the information essential
for creating a cognitive map is thus collected via the visual sensory channel.
However, visually impaired people do not have access to information provided
by the visual sensory channel. As a result, the visually impaired traveller needs
to compensate with information from other sensory channels such as touch or
sound (Lahav & Mioduser, 2008). Therefore, limitations of the visual sensory
channel can severely compromise development of the system of spatial
representation (Morrongiello, Timney, Humphrey, Anderson, & Skory, 1995).
Using the earlier example of a visually able indiviual picking up a cup at a
desk, if information from their visual sensory channel was prohibited and the
cup was moved, they would subsequently have difficulty in locating the cup.
Landmarks are a major contribution in the creation of cognitive mapping and
can constitute anything from an elevator in a building to a large statue in the
middle of a town square. The use of landmarks is highly prevalent on a daily
basis when locomoting. For example, when giving directions to a traveller it
may be stated, "turn left at the statue of Harold Wilson, then go straight until
you get to the fire station". The statue of Harold Wilson and the fire station are
both landmarks that will be used to help create a cognitive map in the
traveller's mind. According to urban planner Kevin Lynch, landmarks are the
most important cue in any environment (Lynch, 1960). For the visually
impaired it is not possible to see landmarks and therefore gain information
from them. This prohibits the visually impaired's ability to orientate themselves
to the same level of accuracy as a visually able traveller. Landmarks tend to be
both static and silent, visable to a traveller before they reach it, upon which
they will then start scanning for the next landmark providing orientation within
the environment. However, the visually impaired user must walk up to the
landmark to identify it and deduce a position within an environment they may
know little about. Therefore developing a system to indicate to the visually
impaired what and where a landmark may be before they reach it could help
greatly in their orientation within an environment.
2.2 – Spatial Coding
According to Millar (1994, p.118), "The term spatial coding is used here for
coding in terms of some form of reference. Thus, the location, distance and
direction of the position of an object have to be specified (whether or not
explicitly) with respect either to oneself, or to external coordinates, or both".
There are two different types of reference cue that summate the basics of
spatial coding and these are called frames. Frames can be self-referent frames
or external frames. Millar expands, "Self-referent frames are centred on the
person’s body. External frames are based on information from the
environment" (Millar, 1994, p.119).
There are two key types of space when it comes to spatial coding: small-scale
space and large-scale space. Small-scale space or haptic space is that which a
person can manipulate or explore without changing the actual location they are
in. In comparison, large-scale space or locomotor space is that in which
locomotion is required to navigate or explore (Unger, 2000). This distinction
can be made clear by considering that in small-scale space, objects should be
within arms reach and therefore found in relation to one's body (self-referent).
Whereas in large-scale space we must locomote to explore and in this the body
has to translate. This means that finding objects with self-referent frames
reduces dependability (Unger, 2000). A major advantage of the visual sensory
channel is that during locomotion external frames are updated thus giving
distal information to the changing environment around the locomotor. This
allows a visually able locomotor to see landmarks and the distance between
them resulting in a more detailed cognitive map.
Another major factor in the development of cognitive maps is the length of
time that the locomotor has been visually impaired. Those who have had some
visual experience will consequently have a heightened spacial ability. For
example, according to Unger (2000) the later a subject loses sight, the closer
their performance in spatial tasks to that of a sighted subject. This notion is
further supported by Dodds, Howarth, & Carter (1982) in a study in which both
congenitally and late blind children were asked to walk a short urban route that
they had been taught and then point out a small number of locations along the
route. They found that the errors increased the further away from the target
the children were. All the children were able to walk the route yet they noted
that late blind children were more accurate at pointing out the locations. They
subsequently surmised that the congenitally blind children coded the route in
terms of the changes of heading they made using self-referent coding. This
meant that they found it difficult to take this information and use it to create
an external representation of the test environment.
2.2.1 Small-Scale Space
According to Millar (1994), people with a small amount of visual experience will
mostly use self-referent coding. She found that children without vision do not
recieve the same quality of information obtained through other sensory
channels as through the visual sonsory channel. Therefore they find other
techniques to code the information, leading to using self-referent coding being
used more than external coding techniques. This could be exemplified by
imagining a person trying to find a static switch that is in the centre of a desk.
A sighted person could easily see the switch and press it using external coding
as they would be able to see other items on the desk and gain distal
information from them. However, if somebody without sight tried to undertake
the same task and used the same method of external coding then they would
have to use the haptic sensory channel to feel out for other objects to be able
to find the switch. In contrast, if they use self-referent coding then they know
how far the switch is away from them and in what direction it is. To expand
upon this, an example described by Millar (1994) concerns being sat on a train,
looking out of the window at an adjacent train. When the other train starts to
move, we may wonder which of the trains is moving until we realise that we
are still stationary using our kineasthetic senses. This also shows that in
certain situations, other sensory channels are more accurate than the visual
channel. In this context, the example shows that the mind chooses the most
appropriate method of collecting the information for each situation. However,
the congenitally blind have never had enough visual experience to have learnt
how to use external coding and take note of external cues. As a result, they
will select other methods that are the most appropriate for them. This notion is
supported by Ungar, Blades, & Spencer (1995b) who investigated blind and
partially sighted children's mental rotation capabilities by asking them to
examine and recreate a layout of shapes, either from the same location or
after a rotation of 90% around the display. The study highlighted that the
children who related objects to each other and to the frame itself found it
easier to locate the objects when the rotation had been made.
2.2.2 Large-Scale Space
When a visually able person wishes to locomote across a large-scale space
such as the lobby of a hotel, avoiding obstacles or judging the distance to the
point they wish to locomote to is not problematic. Furthermore, Rieser,
Ashmed, Taylor, & Youngquist (1990) also highlight that providing a sighted
person has seen the space such as the lobby first, then they can accurately
walk to a given point should you remove their sight. One major problem with
large-scale space is the influence of traversing objects. A visually able traveller
can see these traversing objects using their visual sensory channel and in
doing so rapidly update their cognitive maps allowing them to avoid said
objects. However, a blind or visually impaired traveller does not have this
A problem with large-scale space that applies to both sighted and visually
impaired persons is that a start point and an end point are not close to each
other and so traversing any large urban area has to be done in stages - each
with its own goal (Millar, 1994). To illustrate this notion, Figure 2.1 shows a
hypothetical map of a city centre with the grey blocks representing tall
buildings that line the roads. The traveller must start at point A and finish at
point B.
Figure 2.1 - Points in traversing large-scale space.
This process has to be undertaken in three stages because even a visually able
traveller cannot see point B from point A. They must instead travel to
intermediate point 1, from which point B is still not visible, and so they must
next travel to intermediate point 2. Finally from intermediate point 2, they can
see the goal of point B (if they have visual information) so the third stage of
the journey can be undertaken. Using Figure 2.1 as an example, the
information that is received to navigate from point A to B is done so in three
sequential stages in comparison to a small-scale space such as a desk, where
all the information needed to navigate between points is received in one stage.
If the layout of an environment is complicated enough to require many of
these stages of data retrieval to get to a goal, then it is very possible for even
visually able travellers to become disorientated and lost.
2.3 – Wayfinding
Golledge, Klatzky, & Loomis (1996) argue that, "Wayfinding refers to a
person's ability, both cognitive and behavioural, to find his or her way from a
specified origin to a specified destination". In order for a person to undertake a
wayfinding task they need identifiable landmarks to determine their location,
information on the heading in which they are travelling, turning information
and a form of dead reckoning called path integration. In current literature,
there is almost an integration of the terms, "mobility" and "wayfinding" and
the two are often used synonymously. Gollege et al (1996) state, "mobility is
often used as a synonym for wayfinding". However, mobility has connotations
regarding the restriction of a person's movement due to a physical impairment
of some variety whereas wayfinding has connotations of a cognitive skill more
than a restriction by impairment (Golledge, Klatzky, & Loomis, 1996).
Many blind and visually impaired persons have had specialist O&M training
either using a cane or audio cues from the environment. One of these
techniques is linear feature following which consists of locating the end of a
pavement or a fence and following this directly. It is also often called
shorelining (Golledge, Klatzky, & Loomis, 1996). This technique can be
problematic because some routes will have no curb edge, fence or wall to
follow resulting in travellers finding themselves disorientated thus forcing them
to rely on other techniques of orientation.
Golledge et al (1996) propose that for a blind traveller to improve their
wayfinding abilities two main types of information are needed. Firstly,
information about the "proximal environment" is required which is then
represented in a cognitive map (Golledge, Klatzky, & Loomis, Cognitive
Mapping and Wayfinding by Adults Without Vision, 1996). An example of this
would be the number of doors passed before getting to the correct office.
Secondly, information about the larger environment is also required such as
any sudden drops in ground level as occurring in embankments or ditches. As
stated earlier, this information is easily collected by a visually able subject
whereas a visually impaired user must collect this information via other
methods. Golledge et al (1996) then go on to state that audio perception is
becoming, "the major substitute for the long distance information processing
otherwise provided by vision". Using this research, it can be deduced that
audio feedback may be used to provide landmark information and other spatial
cues at the same distance as a visually able individual would see them to help
visually impaired users orientate themselves.
2.3.1 Path Integration & Dead Reckoning
Dead reckoning is the process in which current position, heading and estimated
speeds over a given time are used to calculate a prospective given position.
Figure 2.2 shows a diagram of a boat sailing in the sea travelling at 10 knots.
The triangle around the boat indicates the direction in which it is travelling and
the line then shows its predicted course. The boat is marked on the diagram at
the time of 14:30pm. The subsequent times marked on the line indicate its
predicted points at 15:00 and 15:30.
Figure 2.2 - Boats predicted path using dead reckoning.
Although dead reckoning is no longer used in this context, it has in the past
been a primary method of marine navigation allowing sailors to navigate
across oceans that have no visual landmarks to use for orientation.
The form of dead reckoning used by animals (including humans) is called path
integration. Etienne & Jeffery (2004, p180) describe path integration as the
"simplest form of navigation (that is used to) ...return to the starting point of a
journey without making use of familiar position cues". This occurs using
locomotion signals and a single reference point such as a nest in the case of a
bird. Path integration can therefore be defined as a number of small
increments of movement, noted in reference to the direction and distance from
the given starting point used to update a cognitive map. This method uses only
one landmark as the starting point used by the animal to remember the
distance and direction in which it has travelled.
It can therefore be suggested that landmarks are not essential for spatial
orientation although they are beneficial. It is possible that a traveller, knowing
the location and direction from which they started, may use kinaesthetic
signals to encode how far they have travelled. Although research cited refers
to animals navigating a featureless environment, it is possible that visually
impaired users can encode each rotation as well as the distal information to
build up a cognitive map of the area using these rotation and transformation
signals and traverse a complex urban environment. Figure 2.3 uses the same
hypothetical city map as figure 2.2 but this time shows the distances the
subject has to travel and the rotations they have to undertake.
Figure 2.3 - Urban area with distances and angles needed to traverse it.
If a subject was to walk the path from A to B then they would have to encode
the route as 20 metres forwards, 90 degree turn to the right, 10 metres
forwards, 90 degree turn to the left and finally 40 metres forward to point B.
Theoretically, if a visually impaired subject was to walk this path with an aid or
using landmarks then they should be able to follow this route a second time
more comfortably by using path integration. This also supports research
regarding wayfinding in that it too uses a string of spatial tasks or decisions,
rather than one continuous operation.
2.3.2 Obstacle Avoidance
One of the most reported problems faced by the visually impaired traveller is
the avoidance of obstacles, which is imperative for safe locomotion in an
environment. A considerable amount of research exists regarding different aids
to help with obstacle avoidance for the visually impaired, for example the
Mowat Sensor and the Sonic Guide (Brabyn, 1982, p. 286). The basic
operation of these systems is not to spatially orientate but help the traveller
avoid everyday environmental obstacles such as lampposts, bollards and
benches. While a cane has a maximum reach of about three feet from the body
some of these sensors have a range of up to eighteen feet giving the traveller
a greater time to react to such obstacles (Golledge, Klatzky, & Loomis, 1996).
However due to the fact that these sensors do not spatially orientate, the user
still has the potential to become disorientated and consequently lost in any
environment. These sensors do not identify individual landmarks and therefore
obstacle avoidable is most useful in a close environment. In the larger
environment however it gives little help in creating a cognitive map of the
environment's layout. For the purposes of a system to aid with spatial
orientation, some form of navigation aid is needed. It is possible that this
feature could be used in conjunction with an obstacle avoidance feature
although care must be taken to not overbear the traveller with information
feedback thus causing confusion.
2.3.3 Navigation Aids
Navigation aids, unlike obstacle avoiders, give the user information about the
general environment they are either already traversing or are about to
traverse. There are a number of aids already on the market that will be
reviewed in detail later in this text. These aids use a multitude of methods to
provide the user with information about the environment. These methods can
include ultrasound, lasers, video cameras and global positioning system (GPS)
with the information often fed back to the user using an audio or tactile
GPS devices normally link in with a geographic information system (GIS) that
triangulates the position of the user and then references it to a map to give the
user information about the surrounding environment. The devices then relay
verbal commands to the user telling them directions to their destination. These
systems rely on the accuracy of the GPS system in use and so will be discussed
at a later point in this chapter.
Navigation aids are a highly useful tool for aiding the visually impaired traveller
with finding new routes through an environment. O&M training teaches the
visually impaired to follow the same paths daily to reach any destination
(Jacobson, 1993). However, if a visually impaired traveller was taking the
route in the urban environment from Figure 2.3, a burst water pipe causing an
obstruction could force traversal of an unknown route resulting in
disorientation without the use of a navigation aid. Figure 2.4 exemplifies this
circumstance. With a navigation aid however a user could locate a new route
from A to B and even navigate the curvature of the route with relative ease.
Figure 2.4 - Roadwork’s
Note: this is one example of a route that a visually impaired user may take.
In Figure 2.4 the visually impaired traveller reaches the road works repairing
the hypothetical burst water pipe and then must rely on a navigation aid to
find the most suitable alternative route. It must be noted that if it was possible
to link the GIS system up to a modern traffic system such as the GPS systems
in use by cars, the road works may have been detected and an alternative
route suggested prior.
2.4 – Summary
The literature reviewed in this chapter suggests that the visually impaired can
navigate large-scale spaces and work within small-scale spaces efficiently.
However, this is made significantly easier with the help of navigational aids
and obstacle avoidance techniques. Research has also shown that cognitive
mapping abilities are stronger in persons with visual experience although the
visually impaired can still create and utilise cognitive maps. Once more, the
use of aids can result in the creation of a more comprehensive mapping which
in turn, allows for easier and safer traversal of a large-scale space.
Finally, the review outlines the notion that although blind and sighted persons
differ in performance when it comes to wayfinding, they have essentially the
same abilities to create cognitive maps and utilise them. In utilising other
sensory channels such as audio and tactile channels, the visually impaired can
gather information to substitute their deficit in vision. However, this
supplementary information does not replace fully that gained from the visual
sensory channel. It is in this deficit that obstacle and navigational aids can be
used to relay additional information to the user. This indicates that a system
could be designed to improve the ability of the visually impaired to navigate
and wayfinding within an indoor environment. The system would also have the
potential to aid in the creation of cognitive maps if it could provide information
that would otherwise be ascertained using the visual sensory channel.
Having researched these issues, the next chapter will explore how they may
be implicated regarding different wireless communication technologies
available and their potential use in the creation of a system to aid orientation
and wayfinding for the visually impaired.
Chapter 3 – Indoor Positioning & Orientation
This chapter summarises different types of indoor positioning technologies
taking into account costs, availability, setup difficulty, accuracy and usability
for the application of spatial orientation for the visually impaired. This chapter
also looks at systems for both the visually impaired and able currently
available on the market.
Over the past decade there has been a great deal of research into aiding the
visually impaired to navigate large-scale spaces and create cognitive maps to
help with wayfinding. However the majority of this research is focussed on
outdoor navigation in which several studies have looked at the use of GPS for
orientation purposes. The initial part of this chapter examines different
techniques for indoor positioning. Whilst the systems to be looked are not
exclusively intended for use with visually impaired users, nonetheless they will
be examined for their potential to be used in this context.
3.1 – Wireless indoor positioning & Orientation
There are three main methods that are used by indoor positioning systems to
locate a position of a person or object indoors. These are Triangulation, Scene
Analysis and Proximity. Using one of these methods allows for a system to
estimate the position of an object or person indoors, however using a
combination of one or more of these methods will allow for more accurate
3.1.1 Triangulation
Triangulation uses two types of estimation to assess a target’s location,
lateration and angulation. The process of lateration uses multiple reference
points to work out how far objects are in distance from each other and then
using this information estimates the position of an object. Angulation, on the
other hand, uses two or more reference points and calculates the position of an
object by detecting the angles at which the signals interject. There are several
algorithms that can be used to calculate these processes using different
methods of distal measurement.
One of the methods used for lateration is time of arrival (TOA). This uses the
theory that the “distance from the mobile target to the measuring unit is
directly proportional to the propagation time” (Liu, Darabi, Banerjee, & Liu,
2007, p. 1068). To calculate an object’s position in two dimensions, a
minimum of three reference points must be ascertained and used to measure
the distance between each and pinpoint the location of the target. Figure 3.1
below shows this.
Figure 3.1 - Triangulation with three reference points. (Liu, Darabi, Banerjee, & Liu, 2007)
TOA is calculated by measuring the one-way propagation time and using this
data to ascertain the distance between the measuring unit and the target. The
process begins by synchronising each of the recievers and transmitters and
placing a timestamp on the signal being sent between them in order for the
distance travelled by the signal to be calculated. There are several methods for
computing TOA, however the basic method uses geometric calculations to
derive the exact point of intersection (Liu, Darabi, Banerjee, & Liu, 2007).
Another method for lateration is time difference of arrival (TDOA). This
technique measures the difference in time that a signal arrives at from three or
more receivers. The technique uses hyperboloids thus is also referred to as
hyperbolic positioning. With two receivers an emitter can be traced onto a
single hyperboloid, and a third receiver is then introduced to give a second
TDOA measurement. From this process, the position of second hyperboloid can
be derived. Taking the two hyperboloids and intersecting them will give a curve
on which the emitter can be placed. Received signal strenth (RSS) can be used
for lateration however several drawbacks exist with this methodology. RSS
needs line-of-sight (LOS) to work accurately, without which the signal would
take multiple paths (Multipath Effect) thus delaying the signal and causing it to
show the wrong position. The multipath effect causes signal delays by
reflecting and refracting the signal causing ghosts which in turn cause errors in
the distal readings given. This is a common cause of the floating of a GPS
position on a mobile device, even when stationary or reading inaccurately a
number of metres. Return time of flight (RTOF) is a method that measures the
time of flight of the signal as it travels from the transmitter to the receiver and
back again. RTOF is very similar to TOA and uses the same type of measuring
mechanism (Liu, Darabi, Banerjee, & Liu, 2007).
The main technique of angulation is the angle of arrival (AOA). This technique
locates a target by plotting angle lines and then finding the intersection point
as shown in Figure 3.2.
Figure 3.2 - Angle of arrival demonstated (Liu, Darabi, Banerjee, & Liu, 2007).
AOA can be used to calculate a 3D location using three or more measuring
units. A major advantage of this method is that the measuring units do not
need to be synchronised. One of the disandantages of AOA is that it can, like
RSS, be affected by the multipath effect.
3.1.2 Scene Analysis
Scene analysis is an approach that first collects features or fingerprints of the
surrounding environment and then estimates the location of an object by
matching online measurements with the collected features. Location
fingerprinting generally relies on collecting and matching signal characteristics
that are location dependant. Fingerprinting has two stages: an offline stage
and an online stage. The offline stage involves surveying the environment and
its locations in which coordinates and signal strengths are collected and
labelled. The online stage involves using a positioning technique to crossreference the data collected online with that offline to estimate a position.
Fingerprinting uses RSS and thus faces the same problems as RSS
triangulation regarding signal strength, refraction and reflection. Because of
this there are several different algorithms that are presently used. The two
main algorithms are the Probabilistic method and the k-nearest-neighbour
method (Liu, Darabi, Banerjee, & Liu, 2007).
The probabilistic method cross references the signal characteristics collected in
offline mode (S) against the live collected positions (Li) which there are “n” ofй
This can then be made into a decision rule equation like this:Pick Li if P(Li|S) > P(Lj|S) for i, j = 1,2,3,…,n and j≠iй
Simplified P(Li|S) refers to the probability that the object or person is in the
location Li given the signal characteristics being received “S” (Liu, Darabi,
Banerjee, & Liu, 2007).
k-nearest-neighbour method uses the RSS to find the k closest matches from
the offline database previously created. Then averaging the k location closest
matches (weighted or unweighted) an estimation of the position can be made.
Neural networks with a multilayer perceptron are another available method in
which there is one hidden layer with offline readings taken to train the neural
network. Online data can then be fed into the network to get a position
estimate. Support Vector Machines (SVM) can also be used which is a
“technique for data classification and regression” (Liu, Darabi, Banerjee, & Liu,
2007). Finally, there exists smallest m-vertex polygon (SMP) that uses RSS
values to find locations with respect to each transmitter independently. Mvertex polygons are created by picking one or more locations from each
transmitter and averaging the coordinates of the vertices gives a location
estimate (Liu, Darabi, Banerjee, & Liu, 2007).
3.1.3 Proximity
The proximity approach uses a grid of receivers with the exact position of each
one placed in the system. A mobile target is considered as nearby when a
receiver detects it. When two or more receivers detect the target, it is
considered to be near the receiver with the strongest signal strength. This
approach is relatively simple to implement and works well with infrared radio
(IR) and radio frequency identification (RFID). This method is also used for cell
phone identification (Cell-ID) as the phone can be tracked to its nearest cell.
However, this is just in the proximity of the cell or the receiver it is near.
3.1.4 Short Range Wireless (Bluetooth + Competitors)
There are several different short-range wireless technologies available however
Bluetooth is the most widely recognised and used. Its current competitors
include ZigBee, near field communication (NFC) and wireless Universal Serial
Bus (USB). These wireless technologies were created focussing on the transfer
of information from one computer device to another. Each has a different
range of data transfer ranging from 20cm with NFC up to approximately 100m
with USB class1 devices. Each of these four different technologies has their
own specialised areas. Bluetooth mainly focuses on mobile phone data
communication and wireless communication between handsets and their
receiver’s domestic phonesй It first appeared in 1994 and was soon used to
transfer data from phone to phone (Hoovers, 2011). ZigBee aims itself more
towards very low power, low cost, self-configuring nodes that could be used in
conjunction with automatic blinds and lights for example (ZigBee Alliance,
2011). NFC is popular in Japan for detecting a wide range of data on mobile
devices from information to advertisements. NFC has a very short range of
only a 20cm radius of the transmission point. Wireless USB uses UWB
bandwidth and therefore does not classify as short range. UWB will be
discussed separately at a later time in the chapter.
The majority of research in positioning is currently on Bluetooth, possibly on
account of its cost and availability. There are two methods of calculating a
location using Bluetooth signals. The first uses trilateration in which
information from several different nodes can be taken into account. Because
the signal energy decreases almost proportionally with the distance between
the station and the mobile device, an algorithm can be applied and an
estimation of the location made (Rodriguez, Pece, & Escudero, 2010). The
second method used by Bluetooth to calculate positioning is proximity sensing
seen as Bluetooth (Class 2). As it has a range of approximately 10m from the
mobile device, it is possible to place a user within a certain distance from a
particular node; this however is not as accurate as the trilateration.
There have been several experiments researching whether Bluetooth is a
feasible and accurate medium to determine location. A large majority of the
experiments use RSS and fingerprinting to obtain a location estimate. RSS is
problematic because the signal can encounter several problems as discussed
earlier regarding RSS. In a study by Rodriguez, Pece, & Escudero (2010),
triangulation was used via RSS obtained from n different access points to
estimate the location of the mobile device. Figure 3.3 shows this.
Figure 3.3 - Triangulation using RSS to locate a mobile device (Rodriguez, Pece, & Escudero, 2010).
The heavy calculations required by the investigation were implemented on a
server that then sent the information back to the mobile device. This allowed
its processing power to be kept to a minimum. The average distance of error
with three access points is <= 1 to 2 metres (Rodriguez, Pece, & Escudero,
2010). The study also noted that the the speed of bluetooth can be
problematic and the inquiring time very slow at approximately 7-8 seconds.
This would cause jittering if Bluetooth were to be used for keeping track of
moving objects and would not display a smooth movement progress.
A significant advantage of using bluetooth is that the majority of modern
mobile devices and personal digital assistants (PDA) are fitted with a bluetooth
transceiver even at the cheaper end of the market, for example the LGA140
from T mobile costing £14.99 (T - Mobile, 2011). Another advantage of
bluetooth is that the signal can penetrate through objects such as walls and
3.1.5 Ultrasound
Ultrasound is sound above the audible level for human hearing. This range
varies from human to human but in healthy adults its average upper boundary
is 20 kHz (20,000 hertz). Figure 3.4 exemplifies the boundaries between sound
and ultrasound. The animal kingdom use ultrasound in the form of
echolocation to navigate and hunt in the environment. The most famed for
using this technique is the bat although it is also utilised by whales and
dolphins. Bats have evolved to use echolocation so effectively that they can
hunt fast moving prey such as moths without the use of sight.
Figure 3.4 - Sound scale identifying Ultrasound (Raferty), 2010)
Echolocation, also known as biosonar, is a process that uses the echo of sound
produced as it reverberates off surrounding objects. The returning echos are
modified versions of the outgoing pulse and thus the brain of the animal can
turn the data into mental images of its surroundings. Distance is calculated
from the time it takes for the echo to return to its source based on a speed of
340m per second in open air. For example, a sound taking 4 miliseconds to
return to its source means the object is at a range of 68 centimetres (Jones,
It can be argued that the most famous manmade use of echolocation would
have to be the sonar pulse system used by navel vessles and applications. It
uses the same principals as echolocation used by the animal kingdom in that a
sound is emitted before a reciever waits for the reflected sound. Figure 3.5
shows animals and a submarine each using echolocation and how the wave
Figure 3.5 - Different examples of Ultrasound (Pitman, 1994).
Although indoor positioning using ultrasound utilises the same fundamental
concepts as echolocation in the animal kingdom and the sonar used on a
submarine or boat, some differences must be clarified. Ultrasound positioning
uses trilateration in which the user wears a transmitter that sends out an
ultrasonic wave to be detected by receivers placed around the environment.
These signals can then be used to estimate the position of the user.
Furthermore, it should be noted that the short wavelength of ultrasound signal
does not penetrate walls or doors thus confining the user to one room
(Greenemeier, 2008). Figure 3.6 shows a diagram of a transmitter emitting
ultrasound waves that travel towards the receivers in a room. Using the speed
that sound travels in air, each receiver can work out the distance of the
transmitter using methods of trilateration.
Figure 3.6 - Ultrasound transmitter within an indoor environment.
One of ultrasound’s several advantages is its low cost with an ultrasound
transceiver costing as little as £12 (Robot Electronics, 2011) or £3.50 per unit
for separate receivers and transmitters. Although careful placement of the
receivers must be undertaken in order for optimal reception of the ultrasound
signal from the transmitter. There also must be some separate system or
computer to take the distance readings from the receivers and trilaterate it
into an estimate of position.
A primary disadvantage of ultrasound for indoor positioning is the interference
caused by environmental changes such as atmospheric temperature or room
temperature. These can cause changes in the propagation speeds by up to 3%.
Another disadvantage is that echoes emitted from other surfaces must be
discounted to avoid inaccuracies in the position estimation (Lorenz, Schmitt,
Oppermann, Eisenhauer, & Zimmermann, 2001).
3.1.6 Infrared
Infrared light is electromagnetic radiation with a longer wavelength than visible
light. Figure 3.7 shows the position of infrared on the light spectrum.
Figure 3.7 – A diagram of light spectrum (The Modern Green, 2008).
Infrared has several uses including night imaging and thermal imaging. More
recently, communication between mobile phones, computers and other devices
has become possible by using infrared and the Infrared Data Association
(IrDA) protocol. Arguably, the most established use of infrared light is with
television remote controllers and wireless controllers for games consoles.
These devices emit infrared light from light-emitting-diodes (LED’s) as a
compact beam that has to be aimed in the direction of the receiver. The beam
transfers data by modulating on and off rapidly, encoding the data towards the
receiver which converts the data into an electrical current to control the device.
Infrared works using line of sight and so, for the purposes of orientation, would
require a lack of obstacles or walls to work accurately. This would obviously
prove problematic for orientation in most indoor environments. Furthermore,
as an infrared beam is fired directionally, the beams would have to correspond
to the position of the receivers in each room. This incurs that infrared is not
the best-suited technique for triangulation. However, it may be suitable for
relaying information such as room names and numbers to a user traversing a
Infrared transmitters and receivers are of minimal cost with starting prices of
£1.00. This would mean that even in a large university building such as
Canalside West at the University of Huddersfield it would be a minimal cost to
place one at each of the doors to the staff offices and labs allowing users to
know what the room is and if it is in use before they enter. This would operate
using the aforementioned proximity method. This was used to create the
Talking Signs system in America to relay audio information to users who
cannot read signs and will be discussed further at a later chapter.
3.1.7 GPS & GPS with Indoor Capabilities
GPS is the most prolific system used for the positioning of people, cars and
other objects within the outdoor environment. GPS uses satellites that orbit the
earth at an approximate altitude of 20200 kilometres over a time period of 12
hours. Each of the satellites is equipped with an atomic clock for high precision
synchronisation. A ground station tracks the satellites ensuring that they are
orbiting the earth correctly and transmitting accurate data. Finally, the GPS
receivers obtain the signal emitted by the satellites and use this internally to
calculate and estimate its position (Chivers, 2006). GPS information is often
fed into a GIS and the device position then shown on a map.
For GPS to work at optimal accuracy it must receive signals from a minimum of
four satellites orbiting the earth. GPS works with line of sight so a clear path
between the satellite and the receiver must exist. Therefore in built-up urban
areas the signal can be weakened resulting in unreliable measurements. The
multipath effect can also occur if the signal bounces off any nearby buildings.
These factors make using GPS indoors difficult because the transmitted signals
have to travel through walls (which can comprise of a variety of different
materials) and in turn slow the signal down or block them altogether (Chivers,
Differential GPS (DPGS) uses a set base station that never moves position; it
then checks the satellite measurements to determine the error amount from its
stationary position. Any mobile GPS receivers in the area that are using DGPS
receive this error amount and apply it to their GPS position. This is because
any two GPS receivers in one area should be affected by the same atmospheric
conditions (Kotsakis, Caignault, Woehler, & Ketselidis, 2001). Figure 3.8 shows
DGPS working by the satellites sending signal to the base station and then
base station on to the mobile device.
Figure 3.8 - An example of DGPS (Note signals from the satellites go to both the mobile device and the base station).
High Sensitivity GPS are extremely fast GPS receivers that work at high speeds
in an open location. The extra power that the receivers use to work at such
high speeds can be used to integrate weak signals that are received at indoor
locations so that they can be processed into an indoor estimation of position
(Wikipedia, 2011). The problem with using high sensitivity GPS is that it is
unreliable in an indoor environment become it is too attenuated to work with.
The major advantage of being able to use GPS indoors is that it does not
require extra systems - it can work directly from the infrastructure that is
already in place. However, the cost of an indoor receiver is quite expensive
and requires the extra cost of infrastructure installation. For instance, the ublox SuperSense high sensitivity GPS receiver costs 85 Euros per unit (U-Blox,
2011) or in GBP, £71.06 per unit (XE Corporation, 2011) (Accurate on
Recently GPS has seen a new development called Assisted GPS (A-GPS) which
is now commonplace in mobile phone technologies. A-GPS improves the
performance of standard GPS by having a different information channel send
information to the GPS receiver.
Figure 3.9 - Shows A-GPS assisting a mobile device that is indoors (Diggelen, 2009).
Figure 3.9 depicts a satellite sending data (the square digital wave) and a
Pseudorandom Noise code (PRN). As the data hits the walls of the building it is
obstructed and cannot pass through. The PRN code however can penetrate but
gets weaker as it passes through more and more obstructions. This is where
the GPS is ‘Assisted’ by the cell tower which sends the same data or its
equivalent to the mobile device. As the diagram illustrates, the receiver still
obtains the information even though it is blocked by several obstacles. Also, it
must be noted that even with no obstacles, the signal transmitted from the
satellite to the A-GPS receiver will receive the information faster than a
standard GPS receiver (Diggelen, 2009). Incidentally, each GPS satellite works
on a different frequency to compensate for the Doppler shift effect because
most satellites travel at over 3km/s (Diggelen, 2009).
A primary advantage of A-GPS is that it informs the receiver which frequencies
to search in order to find the satellite’s signal followed by the data from the cell
tower instructing the receiver where to look to locate the satellite positions.
This not only increases the dwell time (the time the satellite sits over a certain
part of the earth) that in turn increases the amount of signal received at each
frequency but makes the system much faster overall. This also means that the
sensitivity of the A-GPS receiver is increased and can therefore acquire signals
at much lower signal strength than standard GPS (Diggelen, 2009).
3.1.8 Radio Frequency Identification (RFID)
RFID is a method of storing data on tags and then retrieving the data with a
receiver. It has several modern day applications such as tagging library books
and keyless entry for doors for example, an access door in the Canalside West
building at the University Of Huddersfield in which only staff and those with
disabilities have access. An RFID tag, also referred to as a transponder, has a
small amount of data storage that it broadcasts using a built-in antenna. There
are two main types of RFID tag: passive and active. An active tag has a power
source such as a battery whereas a passive tag runs from power induced from
the receiver (Hunt, Puglia, & Puglia, 2007).
There are two methods in which RFID can be used for positioning. The first is
the proximity method that relies on low frequency, passive RFID tags
transmitting up to approximately a third of a metre (RFID Journal LLC, 2011)
meaning that if a user is to pick up tranmission from the tag they must be
within its proximity range. Information can then be relayed to the user
regarding their position. For example, if each door in a corridor was equipped
with an RFID tag that explained to visually impaired users which room they
were outside and what it contained, it could aid the creation of cognitive maps
thus helping to orientate them within the building. The second method uses
active RFID tags within an environment to triangulate the position of the
receiver using RSS. Because active tags have a power source and are
fundamentally small transceivers, this means that they have a broadcast range
of upto tens of metres (Liu, Darabi, Banerjee, & Liu, 2007). The downfall of
using RSS with active RFID is that the error margin can be reach upto 2m.
Using this method for spatial orientation could be problematic as users could
be misguided into objects or become disoriented jeapodising user safety.
3.1.9 Ultra-Wideband (UWB)
UWB is radio signal communication that generates signals with a bandwidth
wider than 500MHz. It works by sending incredibly short pulses (less than 1
nanosecond) using very low power (Arslan, Chen, & Benedetto, 2006). The
maximum data rate that a digital radio signal can carry is roughly
proportionate to its bandwidth providing the signal is above the receiver noise
and is unaffected by it (Ingram, 2006). It is for this reason that the first
implementations of UWB was in wireless USB because it enables large amounts
of data to be sent over very short time periods, which is needed to replace the
USB2 data rates of 400MHz. Furthermore, if the data rate of UWB are slowed
down additional advantages appear such as increased transmission range and
lower transmission powerй This means the UWB’s original range of 10m or less
can be increased greatly making it suitable for covering a large indoor
environment. Unlike many other radio frequency signals, UWB does not suffer
from the multipath effect. This is because it uses such short pulses to send
data so any reflections of the pulses are easily filtered. Also, the signals can
pass easily through brick, plaster, clothing and equipment however it does
have interference troubles when trying to pass through liquid or metallic
materials (Liu, Darabi, Banerjee, & Liu, 2007).
A disadvantage of UWB is that it uses a large proportion of the radio spectrum
that may be in use by other services. Satellite television works at frequencies
above 10000MHz and so from 1MHz to 10000MHz is divided between other
applications such as mobile phones, satellite communications and military
applications. Figure 3.10 shows the main European spectrum allocations.
Figure 3.10 - European radio spectrum allocations.
As illustrated in figure 3.10, a large ratio of the spectrum is already allocated
to other services and so it is complicated for UWB to share it without
implicated interference.
The short-pulse waves used by UWB allow accurate placement by using the
TOF and TOA algorithms on the burst transmission from the transmitter to
receiver. Also, UWB can utilise the AOA algorithm to enable 3D positioning.
Using these algorithms and UWB short-pulses, a high level of precision can be
achieved (sub 20cm) (Liu, Darabi, Banerjee, & Liu, 2007). Such accurate
indoor location sensing make UWB ideal for use in spatial orientation of the
visually impaired - this is because accuracy of 20cm and less is within human
reaching distance meaning the possibility of user disorientation is low.
3.1.10 Wireless Local Area Network (WLAN)
WLAN is a radio frequency technology that gained its reputation for wirelessly
connecting computers and other wireless devices such as mobile phones, PDA’s
and laptops to networks and most commonly, the internet. This connection is
often advertised as Wireless Fidelity (Wi-Fi) and a large amount of café’s, bars
and restaurants now use this to allow their patrons to have free access to the
internet. WLAN works in both indoor and outdoor environments. An example of
this would be the University Of Huddersfield’s campus network which mobile
phone users can access anywhere on campus.
The accuracy of WLAN positioning systems ranges between 3m to 30m and
updates every few seconds. Considering the number of wireless points within
the general urban area, the possibility exists for information on the local
environment’s to be relayed to users including the visually impairedй A major
advantage here would be the ability to use existing infrastructure, thus
incurring no extra costs to implement. It should be noted that a vital
disadvantage would be WLAN’s inaccuracyй Furthermore, some areas in a city
may have no WLAN access points for hundreds of metres thus a user could be
left disorientated (Liu, Darabi, Banerjee, & Liu, 2007).
Another method of positioning using WLAN uses the k-nearest-neighbour
technique in which an environment is set up with specifically placed WLAN
transceiver points which are then used to estimate the position of the user.
This method yields a much more accurate estimation, however it would require
the installation of more WLAN points thus adding a greater cost to the project.
It also requires to be set up specifically for each environment. There are
several methods and packages already on the market to be subsequently
discussed that use different techniques however each requires specific
adaptation of the environment by adding WLAN points and so falls prey to the
problem of having to purchase the infrastructure.
3.2 – Visually Impaired Orientation & Wayfinding Systems
This section aims to review systems previously discussed in their suitability to
aid the visually impaired using the techniques mentioned in Section 3.1. These
systems are not commercially available to buy but have been part of other
research projects and tested to evaluate their use in orientation and
wayfinding of visually impaired persons.
In 2004 collaboration between the Hong Kong University of Science and
Technology and Michigan State University developed LANDMARC, an indoor
location sensing system using active RFID. LANDMARC uses active RFID tags,
much like the landmarks in daily life, to help estimate the location of the user.
The LANDMARC project test environment consisted of a network of active RFID
tags and some standard passive RFID tags, a wireless network with internet
connection, an RFID receiver that can communicate with the wireless network
using IEEE 802.11b, a server to implement the calculations and a mobile
device used to contact the server via the wireless network. Figure 3.11
illustrates this.
Figure 3.11 - LANDMARC network diagram.
As shown in Figure 3.11, the RFID tags broadcast to the reader that in turn
contacts the wireless network sending the data to the server. This is then
transmitted to the PDA or other mobile device in use so information can be
relayed to the user.
LANDMARC uses k-nearest-neighbours for scene analysis by using RSS
strengths of the nearest RFID tags. However, as noted by the group, RFID tags
do not allow access to the RSS directly. They instead only report the power (on
a scale of 1-8 in this instance) to the reader. Consequentially, they may
measure what distance each power level corresponds with but as was found,
this may only work correctly in free space. This meant that it is not possible to
calculate distance accurately using the power levels and, as concluded by the
test group, to accurately use RSS to estimate position would require RFID
manufacturers access to granted to the RSS value.
The LANDMARC system showed that RFID is a cheap and viable option for
tracking a position within an indoor environment. However, three major
limiting factors were highlighted. Firstly, the final conclusion mentioned that
the RFID vendors need to allow direct access to the RSS to gain greater
accuracy of estimationй Secondly, latency between the tags emitting their ID’s
needs to be shortened although this is currently unachievable as RFID
manufacturers lock the time interval with the average used in the LANDMARC
study being 7.5 seconds. Lastly, different tags, even those produced by the
same manufacturer using the same batteries, may provide different power
levels and different signal strengths. This may result in inaccuracy when trying
to calculate the signal strength as no two tags are identical and calculations
must assume all tags have the same strengths. The disadvantages of this
system outweigh the advantages and as such, the system cannot be used
accurately for orientating the visually impaired within an indoor environment.
The research did, however, provide great insight into the pitfalls of RFID for
scene analysis(Ni, Liu, Lau, & Patil, 2004).
3.2.2 UCSB Personal Guidance System (UCSB PGS)
Between 1985 and 2008, Jack M Loomis led a team from University of
California, Santa Barbara conducted a project concerned with researching and
developing a GPS-based navigation system for the aid of the visually impaired.
During the extended research period, technology advanced and with it, the
project, from using older GPS-based receivers to modern DGPS receivers
integrated with a GIS. The system built upon standard GPS with addition of
narrated instructions, such as “go forwards” and “turn left” and other points of
interest such as landmarks in parks, specific doorways, car parks and other illdefined areas such as more rural areas. The idea of the UCSB PGS was not
only to convey a route to get from point A to point B but also to relay nearby
points on interest. This was done, not just by speech but also by giving an
impression of auditory virtual reality by spatially orientating sounds to seem as
if they are emitted from the position of that particular point of interest.
The study used several different methods to convey information to a user
alongside different GPS receivers and other pieces of hardware equipment. The
idea of the project was to test each different type of feedback and different
types of GPS receiver position such as GPS and DGPS to get the most accurate
position. Combined, it aimed to identify the optimal way of orientation using
GPS, GIS, auditory and haptic feedback. Figure 3.12 shows a diagram of the
system and how it works. Figure 3.13 shows Reginald Golledge with the
system on him.
Figure 3.12 – A diagram of UCSB PGS (Golledge, Schematic of the PGS System).
Figure 3.13 - Reginald Golledge demonstrating UCSB PGS(Golledge, Schematic of the PGS System).
One of the downfalls of the UCSB equipment is the size and weight of the
equipment. As Figure 3.13 shows, the equipment is very large and for some
weaker people may be overly cumbersome when travelling any substantial
distance. This could be one of the reasons the system did not reach the
commercial market.
3.2.3 MoBIC
The MoBIC project (Mobility of Blind and Elderly People Interacting with
Computers) was run by the University of Magdeburg with the aim of helping
blind and elderly people in fixing their position and navigating towns. It used
GPS in combination with maps and, unlike some other systems, had a prejourney system where the user can make provisions before they travel with the
MoBIC outdoor system taking over in outdoor environments.
The MoBIC team first tested use of the route planning system which involved
using a computer with a synthetic speech output. The user then explored
electronic maps alongside this allowing the software to find an optimum route.
This information was then relayed to the user enabling the creation of a
cognitive map of the environment prior to entry. The outdoor system then uses
DGPS to give relatively precise information regarding the user’s location and an
electric compass to give the user bearing. The interaction with this system is
done via a small handheld keypad that relays information via artificial speech
in the format of clock-type directional commands. An example of this would be,
“change direction to 3 o’clock and continue for 100 metres to High Street”(Gill,
1996). A specially designed serial earphone is used allowing the audition of
commands without preventing external environment sounds being heard.
MoBIC ran a number of field tests, one of which took place in Berlin with six
people attending 25 tutorial sessions before being asked to test two unfamiliar
routes of approximately 1200 metres. Both aspects of MoBIC performed to a
high standard with feedback from participants positive and testers impressed
with the accuracy of the system.
One of the disadvantages to MoBIC was reported in the first part of the system
in which users familiarise themselves with an environment before traversal.
This made the assumption that the user knows exactly where they are going
having an end destination and relies on the system being up to date. If, for
example, road works were taking place on the day of travel they would
inevitably have some problems navigating the area due to the cognitive map
created prior not being wholly accurate.
3.2.4 SpotON
SpotON was a project run by the University of Washington, which aimed to use
radio signals, and RSS to obtain distal information to estimate the location of a
user in a small-scale indoor environment to guide visually impaired users
around indoor environments. The team began by evaluating a piece of RFID
equipment currently on the market for the purposes of triangulating position of
the user using its RSS. However, after large amounts of experimentation it was
found that the accuracy - whilst capable for automating lights in a room upon
user entry - is not great enough for the purpose of accurate positioning. Issues
were also identified with the time taken to obtain readings from the sensor (in
the region of ten to twenty seconds), which was deemed too great for use with
moving objects.
The team then developed and engineered a customised piece of hardware
aiming to overcome these problems and so track people with greater accuracy
and at a faster rate. The customised hardware incurred some extra cost to the
project with the hardware cost $120 to manufacture (it was noted, however,
that the mass production costs would fall between $30-$40). There were
further issues regarding the power needed to run the hardware and as a result,
larger batteries were needed in a revised system increasing the weight and
size of the product.
One disadvantage of the SpotON system’s functionality was the design feature
of processing data using a server before relaying it back to the user. As the
project was conducted in the late 1990’s to the early 2000’s, mobile
technology was not yet powerful enough to compute the data at a speed
necessary to keep track of moving objectsй The system’s size was
advantageous, however, being small and compact unlike the earlier mentioned
UCSB PGS. Figure 3.14 shows an image of the SpotON device in scale against
a biro pen.
Figure 3.14 - SpotON compared to a standard biro pen.
The SpotON team concluded that although the system may be accurate to a
certain degree, to build a ‘complete’ system other features such as compass,
accelerometer and other location devices such as GPS should be
used(Hightower, Borriello, & Want, 2000).
3.2.4 Drishti
Drishti is a wireless navigation system that uses several different technologies
such as wearable computers, voice recognition systems and voice synthesis
systems alongside wireless networking, ultrasound, GIS and GPS. Drishti finds
optimised routes through both urban and rural areas taking into account static
and dynamic data, such as road works, roadblocks and traffic congestion. The
user of the system can command it by using voice input prompting the system
to relays its information back via a set of headphones built into the headset as
can be seen in Figure 3.15.
Figure 3.15 - Shows the Drishti wearable system(Helal, Moore, & Ramachandran, 2001).
The system estimates the user’s position using DGPS before voice inputs
request route information or information about the environment. When a user
enters an indoor environment where GPS no longer functions accurately, the
user can command the system to change to ultrasound navigation. Here, two
ultrasound sensors control navigation using the time of arrival algorithm.
The accuracy of this indoor system was at its worst 22cm with a mean of 10cm
or less out of 22 test cases(Ran, Helal, & Moore, 2003). This confirms that the
system can work with reasonable accuracy indoors allowing for general
navigation. However, obstacle avoidance may be problematic for the user thus
total reliance on the system would not be possible. As mentioned in the prior
discussion on ultrasound, it suffers from reflection, which would cause severe
inaccuracies in cluttered environments. Another major disadvantage of Drishti
is apparent in Figure 3.15, where the equipment is ergonomically cumbersome.
Not only does it appear uncomfortable to wear, it could possibly create further
feeling of exclusion from visually able society.
3.2.5 Cyber Crumbs
Cyber Crumbs was created in collaboration between Atlanta VA Rehab R&D
Centre, Charmed Technology Inc and Georgia Institute of Technology. It is an
indoor orientation and wayfinding infrastructure that was designed and created
with the visually impaired in mind. Cyber Crumbs works by using infrared to
transmit data between the “crumbs” (which are small IR transmitters) and a
badge worn around the user’s neckй These crumbs can be compared to a trail
of breadcrumbs leading users around buildings. When a user enters a building
they must go to an information desk and select their chosen
destination/destinations within the building. The information desk then
calculates the simplest route to each destination and downloads this route to
the user’s badgeй These directions are in the form of sequential Cyber Crumb
ID numbers and a set of directions that correspond to the landmarks that each
crumb is related with(Ross, Lightman, & Henderson, Cyber Crumbs: An Indoor
Orientation and Wayfinding Infrastructure, 2005). When traversing the
environment the badge receives information from the crumbs. This information
is processed and audio information is then relayed via earphones. The crumbs
are only placed at strategic points such as office doors, bathroom doors,
staircases and hallways junctions thus cutting back on the cost of the
infrastructure and only sends relevant information to the user in an attempt to
not overburdening them with information.
One advantage of the Cyber Crumbs system is that the crumbs themselves lay
dormant in sleep mode to conserve power and thus stop the need for constant
battery changes. The crumbs only become active when they sense a badge
within up to 12feet. Conversely, a disadvantage of the Cyber Crumbs
infrastructure was the high cost of the components needed. Each badge cost
$600 to produce and the team were quoted $25 per crumb if they purchased
100 at a time falling further to $5 a crumb if they purchased 100,000. For a
facility such as the Canalside West building, University of Huddersfield, careful
observations of the schematics revealed it would cost approximately $5000 for
crumbs and up to $6000 for the badges if a high volume of visually impaired
users were to be catered for summating in a total of $10000 or £6710
(accurate on 20/04/2011)(Ross, Cyber Crumbs for Successfull Aging with
Vision Loss, 2004). Another problem was encountered with signal reflections
over the distance travelled by signals resulting in system inaccuracy.
3.3 – Commercial Products
During investigations into systems aiming to aid orientation and wayfinding for
visually impaired persons, commercial products were found for both indoor and
outdoor use. This section briefly examines major products that were uncovered
during research.
3.3.1 Loadstone GPS
Loadstone GPS is a free GPS package developed for Symbian-based mobile
phones that use the Series60 platform(Kirkpatrick, 2011). Loadstone involves
transmitting satellite GPS signals to a Bluetooth GPS device connected to a
mobile phone that, using a screen reader, relays information such as points of
interest and distance to landmarks to the end user. Loadstone prides itself as a
cheap solution, excluding the price of professionally designed maps and
expensive pieces of additional hardware. It also allows users to add their own
landmarks such as friend’s houses, supermarkets, bus stations and train
stations. This function will further aid the user to build their own cognitive
maps of an environment, increasing familiarity and in turn, confidence and
inclusivity into general society.
The major advantage of the Loadstone system is that it is free to use and the
user only needs a mobile device compatible with the system. The Loadstone
system can be downloaded for free and used as a supplementary system to
any existing orientation and wayfinding systems. One disadvantage of the
system is that the user must purchase a Bluetooth GPS receiver to be able to
triangulate their position to work with the Loadstone system and as such, the
system requires financial outlay before it can be functional. Obviously, owning
a mobile phone with GPS already installed would negate this disadvantage, as
no additional expenditure would be required. Also, by not using a GIS system
thus saving costs incurs that accuracy to the user’s environment is totally
reliant on the addition of personal landmarks. This would be problematic when
the user encounters an unknown environment. Therefore, using the Loadstone
GPS for exploration is not possible.
3.3.2 Trekker Breeze
Trekker Breeze is a talking GPS unit manufactured by
Humanware(Humanware, 2011)й Humanware bill the product to be “as simple
as your TV remote” and it consists of a controller that can be used in one hand
allowing the other to be used with a cane or hold the lead of a guide dog.
Trekker orientates the user by giving a verbal description of where they are at
the touch of a single button. This information includes that on businesses and
public transport buildings such as bus stops and shops. The Trekker handset is
lightweight at 7 ounces and has eight hours of functional battery life, which
means a user can have full use of the device for an entire day. Similar to
Loadstone, Trekker allows users to store landmarks such as their friends’
houses or favourite restaurants, allowing them to build a cognitive map of
familiar environments. However, unlike Loadstone, Trekker allows full
navigation routes in which it can direct the user from a present position to a
designated landmark or location. Therefore Trekker can theoretically be used
to explore unfamiliar environments.
Unlike other systems, Trekker does not use headphones to relay information
instead featuring a speaker built-in to the unit. This feature has both
advantages and disadvantages. For instance, the system allows users to hear
environmental cues such as moving traffic that earphones may otherwise
mask. Likewise, the system may also succumb to ambient noise in the
environment, which could mask the information being relayed by the speaker
itselfй One of Trekker’s major disadvantages is the high cost for a piece of
equipment with a single functionality. Its cost of 945 Canadian Dollars is
equivalent to approximately £600.66 (XE Corporation, 2011) (Accurate on
31/01/2012) which is financially unviable for a visually impaired potential
buyer in the UK, should they be unemployed and reliant on DLA.
3.3.3 BrailleNote GPS
BrailleNote GPS is another product from Humanware, which uses two pieces of
equipment to aid user’s orientation. The system comprises of a custom-built
BrailleNote or VoiceNote input/output device that connects to a DGPS receiver.
Used in combination, the user can input a destination and then output the
needed route from their current position. Similar to Trekker, this system allows
the user to store personalised points of interest, but differs in the user
interface with the DGPS unit.
A major downfall of this system is not only the size of the BrailleNote or
VoiceNote, but also that users must be connected to the DPGS unit via a wire.
This could prove ergonomically cumbersome, especially for long distances.
Because BrailleNote or VoiceNote can be used for interfacing with other pieces
of software such as Windows, it could be considered an advantage that any
expenditure for the hardware can be rationalised by using it in several different
ways (including orientation and wayfinding). Humanware although
manufacturing BrailleNote GPS, market Trekker as their main orientation and
wayfinding system.
3.3.4 RNIB React
RNIB React is a talking sign system to help the visually impaired orientate
themselves within town centres and is used as a device to improve user
confidence and independence. The system is designed so that in addition to
orientation, it can deliver additional information such as tourist information and
real time passenger information (RTPI). RNIB React uses radio frequency
technology to contact speaker units placed in strategic locations to a fob in the
possession of the user. When a speaker unit picks up a broadcasted signal
from the user’s fob, a recorded message is emitted at a suitable volume for the
surrounding ambient noise conditions. The opening message gives information
to orientate the user to their current location followed by further, more
detailed information including RTPI. The system has been installed in a number
of locations throughout the UK including Newcastle City Centre, Newcastle
Metro station, Birmingham City Centre and First Scotland Rail Stations(RNIB
Business Development Team, 2008)й A primary motivation of RNIB React’s is
to enable social inclusion of the visually impaired. The system is designed to
facilitate independent travel around urban areas and public transport. This
ensures that councils and other authorities fulfil part of their obligations as
outlined by the Disabilities Discrimination Act.
A large advantage of the RNIB react system is that the transmission fob
carried by the user universal across UK locations and as a result a user can
travel from Newcastle to Birmingham and the fob will work in both locations.
Another major advantage of this system is that it may be use to foreign
visitors to the UK. This is because the messages have the potential to be user
specific for instance; Russian tourists could use a specific fob that prompted
Russian messages. One disadvantage of the system would be that it uses
speakers for message playback which need a 120/240v ac 5A-fused power
supply. Consequentially, the system is significant in size and so its location
requires sufficient space and a readily available power supply(RNIB Business
Development Team, 2008). This also requires the RNIB to employ people to
install the units correctly and safely resulting in greater expenditure. Figure
3.16 shows the size of the speaker unit fitted to an electronic bus stop.
Figure 3.16 - RNIB React Speaker (RNIB Business Development Team / SFX Technologies, 2007).
The cost of the RNIB React system is illustrated in the table below and can be
considered expensive in comparison with other available systems on the
market(RNIB Business Development Team, 2008). As is shown, the system
can cost up to several hundred thousand GBP to equip a large facility such as a
university campus.
RNIB React 3 speaker unit
RNIB React RTI speaker unit
RNIB React Trigger Fob
RNIB React Fixing Bracket
RNIB React Message recording +
Implementation up to 50 messages
RNIB React Message recording +
Implementation up to 51 - 100 messages
Price (ex. VAT and
£2000.00 Plus VAT
£3000.00 Plus VAT
£25.00 Plus VAT
£27.00 Plus VAT
£750.00 Plus VAT
£1500.00 Plus VAT
3.3.5 Talking Signs
Talking Signs is a system that reads signs and points of interest out aloud to
users by utilising infrared technology. The project was initiated as collaborative
research between Smith-Kettlewell Eye Research Institute and the
Rehabilitation Engineering Research Centre in San Francisco, California. The
system works using permanently installed transmitters that send infrared
signals to a handheld receiver, which then decodes the signal and outputs it as
an audio voice message to the user. The transmitters are placed at landmark
locations such as information desks, stairs, cash machines and pedestrian
crossings. There, by continuously looping, they send the message to the
receivers aiding user orientation. Figure 3.17 shows a diagram of Talking Signs
implemented in a town centre in which information at a pedestrian crossing
can be sent to the user replacing the “cuckoo and chirp” noises more
commonly used as a method of indicating when it is safe to cross a road(Noyce
& Barlow, 2003, p. 26). This means that information can be sent directly to the
user rather than it being emitted by a speaker and being masked by ambient
environmental noise. One of the major advantages of the system is its high
versatility and ability to work in both indoor and outdoor environments
providing the transmitter has access to a 12VDC current. However, as seen in
Figure 3.18, an urban environment may result in obstructions to the infrared
signal which relies on line of sight communication.
Figure 3.17 - A town centre scene demonstrating how Talking Signs works(TalkingSigns).
Figure 3.18 – Shows a transmission beam hitting pedestrian thus rendering the system non functional. (Pedestrians
from (TalkingSigns)).
3.3 – Summary
The literature reviewed in this chapter gives insight into the functionality of
positioning technologies highlighting both benefits and drawbacks of each. As
such, the literature suggests that although some boast a much greater
accuracy than others (such as ultrasound technologies), at times the cost of
the infrastructure may be too great for widespread installation and deems
some technologies unsuitable for spatial orientation of the visually impaired in
an indoor environment. Other drawbacks revealed include the need for line of
sight between transmitters and receivers, showing that indoor environment
with walls, desks, pillars and office equipment or irregularly shaped rooms may
not be conducive for technologies such as ultrasound. Additional issues of
consideration were highlighted during the review of existing projects and
commercial technologies such as the UCSB PGS which, although highly
acclaimed and advanced, is also ergonomically cumbersome. With one of the
universal motivations for creating assistive technology being the improvement
of integration and inclusivity of the visually impaired into the general public,
wearing such a piece of equipment can prove counterproductive to this aim.
Taking each of these factors into account, a suitable technology to use in
conjunction with existing mobile technology must be selected. The most
accurate technologies are suggested to be Ultrasound and UWB, however the
drawbacks outlined during the review deem them unsuitable for the purposes
of solo spatial orientation of the visually impaired using mobile technology.
Other technologies such as infrared also use line of sight as exemplified by
Talking Signs, but using line of sight in an indoor environment can cause
problems due to furniture, walls, pillars and other obstacles. This means that
infrared alone is not a viable option for spatially orientating the visually
impaired in an indoor environment using mobile technology.
Two technologies that have a relatively cheap infrastructure are GPS and RFID,
with standard GPS being available on most modern mobile devices. However,
as explained earlier, GPS has a problem with indoor use due to its line of sight
nature. Although it is possible for signals to penetrate some materials and
thus, if used under certain circumstances, it may be possible to use standard
GPS for the orientation of the visually impaired within an indoor environment
using the mobile devices built-in GPS. Secondly, the cost of the RFID
infrastructure is relatively low and can be installed within an indoor
environment with relative ease. Using a modern mobile device to process the
signals within the environment and give feedback to the user is possibly a
viable option for spatially orientating the visually impaired within an indoor
environment. Furthermore, using a mobile device as the transceiver means the
device is discrete and manageable in contrast to the more cumbersome
designs previously discussed. This would be conducive to the aim of improving
user integration with the general population. If the mobile device has other
functions such as GPS, then it is possible for a user to use it to guide them into
an indoor environment before an alternative system takes over orientation. To
summarise, it is clear that some technologies lend themselves more
successfully to the task given based on issues of cost, system design and
finally the discreteness of the device. The two that have stood out as potential
systems are RFID and GPS.
Chapter 4 – Mobile Technology
The recent appearance of Smartphones on the UK market with the capability
for allowing applications to run from the mobile device, built-in GPS and
features such as the screen reader for the iPhone (for use with the visually
impaired) has presented several options for the development of an indoor
positioning system. This chapter reviews a variety of current mobile phones
that may be suitable to use with a specialised system for helping orientate the
visually impaired within an indoor environment. Phones will be analyzed for
ease of use for the visually impaired, cost and additional services.
4.1 – Samsung Galaxy S II
The Samsung Galaxy S II is a touch screen phone that has no accessibility
features for the visually impaired making it extremely difficult for them to use.
The phone is free for those who sign to a minimum contract of 24 months at
£41 per month (Vodafone, 2011) or can be purchased outright for
£499.99(Three, 2011). The device can run a multitude of different applications,
however it has been noted that due to the number of different Android
operating system based phones available on the market, some have trouble
supporting a great deal of applications(Popular Mechanics, 2011). This could
inhibit a user successfully obtaining valuable applications that can offer
positioning support to the visually impaired. The device does however have
GPS and comes with a built-in GIS for orientation.
4.2 – HTC Sensation
The HTC Sensation is a touch screen phone which, like the Samsung Galaxy S
II, has no accessibility features for the visually impaired. Due to its complex
nature, it has no standard layout which would prove difficult for visually
impaired users. The phone is free if a contract is taken out for a minimum of
£34 per month for 24 months (Vodafone, 2011). It can also be purchased
without a contract for £449.99 (Three, 2011). Again, and with the same
drawbacks as the Samsung Galaxy S II, the HTC Sensation uses the Android
operating system. An advantage of the HTC Sensation is that it comes with
built-in GPS and GIS allowing a user to orientate themselves within an outdoor
4.3 – BlackBerry Torch
The BlackBerry Torch has both a touch screen for selecting applications and
keypad for typing text which would allow visually impaired users to more easily
use services such as email and the internet as long as they are familiar with a
QWERTY keyboard. An application for the BlackBerry has recently been
released that reads aloud the screen contents which would allow visually
impaired users to navigate the phone’s features with greater ease while using
services such as the built-in camera or the navigation software. This piece of
software is supplied by HumanWare at a single expenditure of £280 and is
called Oratio (HumanWare, 2011). The phone is free if a contract is undertaken
for a minimum of 24 months at £31 a month (Vodafone, 2011). It can also be
bought outright for £414.99(Three, 2011). Including the added cost of Oratio
the phone can be considered expensive for the visually impaired. It does
however have built-in GPS and GIS allowing the user to orientate themselves
within an outdoor environment and this feature used in conjunction with Oratio
could be used to enhance independence of the visually impaired in traversing
outdoor environments. Not all of the BlackBerry applications will be compatible
with Oratio thus limiting it usability reducing the Smartphone in functionality.
4.4 – Apple iPhone 4
The Apple iPhone is arguably the most established Smartphone on the market.
It is a touch screen device but does have a built-in advanced screen reader
allowing visually impaired users to use a wide range of applications and
features. The screen reader narrates the text alongside additional information
such as the battery life, time and position of the icons on the screen. The
screen reader, called VoiceOver by Apple, allows a user to tap the icon to
select an icon and hear its name before tapping again to launch the program
selected (Esquirol, 2011). The problem with locating icons and the keyboard
for creating emails and text messages has since been negated by Bruno Fosi’s
design of silicone casing embossed with the keyboard layout and icon
positioning. Using their haptic sensory channel, the visually impaired may now
locate applications and the keyboard (Fosi, 2008)й Figure 4й1 illustrates Fosi’s
design retaining all screen functionality of sliding and touching.
Figure 4.1 - Bruno Fosi’s silicone case for the iPhone that allows users to feel the screens layout.
Various iPhone applications aimed towards visually impaired users currently
exist and it may be because of this that the iPhone was recently reviewed as,
“the most revolutionary thing to happen to the blind for at least the last ten
years” (VanHemert, 2010). This statement occurred after a reviewer reading of
blind iPhone user Austin Seraphin, and his utilisation of an application using
the iPhone camera to identify colours enabling him to see the sky and his
pumpkin plants for the first timeй Seraphin goes on to say that, “I love my
iPhoneй It changed my universe as soon as it entered itй”(Seraphin, 2010). The
application, called ColorID, takes a photograph of a scene and analyses it
before reading out the colours on screen citing them in a range of descriptions
from “orange” to “atomic orange”й Figure 4й2 illustrates ColorID running a
photograph analysis.
Figure 4.2 - ColourID identifying colours from a photograph taken with the iPhone (Esquirol, 2011).
Another iPhone application for the visually impaired is the money reader,
LookTel. Whilst this application only currently works with US dollars, similar
applications are currently being developed to work with different currencies on
a global scaleй LookTel works by scanning money with the iPhone’s camera,
analysing the picture before announcing to value of the money to the user.
Figure 4.3 shows LookTel identifying some American dollar bills.
Figure 4.3 - LookTel identifying American dollar bills (Esquirol, 2011).
There are hundreds more applications for the iPhone that help the visually
impaired. Audiobooks is an application allowing users to listen to over
2,200,000 audio books for free (Apple, 2011) while the oMoby application uses
the iPhone’s camera to identify everyday shopping items(Apple, 2011). There
are also applications aimed at the general public that in conjunction with
screen reader prove useful to the visually impaired. These include ToiletFinder
UK which locates the nearest public toilet. When considering the multifunctionality of a standard 16 gigabyte iPhone 4, its contracted tariff of 24
months at £40 a month (Vodafone, 2011) or cost of £499.99 without a
contract (Three, 2011) may be justifiable for a visually impaired individual on
DLA. An additional benefit of the iPhone 4 is its A-GPS functionality providing a
higher accuracy than standard GPS. It also comes packaged as standard with
Google Maps to use as a GIS with the option of several other GIS packages
available at varying costs.
4.5 – Summary
Comparatively, most modern Smartphones are similarly priced and all can be
obtained for free by signing a contract with a phone company. Therefore, as
most members of the public already own a mobile phone, adapting a
Smartphone as a wayfinding and navigation device would further expand its
multi-functionality. This information supports the notion that a Smartphone
could be used in conjunction with wireless communication technology to
provide a solution to a commercially viable system for the indoor navigation of
the visually impaired.
Whilst all Smartphones are powerful, some are completely inaccessible to the
visually impaired and so even if these phones have applications aimed towards
visually impaired users, these are redundant if the user cannot navigate the
phone’s menus due to lack of screen readers and assistive technology. As can
be seen in Table 4.4 below, the Samsung Galaxy S II and HTC Sensation do
not come equipped with screen readers nor can one be purchased and
consequently are not suitable options.
The BlackBerry Torch is a phone with GPS capabilities and a screen reader.
However the cost of the screen reader is not favourable in comparison with the
Apple iPhone which has the capacity for a free screen reader with an arguably
better functionality (Schroeder, 2010). Therefore, the iPhone is clearly the
market leader for use with the visually impaired on account of its free screen
reading functionality, multiple assistive applications currently in development
and silicone overlays by the aforementioned Bruno Fosi. This is illustrated in
Table 4.4. The iPhone is a multiple function device that can act as a mobile
phone, navigation aid, internet browser, item identifier and a monetary
identifier. It also supports additional applications that give visually impaired
users the ability to see colours around them and have them described in detail.
This explains why people such as Paul Schroeder, vice president of programs
and policy at the American Foundation for the Blind, praises the iPhone and the
assistance and advancements it is pioneering for the visually
impaired(Schroeder, 2010).
Price (phone
£41 pm
Galaxy S
iPhone 4
£34 pm
£31 pm
£40 pm
Table 4.4 Summary of Smartphone functionality
Chapter 5 – Evaluation of Technology
The information outlined in earlier chapters of this thesis suggests that
obtaining accurate positional information within an indoor environment is a
difficult task. The following chapter documents testing undertaken to
investigate the selected mobile device, the Apple iPhone 4’s ability to help
spatially orientate the visually impaired within an indoor environment using its
AGPS. Based on the results of this investigation, which highlighted the
unsuitability of the iPhone 4’s AGPS, further testing was then undertaken to
explore if RFID technology could provide a suitable alternative (because of the
advantages outlined in Chapter 3)й Here, the chosen RFID receiver’s read
range was investigated to ensure that it was capable of being used in
conjunction with a peripheral system.
5.1 – iPhone Indoor GPS Performance
The first investigation was aimed to test the accuracy of the iPhone’s Assisted
GPS chip by locating a singular position inside various public environments
such as a bus station or a public library. The public environments that were
tested differ structurally regarding the materials they are constructed from and
it was therefore important to test the iPhone’s A-GPS chip’s ability to receive
data using a Pseudo Random Noise (PRN) code in a variety of conditions. The
Apple iPhone is advertised to have A-GPS that, as explained in Chapter 3, uses
a ground station and a PRN code to transmit the same information as the
satellite through numerous obstacles. If the standard analogue satellite signal
cannot penetrate the structural materials of the building or its potential
obstructions, then the PRN code has a greater chance of penetration due to its
robust digital nature. Initial anecdotal reports of the iPhone and its A-GPS
system ascertained from personal use suggest that this system does not
perform adequately in indoor environments and suffers greatly from signal
attenuation, often reverting to the use of its Wi-Fi positioning system. I foresee
the results of this investigation aligning with the aforementioned reports
rendering the iPhone’s A-GPS unsuitable for indoor orientation of the visually
This investigation was initiated by finding suitable locations within public indoor
environments throughout a town centre. The locations (as seen below in table
5.1) were chosen on recommendation by the Batley Blind Association due to
their common usage by the public and the visually impaired. A high accuracy
GPS receiver was used to record the exact latitude and longitude of a specific
point in each location with that position then marked on the ground using a dot
stamp. This point was then referenced against the exact latitude and longitude
as recorded by a custom-built iPhone application implemented in the same
location. Following this process, the latitudinal and longitudinal position of both
sets of data were recorded into a geographic information system in order to
measure the difference in metres between the two points. This data provided
the discrepancy of the iPhone’s A-GPS in comparison to the high accuracy GPS
reciever within the indoor environments.
One problem that may have compromised the data originated from the
iPhone’s Wi-Fi location function. This occured when the nearest Wi-Fi point was
used to calculate the position of the phone because an A-GPS signal was not
available. To negate this problem, Wi-Fi communication was prohibited during
all the investigations forcing the phone obtain its location using other methods.
In order for the iPhone’s A-GPS to be considered accurate enough for use in an
indoor environment, the discrepancy between the device and its high accuracy
counterpart must be suitably low. If the discrepancy between the device and
its counterpart is too great, guiding users in possible confined spaces could
become problematicй It should be noted that the iPhone’s GPS data is accurate
to the fourth decimal and as such, the data of the high accuracy GPS device
has been rounded accordingly. The results of the aforementioned investigation
examining the accuracy of the iPhone’s A-GPS chip can be seen below in table
Location Name
University Building
Students Union
Shopping Mall
Bus Station
Train Station
Actual Location
53.6410, -1.7786
53.6430, -1.7784
53.6449, -1.7794
53.6461, -1.7812
53.6465, -1.7827
53.6457, -1.7829
53.6448, -1.7827
53.6455, -1.7863
53.6486, -1.7847
iPhone Location
53.6374, -1.7944
53.6412, -1.7733
53.6452, -1.7794
53.6458, -1.7812
53.6461, -1.7729
53.6457, -1.7829
53.6461, -1.7729
53.6374, -1.7944
53.6461, -1.7729
Table 5.1 Results of location with both high accuracy GPS and iPhone GPS
As can be seen from the table of results, there was a variable amount of
discrepancy between the GPS locations of the iPhone and the high accuracy
GPS device. A discrepancy range between 14.7m and 1,108.79m shows clearly
that the iPhone 4 cannot be relied upon for accurate indoor spatial orientation.
Furthermore, even if the discrepancy was not over such a wide range, the
smallest discrepancy of 7.78m would not be suitable in the context of
orientation in an indoor environment. In conclusion, this data rules out the use
of the iPhone’s built-in A-GPS to orientate the blind within an indoor
The iPhone was chosen as a multi-function device based on its commercial
viability. Although, there are some GPS receivers on the market that boast
indoor functionality on account of their ability to track extremely weak signals
alongside using additional algorithms that help reduce the multipath effect and
potential signal reflections. One such example would be the aforementioned
uBlox SuperSense (uBlox, 2011). However, due to the high cost of the uBlox
chip, the net cost of a working product may prove beyond the realms of
commercial viability especially considering the average income of an individual
on disability benefits. In conclusion, on account of the fact that the Broadcom
BCM4750 (Chipworks, 2011) (iPhone’s standard GPS chipset) does not have
the required accuracy to safely orientate the visually impaired within an indoor
environment, an alternative solution must now be identified.
Figure 5.2/5.3 – Showing the use of GoogleMaps to work out straight line distance between points (Google, 2011).
Figure 5.4 – Showing the use BROADCOM BCM4750 Chipset in the iPhone4 (Chipworks, 2011).
5.2 – RFID Receiver Read Range
There are several frequencies of RFID, however to keep costs to a minimum,
low frequency passive RFID tags running at 125khz were selected. An RFID
reader must also be selected so that its data can be sent to the Apple iPhone
for interpretation. The RFID reader chosen was the ID-12 Innovation chip that
is retailed by SparkFun Electronics (SparkFun Electronics, 2011). The aim of
this investigation was to evaluate the distance from which the chosen receiver
can read RFID tags. The reader specification states that the device has a read
range of 10cm (SparkFun Electronics, 2011). Therefore this investigation
aimed to test that the specified read range upholds when the device is
implemented in a peripheral system. It is to be noted that various methods of
RFID card placement within the peripheral system were considered and that
attachment to the tip of a blind cane with tags hidden under flooring was
deemed the most efficient. Furthermore, the investigation acted as an initial
testing of the overall functionality of the device.
To effectively evaluate the read range, it was measured at a number of
different angles and through a multitude of materials such as carpet, laminate
flooring and wood flooring alongside a control condition using no material
obstacle. This aimed to imitate the real life scenario of a visually impaired user
scanning with a cane. In the context of the investigation, this was the receiver
system being used from a variety of different angles and through various
materials. Figure 5.6 demonstrates the specific angle positioning of the RFID
card in relation to the receiver. A controlled testing environment was designed
using a 30cm rule secured to a wooden base with the device reader secured at
the 0cm mark (see Figure 5.5). An RFID card was then attached to a purpose
built slide, allowing for accurate intervallic movements of 1mm. In order to
obtain the average range, the cards were moved at 1mm intervals towards the
ID-12 chip from 15cm onwards, through the selected test materials. The ID-12
was then linked up via the peripheral system to the iPhone, which was
programmed to play an alert tone when the chip had been detected. Once the
chip had been detected, the distance was recorded and the process repeated.
One problem that may have occurred during the investigation was that the
acquired batch of manufactured cards could have been defective, therefore
affecting the read range of the peripheral system. To ensure that the results
for the investigation were not biased according to a single batch of RFID chips,
five separate cards from five separate batches were used alternately to ensure
accurate data was collected. It should be noted that when placing the material
obstacles in front of the RFID reader, the material’s reverse side was placed
flush to the chip before recalibrating the measuring system to account for the
added thickness of each different material that was measured.
Figure 5.5 – Diagram of experiment configuration.
Figure 5.6 – Diagram of experiment angles and positions.
Singular intervals of 1mm were deemed suitable in the investigation as they
provided sufficient accuracy to give useful end data but without being so
diminutive a unit that it would not have an active effect on the system. To gain
an accurate average, the readings were taken 25 times at each angle through
each material respectively. Once the experiment had been conducted, the data
was analysed and a mean value for each angle within each material calculated.
The data recorded was not anticipated to have a large enough deviation from
each another to warrant using mode or median averages and therefore
calculating a mean average was deemed sufficient to show the differences
between angles and materials. A desirable outcome would show that the
average read range at different angles and through different materials is as
close to 10cm as possible, with the ideal outcome being that all eventualities
have a maximum read range of exactly 10cm. However, if the investigated
read range does not align with the 10cm specified by Sparkfun, it is anticipated
that the system will still be viable. This is because, even with a small read
range, the cane would still function successfully as the electromagnetic field
produced will still contact the RFID card allowing it to transmit its data
After the investigation was conducted (data in Appendix B), the following
graph was generated using the mean averages of the data collected
ascertaining the read ranges of the RFID cards through a number of materials
at given angles. The red line indicates the read range as specified by the
manufacturer of the ID-12 RFID reader.
Figure 5.7 – A graph to show read ranges of RFID chips through various materials at a number of specified angles.
The initial test was conducted using no material obstacle. The graph shows
that the specification of 10cm read range given on the website of the chip
manufacturer was not accurate in this case, with a minimum deviation of
3.8cm. As discussed prior, it is still possible for the peripheral system to be
successful with read ranges smaller than 10cm. With no obstacle, the lowest
read range recorded was 3.2cm, which is still adequate enough for a user
using the RFID device to read the tags placed on the floor. A standard card
read would most likely be “Angled Centre” or “Angled Centre Left/Right” in
which case the mean range tested at 5.5cm or 4.2cm respectively, which is
more than adequate for the purpose of the peripheral system.
Testing on both 4mm reinforced plastic flooring and 5mm wood flooring yielded
very similar results, which also correlated to the initial results gained without
material obstruction as discussed prior. Both materials cause little to no
disturbance of signals by the RFID reader. This suggests that both material
types should work effectively with the RFID reader.
As can be seen from the graph at Figure 5.7, the marble flooring caused a
decrease in the read range. This is because the material affects the penetration
depth of the electromagnetic field given off by the RFID reader. Despite the
reduction in read range, marble flooring is still adequate for use with the
peripheral system. However, these results highlight the fact that some
materials can cause the penetration depth of the electromagnetic field to
diminish. As a result, penetration depth is a concern when placing the
peripheral system in environments with metallic robust flooring. It is possible
that this type of flooring material may detrimentally alter the penetration
depth of the electromagnetic field thus rendering the peripheral system
inefficient. To investigate these concerns further, metallic steel flooring (as
used in industrial sites) was tested to ascertain if penetration depth of the
electromagnetic field would be affected.
As Figure 5.7 shows, the metal flooring did not allow the electromagnetic field
of the RFID reader to penetrate. This proved to be the case in all eventualities
of angle placement of the cards. It should be noted that the primary use of
steel flooring is for high traffic areas where durability is key, for example
industrial warehouses. Although it must be stated that it is possible that a
visually impaired person may want access to such an environment, a primary
aim of the system is to be installed in commonly used indoor environments.
Figure 5.7 clearly shows a trend in the read ranges at the different angles
through the range of materials tested. This excludes 2mm steel flooring as an
exception to the trend due to it being non penetrable by the RFID receiver.
Head on centre and angled centre show the highest read ranges throughout all
materials with head on left/right and angled left/right consistently showing
similar values but with a decrease of 2cm compared to the centre readings.
This can be equated to the electromagnetic field not contacting the copper coil
inside the card as efficiently when the card is held at an extremity, figure 5.8
demonstrates this.
Figure 5.8 Magnetic field induction on RFID cards at different angles.
This investigation highlights that it is possible for the system architecture to
function correctly in a multitude of different environments and angles.
Although when embedded behind certain materials the read range of the
system is adversely affected, it still functions to a level that would allow the
system to operate correctly. It also highlighted that certain flooring materials
such as metal will cause the system to cease functioning. However, it was
concluded that as the target installation environments would not commonly be
constructed from such materials, the system should be able to be successfully
installed. From this, the project can progress onto the next stage in which this
information will be used to design and engineer a commercially viable and
accessible system to aid the visually impaired with indoor navigation and
Chapter 6 – Design of a Peripheral Device
Using information derived from the previous chapters, a design for a peripheral
device to be attached to a mobile phone for the use with spatial orientation
and the creation of cognitive maps for the visually impaired will be outlined.
Used in conjunction with the mobile device, the peripheral system should be
able to identify landmark positions within the indoor environment, identify
obstacles such as pillars and stairs and inform the user of each room they pass
as “Computer Laboratory” or “Gary Jones’s Office – Head of Computer Games
Programming”й The device should also be able to identify supplementary items
or objects such as artwork, emergency posters, fire extinguishers and
windows. The system will be referred to from this point onwards as the ASOVI
system (Audio based Spatial Orientation for the Visually Impaired).
6.1 – Technology Used
Chapter 3 outlined and analysed different wireless communication technologies
and current systems on the market utilising these technologies. It was
deduced that RFID is conducive for use with a mobile device. Strategically
placed RFID tags should be able to identify landmarks, rooms, obstacles and
any other points of interest such as works of art, water fountains etc. It also
has a relatively cheap infrastructure allowing buildings to be appropriately
labelled for a small financial outlay.
Chapter 4 outlined and analysed available Smartphones currently on the
market. The Apple iPhone 4 is the most easily accessed by the visually
impaired and boasts a wide range of applications with the potential to prove
most valuable to visually impaired users. As such, a combination of the Apple
iPhone 4 and RFID will be used to create the ASOVI system attempting to
spatially orientate the visually impaired within an indoor environment.
6.2 – Design of the Electronics
As outlined in the earlier chapter, the 125 KHz ID12 RFID receiver was
selected (SparkFun Electronics, 2011)й The chip’s features are outlined on the
company’s website as:
• 5V supply
• 125 kHz read frequency
• EM4001 64-bit RFID tag compatible
• 9600bps TTL and RS232 output
• Magnetic stripe emulation output
(SparkFun Electronics, 2011)
The chip is to be soldered to a breakout board in order to then solder it
successfully into a working circuit. Breakout boards are used in electronics to
allow wires to be hooked to the chip or device that then send the signal onto
other chips or devices (CNC Router Source, 2011).
The battery power of an iPhone is not sufficient to sustain the 5V supply of the
peripheral system and so an additional power source must be considered.
Several battery packs are available that are capable of generating the specified
5V however the need for the device to be small and lightweight must be
considered in order to maximise the comfort and discretion of the end user. As
such, an AA battery pack that uses a 5V DC to DC step-up to take the 3V
outputted by the two AA batteries and step it up to 5V was selected.(SparkFun
Electronics, 2011).
Once the system was engineered to have the ability to read data, it must then
be capable of transmitting this data to the iPhone. The built-in serial port of
the iPhone allows communication between peripheral devices and the phone.
The serial communication uses 3й3V which conflicts with the RFID chip’s output
at 5V and so connecting the two would cause damage to the iPhone. A level
converter to step down the voltage level from 5V to 3.3V is therefore required
to prevent this circumstance (SparkFun Electronics, 2011).
Once the level converter has been connected, a means of sending the data to
the iPhone must be implemented. This is achieved by reconfiguring a standard
USB iPhone or iPod data cable. The head of the cable must be stripped to
reveal a 30 pin connector, with 4 pins attached to the 4 cables in the USB wire.
Figure 6.1 shows a diagram of the 30 pin connector.
Figure 6.1 – A diagram of an iPhone cable 30 pin connector (Pinouts.RU, 2010).
Using the configuration specification found at Pinouts.RU, the 4 cables are
removed from their current positions and reconfigured into the following
Pin 1 for ground cable (GND)
Pin 13 for the serial data receive cable (Rx)
Pin 18 for the 3.3V power cable (+)
The fourth pin is removed as it is redundant.
The USB head on the opposite end of the cable is then removed in order to
place the wires through the appropriate holes on the level converter. This will
allow data sent through the converter at 5V into the cable, which operates at
The ASOVI system was then soldered with the addition of two wires (from pin
7 to pin 1 and from pin 11 to pin 2) to the RFID breakout board. This is to
allow the data to be outputted in ASCII (American Standard Coding for
Information Interchange). Finally, a switch was added to allow the power to be
turned on or off to conserve battery life. Figure 6.2 shows a diagram of the
Figure 6.2 - A diagram of the circuit used for the peripheral device.
Figure 6.2 illustrates that the battery pack provides power to the ID-12 RFID
breakout, soldered into the RFID receiver chip. This in turn powers the chip
which transmits its data to the level converter as well as passing current and
ground cables through to it. The converter then steps the power from 5V to
3.3V and sends the data through the Rx (receive) channel up into the iPhone
along with the 3.3V power and ground.
6.3 – Design of the Application
Data is then to be transmitted into the serial port of the iPhone from the RFID
reader. This process requires the bespoke coding of an application to open the
serial port and listen for any data being received. However, the iPhone
software development kit uses Objective C, which lacks the specific
functionality needed to access the serial port. As an alternative, a framework
called OpenFrameworks is used. OpenFrameworks allows the coding language
C++ to be used and its functions accessed easily. Using this, an application is
then created to allow a serial port to be opened and listen for incoming data
which is then dealt with as necessary. The code for the program can be found
in Appendix A.
The 125 kHz tags in use each have a 20 digit unique identification number that
is broadcast to the reader. The screen of the iPhone outputs this identification
number so that it can be noted. Once this process is complete, it is hard coded
into the application. The phone can then output a specific response if a specific
identification number or a set of running identification numbers is scanned. For
example, a loud warning sound could emit to advice the user of any imminent
obstacles followed by speaking the name of the imminent obstacle whilst the
phone vibrates to warn the user.
6.4 – Physical Design
Low frequency RFID has a limited broadcast range and so a method must be
created to ensure that the signals broadcast by the strategically placed tags
are successfully received by the reader. This method must allow the user to
traverse indoor areas and easily detect signals from tags placed at obstacles,
doors and other points of interest without the need to move the reader
The proposed solution was to modify a standard guide cane used by the
visually impaired to scan areas for imminent obstacles such as steps and
curbs. The modification would allow the RFID reader to be situated on the end
of the cane in order to detect tags placed on or under the floor (depending on
the building itself). This will allow the RFID tags to be detected by the cane,
which will then send the data to the user’s iPhoneй Figure 6.3 shows a design
for the modified cane. It should be noted that the wiring shown in Figure 6.3 is
not accurate and purely symbolises a completed electrical circuit. The diagram
illustrates the position of the RFID reader on the bottom of the cane, the
position of the battery pack and a holster in which to place the iPhone. This
holster allows the user to use the cane without having to carry the iPhone
separately. It also allows the vibrations of the phone to transfer through to the
cane giving haptic feedback to the user.
Figure 6.3 - The design of the modified RFID device.
Figure 6.4 shows how the cane would react when a user is in danger of
colliding with an obstructive obstacle, such as stairs or a plant pot. In the
diagram, the receiver induces the tag to send it the unique identification
number. This in turn is transmitted to the iPhone which then processes the
data prompting an immediate response. As a result, the iPhone will vibrate and
emit a loud auditory warning alerting the user to an imminent obstacle. Both
haptic and audio information has been implemented aiming to bridge the
sensory gap due to the visual impairment of the user. Figure 6.5 shows tags in
an arc formation that will be situated at points of interest such as doors,
artworks and fire extinguishers. Because the tags are not directional, the arc
formation is used to enable a user to detect the direction of their traversal and
which side of a corridor a door may be situated, for instance. This is due to the
fact that as the user scans with the cane, the tags will only be activated if the
cane is on the correct side to receive data from the point of interest. The
design also incorporates a large enough projection of the arc from the point of
interest to allow the user to easily scan it without excessive searching.
Figure 6.4 - RFID device heading towards an obstacle.
Figure 6.5 - Arcs of tags showing points of interest.
Chapter 7 – Evaluation
The following chapter evaluates the ASOVI system (as outlined in chapter 6)
and its ability to aid the visually impaired in the navigation of indoor
environments, aid in the creation of cognitive maps and finally, its general
reception with the target market. It outlines the investigation conducted with
an end-user group of participants before analysing the observational data
recorded and the subsequent feedback gained during qualitative focus group
discussions. This will be followed by outlining and analysing the investigation
designed to test the ability of the ASOVI system to aid in the creation of
cognitive mapping in the visually impaired. Both investigations will provide
information to aid the future development and refinement of the ASOVI
7.1 – ASOVI’s Ability to Aid in the Navigation of Indoor Environments
and its Reception with the Target Audience
The goal of this investigation was to test the ASOVI system’s ability to help a
user navigate an indoor environment and locate specific landmarks or points of
interest. Furthermore, the experiment aimed to highlight any successful points
or problems with the ASOVI system by means of observation of the
investigation and focus group discussion.
Based on the research summarised in Chapters 2 and 3, the ASOVI system
should successfully be able to aid users in their navigation of indoor
environments. The end goal is to use a combination of auditory and haptic
means to provide visually impaired users similar information gained by sighted
individuals using their visual channel. Secondly, the ASOVI system combines
the use of a standard blind cane with RFID capabilities. This function allows for
the user to utilise the standard functionality of a cane for general obstacle
avoidance, such as another individual moving along the same path, alongside
the aforementioned information acquisition. It should be noted that the ASOVI
system uses the same functionality, such as haptic and audio feedback, as
other successful commercial systems. However, by using techniques that are
already successful but applying them in a different context and combination
with a refined systemic methodology, more successful results should be
The investigation was initiated by taking the schematic of the designated test
environment in order to locate and mark potential obstacles and points of
interest. The selected target environment for the test area was the fourth floor
of the Canalside West building on the University of Huddersfield Campus. This
particular area was selected on account of the high level of points of interest in
one environment allowing for optimal test data to be collected. Once the
testing environment had been designed, measurements were taken and the
number of RFID cards to mark the area calculated. The cards were then
scanned and their unique identification numbers placed into the iPhone
program so that each card would react upon scanning. A voice output was then
created to give information, instruction or alert upon the scanning of any card
in a particular card ark.
Figure 7.1 shows the schematic of the floor and the proposed arcs and lines of
cards. Table 7.2 describes the different obstacles and points of interest that
were marked on the schematic paired with their respective instructions or
Figure 7.1 Schematic of floor four Canalside West with mapped card arc locations.
Following preparation of the environment which involved placing the tags in
their designated locations, a trial investigation was implemented to ensure that
each of the card arcs functioned correctly, sending both haptic and audio
feedback to the user ensuring optimal data collection and user safety. Once the
integrity of the system had been confirmed, the participants were taken into
the environment and situated in the room CW4/07. A number of test arcs were
also installed in room CW4/07 to allow the participants to become accustomed
to the ASOVI system in a controlled environment. This allowed for the
participants to become familiar with the vibrations sent down the cane and also
the style in which the audio instructions and observations are relayed to them.
Once the participants were comfortably accustomed with the ASOVI system,
they were then instructed to begin the investigation.
Table 7.2 Obstacles / points of interest and the instruction given by them.
The investigation began by asking a participant to find a room, piece of
artwork or any other obstacle or landmark. This aimed to imitate real life in
which people would normally have a point of departure and destination to a
journey. During this process, observations were recorded on the participant’s
progress noting how confidently they use the system, any hesitations or
problems that occur and finally how successfully and efficiently they traverse
to their target. This process occurs three times, each with a different target
destination before the participant was once again seated in CW4/07. The entire
investigation was then repeated with all participants before a focus group
discussion was initiated to gain the participant’s personal feedback on the
ASOVI system. A focus group discussion was chosen to allow qualitative data
to be collected by allowing participants to share and build upon each other’s
One of the problems anticipated during the investigation is due to the RFID
cards being unable to be placed under the flooring of the environment. In a
real life situation, this would be desirable but as this would incur construction
work to achieve in the test environment, this is not feasible. Therefore, it
should be noted that the environment does not completely represent a finished
installation of the ASOVI system. However, as shown in Figure 5.5, the system
works adequately through the non-metallic materials tested. The intended
environment uses stone and carpet flooring indicating that the RFID receiver
should have sufficient penetration depth if the infrastructure were fully
installed. Another anticipated problem was that participants may prefer the use
of their own personal cane. This is because canes for the visually impaired
come with a variety of tips from large balls to straight nibs. Each user will have
their own preference of tip and each cane is also size specific to the user’s
height. This means the custom-built cane that was modified as part of the
ASOVI system is rendered useless. To negate this problem, a separate system
was assembled that could be easily affixed to each individual’s specific caneй
Once the investigation was complete, the observations were analysed and
cross-referenced. Analysis comprises of identifying any trends, similarities or
patterns as observed during each participant’s completion of the taskй For
example, if more than one participant find some of the audio information
overbearing and that this caused them consequent difficulty or disorientation,
it may be concluded that this problem would be encountered en mass
prompting further analysis, implications and possible solutions to be noted. The
focus group feedback was also analysed in a similar manner to the observation
data, identifying trends, similarities and patterns before looking for possible
solutions and implications of given feedback.
Upon completion of the investigation, the following conjectures were drawn
from the noted observations (Observations in Appendix C). During the
investigation it became apparent upon observation that each of the
participant’s RFID receiver was snagging on the floor of the environmentй As a
consequent of this, the participants often mistook this snagging for an obstacle
causing false positives to occur in the cognitive mapping process. This
suggests that the cane needs to be physically modified so that the RFID
receiver is situated inside the tip of the cane allowing the ball on the end to
rotate and maintain constant contact with the floor. This should be noted when
modifying the design for the cane in future development.
Cross-referencing observations made for each participant confirmed that, on
occasion, the participants dislodged RFID cards from their position on the floor
with their cane. It is clear that this was due to the position of the RFID cards
on top of the flooring material. As the intended design would place the
infrastructure under flooring material, this should not cause a problem in a real
life situation.
Another similarity observed in all participants was that as they spent longer
using the ASOVI system, their confidence increased in traversing the
environment. This could be attributed to initial nervousness regarding an
unknown situation with an unknown device that they are reliant on for safe
orientation and wayfinding. Upon full accustomisation of the user with the
system, suitable trust is established as the user gains confidence. It could be
argued that this would be the case for any method of navigation new to the
user, such as a guide dog or long cane.
It was commonly observed among the different participants that they
occasionally failed to scan card arcs on account of the methodology of their
cane usage in the environment. The card arc design for the placement of the
RFID cards was intended to be a solution for directional problems. For
instance, a user hitting a straight line of cards would want to know which side
of a corridor a door was on, and for this to occur, information such as “door on
left” would have to be conveyedй This information would be problematic if a
user encounters the line of cards from the opposite direction due to the
aforementioned door ‘on the left’ now being on the user’s rightй As a result, a
false cognitive map would be created. However, as a result of the potential for
card arcs to be missed in some cases, a straight line system must be
implemented. The solution to the previous problem posed by a straight line
design would be the use of triggers in the programming of the system. This
would mean that when a user encounters a line of tags, the system would log
the scanned line and then upon encountering the next line of tags it will be
able to determine which direction the user has arrived from and give the
correct information – in this instance, either “Door on left” or “Door on right”й
This suggests that using lines of tags across the corridor or protruding from
any landmark would be more effective for orientation and the method should
be adopted during further development of the ASOVI system.
Another common observation is that card arcs were missed because some
participants scanned the environment at great speed. This suggests that when
the receiver is moved at pace, it does not have enough time for its
electromagnetic field to pass sufficient power to induct the card to then pass
its data to the receiver. This shows that although the read ranges of the RFID
receiver seemed to be acceptable, as shown in the earlier investigation, this
range may drop when the system is moving in situ. A possible solution for this
would be to use a RFID receiver with a larger read range meaning that the
larger electromagnetic field will be in contact with the RFID card for longer
thus allowing it have the power to send its data to the receiver, even when the
system is used at speed.
Finally, a correlation was noted in observing that participants were successful
in locating targets during the task. This suggests that the system can
successfully aid in the traversal of environments by alerting users to landmarks
and features that are not normally apparent. Although, there were variations in
the time it took each participant to correctly identify target locations. This
implies that some of the slower timings were not caused by problems with the
ASOVI system but by participatory discrepancies and accessibility of selected
locations. It is also possible that these timing differences would be apparent in
sighted individuals when trying to find different locations within an indoor
environment, with the more perceptive individuals noticing some of the more
obscure and less accessible locations faster than others.
Following completion of the task, the focus group discussed feedback on the
system. It should be noted that although the focus group was question
initiated, the general discussion did not follow a structured path and
subsequently, responses have been divided into categories and analysed
7.1.1 Information feedback from the device
The response from the participants with regard to the audio and haptic
feedback suggested that some improvements could be made to the design of
the ASOVI system. During the discussion it became apparent that the
participants thought the ability to increase the volume of audio feedback would
be essential in busy, loud environments. However, there were concurrent
concerns about the system’s audio distorting or completely masking the
ambient noise in the environment that is often used as a source of orientation.
As a proposed solution to this, wireless headphones that allow cleanly audible
sound to be heard while simultaneously allowing the audition of ambient noise
could be implemented in future development.
Another issue was discussed when one member of the group observed that
“sometimes the phone was still playing the last instruction when I found a new
object”. This suggests that audio overlap occurred due to the close proximity of
the card arcs. The participants then proposed a solution stating that they felt
the speed of the audio could have been increased, allowing for multiple arcs in
close proximity to be scanned without this audio overlap. This modification
would have to be executed with particular care as to avoid the audio becoming
unintelligible. And so, as a proposed solution to this problem, the information
conveyed at each landmark or point of interest could be reduced in length by
conveying only vital information in a concise manner.
During the focus group’s discussion of haptic feedback, one participant stated
that if they “know when an audio message was coming” they were prepared to
focus on the audio feedback. They went on to explain that if an auditory
warning was transmitted, it was much more clear what obstacle was
approaching rather than the “guesswork involved with vibrating products”й This
feedback has large implications for the ASOVI system on account that if its
vibrations are redundant, the mobile device can be detached from the cane
and be stored elsewhere. This detachment of the mobile device reduces the
weight of the cane, making it closer to that of a standard cane. This would
relieve users from manipulating extra weight repeatedly during traversal. Also,
this detachment would reduce the risk of theft involved with the public display
of an expensive piece of technology.
7.1.2 Alternate delivery methods and mobile devices
During the discussion, feedback on the delivery method of a cane and the
mobile device was requested. One participant identified themselves as a
qualified cane instructor and then went on to say that “for the majority of
people that mobilise, the cane is their first choice”й This suggests that the cane
is possibly the best method of delivery for the ASOVI system. However, it
should be mentioned that another participant was primarily a guide dog user
and although capable of using a cane, they primary preference for orientation
and wayfinding was with the use of a guide dog. It was mentioned by this
particular participant that although the system could help identify and give
information that their dog could not, they would not return to using a cane as
it is unable to offer “the companionship of a dog”й This lead to the suggestion
of a clip-on ASOVI device, which could be attached to a dog’s harness or the
shoe of a user. Taking into account the earlier proposal of detaching the mobile
device from the cane, the ASOVI system has the potential for multiple
application methods increasing its accessibility to the target audience.
Following the discussion on delivery methods, the use of the Apple iPhone was
considered. A unanimous agreement between the participants was reached
that if a user already has an iPhone then it would be “perfect”, however many
of the target users may already own one of the other smartphones available on
the market and may therefore be reluctant to purchase another with the sole
purpose of use with the ASOVI system. It was then suggested that making the
ASOVI system compatible with multiple platforms such as “Android” would
make it more appealing to the target user base.
7.1.3 ASOVI Marketability
When the participants were asked about the comparison between the ASOVI
system and other market competitors, the common consensus stated that
ASOVI was superior and had a significant amount of potential. One participant
commented that “…other systems alert you to approaching things but they
don’t tell you what it is” whilst another commented that the ASOVI system
“makes you feel a lot more comfortable” than the market competitorsй This
suggests that the ASOVI system compares well to the other technology
available on the market and that it could possibly prove more successful than
these products.
Furthermore, the participants were asked if they would trust the system to
orientate them and aid in the traversal of an environment. Although each
participant articulated this differently, they were all in agreement that a largescale investigation must be performed where full building traversal and larger
test group was availableй One participant commented that, “You have always
got to know where you are, however it has the makings of a brilliant prompt to
assist in getting from A to B”й Another commented that, “going out of one
building and next door” would allow them to become more trusting of the
system as it would help highlight any potential weaknesses. It may therefore
be deduced that further investigation should be conducted over a larger test
The participants were questioned whether they would purchase and use the
system if it became commercially available. One participant responded that
they would, “definitely use the system” with another agreeing that it would,
“definitely improve mobility around a building”й This concludes that the system
was well received by the participants and suggests that the system would be
commercially viable given the correct testing and marketing in order to enable
it to compete with current technologies available on the market.
7.1.4 Requested additional features
In conclusion to the focus group, participants were invited to suggest
additional ideas and improvements to the ASOVI system. In response, they
voiced that the system conveyed too much information. For example, one
participant said, “you had a fire extinguisher down there, if there’s a fire the
last thing I want to do is fight itй I’m running anywhere I can”й Another
participant passed comment on the artwork on the wall saying, “The artwork’s
all good, but if I’m in a hurry I don’t want to take note of the artwork anymore
than you would if you were in a hurry” further stating that, “…if I was waiting
about then maybe I would, so having a way of changing between levels of
information would be good”й In response to this observation, the newly
released iPhone 4S and its voice recognition functionality was highlighted and
when discussed further, the participants unanimously suggested that they
would prefer a method of asking the ASOVI system to direct users to
landmarks within buildingsй One participant stated, “If you could ask it to take
you to the front entrance or whatever it would be amazing”й This suggests that
the ASOVI system should utilise the iPhone 4S “Siri” voice recognition software
to allow a checkpoint system that can guide a user, step-by-step, to their
intended goal. This voice recognition software would also allow users to select
the level of detail conveyed by the system to allow superfluous features such
as artwork and points of interest to be optional. Another advantage of this
feature would be that, in the case of an emergency, the user would only
receive information guiding them to the most efficient escape route.
To summarise, the investigation has highlighted that the ASOVI system can
successfully aid the navigation of the visually impaired through an indoor
environment. It has also shown that the participants from the target user
group received the system well and that if developed further, they would
purchase and use the system to aid navigation through indoor environments.
7.2 – ASOVI’s Ability to Aid in the Creation of Cognitive Maps
The goal of this investigation was to test the ASOVI system’s ability to help a
user create a cognitive map of an environment. Currently, there exists no
standardised methodology to investigate if a successful cognitive map has
been created within a participant’s mind. While various approaches have been
taken to do so for example (Lotfi & Sanchez, 2009) get participants to recreate
the area using building blocks, nothing has been standardized. Consequently, a
methodology must be designed to investigate successful cognitive mapping.
Previous investigation has suggested that the ASOVI system can successfully
aid the visually impaired with wayfinding and orientation. Once information
about an environment has been successfully conveyed using the ASOVI
system, as highlighted in the earlier investigation, it should be retained in the
mind as a cognitive map. The amount of accuracy and detail of the cognitive
map should be positively correlated to the area explored and time spent in the
environment. However, due to the number of willing participants it was unable
to take separate test groups of participants and allow them different amounts
of time in the environment to test this theory. It should also be noted that
some fluctuation of results between participants can be expected on account
that even the sighted population’s ability to create cognitive maps varies from
person to person.
The investigation began by taking each participant into a separate office to
ensure that the other participants did not overhear the questions or answers
given which may have resulted in them inadvertently adding additional data to
their cognitive map. The participants were asked to describe a route to a
specific location in the test vicinity using data ascertained by the ASOVI
system in the previous investigationй The participant’s answers provided data
that will then be cross-referenced to determine whether the system succeeded
in aiding the creation of cognitive maps.
One foreseeable problem with this method of investigation is that it does not
allow the testing of distal reading within the cognitive mapping process.
However, as the ASOVI system does not give distal data in audio or haptic
format, it still relies on the user’s human dead reckoning system as they are
traversing the environment. This means testing of distal reading within the
cognitive mapping process is extraneous to the ASOVI system’s ability to aid itй
Once collected, the data was cross-referenced between participants and
analysed quantitatively in graphical format for the number of correct locations
identified. Two graphs were plotted; the first shows each participant’s correct
and incorrect answers against each other in an attempt to quantify the
accuracy of the cognitive maps created and the second shows the question’s
success rate against the other questions in order to highlight any particular
area that the participant group struggled or excelled at in the test
After the investigation was conducted (Questions in Appendix D), the following
graphs were generated using the amount of correctly answered questions.
Number of Correct Answers
Correct Answers
Incorrect Answers
Participant Number
Figure 7.3 A graph to show the number of correct and incorrect answers by each participant.
Number of Correct Answers
The graph in figure 7.3 shows that, apart from participant 4, each of the
participants answered more questions correctly than incorrectly. This suggests
that the ASOVI system does aid in the creation of cognitive maps. Participant 4
only answered one question correctly and it is likely that this is due to the
individual traversing the environment rapidly resulting in the system being
unable to read any card arcs. Subsequently, the ASOVI system was then
unable to convey the information back to participant 4, as discussed in the
previous investigation. The first three participants have a very similar quantity
of correct and incorrect answers. This implies that the fourth participant is an
anomaly and their lack of cognitive mapping of the test environment is due to
the speed of their movement and cane technique.
Number of correct
participant answers.
Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8
Question Number
Figure 7.4 A graph to show the number of correct answers on each question.
The graph in figure 7.4 shows that the average number of correct answers per
question is two. This equates to half of the participant group being able to
correctly answer the questions, taking into account the fact that one
participant did not receive full environment information. Furthermore, this can
then be roughly equated to state that two thirds of the participants
successfully used the ASOVI system to aid in the creation of cognitive maps.
Question 5 is a clear anomaly due to its large deviation from the rest of the
resultsй Question 5 asked for the route, “From Lifts to the Security Phone”й The
security phone card arc was nestled in between CW4/01 and the first-aid box.
This may indicate that the participants did not scan the card arc when
traversing the area or that there was too much information placed in such a
small area. This issue could be solved by changing the card arc system to
straight lines, providing a longer time between the points of interest, as
proposed in the previous investigation. It may also be the case that the
questions that were answered incorrectly by participants would have yielded
correct answers had the participants used the system in a particular
environment for an extended period of time and on multiple occasions as
would be likely in the real life implementation of the system as previously
Overall, the data gathered from the questions posed to the participants
suggests that the ASOVI system (although dependant on some minor
amendments as discussed) can help with the creation of cognitive maps.
Chapter 8 – Conclusion and Further Research
This chapter will conclude the thesis by summarising and evaluating the
research that culminated in the development and testing of the ASOVI
system. It will then continue to outline further research and development as a
result of these findings.
8.1 – Summary of Findings
It was established in section 1.1 that there are currently no commercially
viable and accessible systems available to aid the orientation and wayfinding of
indoor environments for the visually impaired. It was consequently proposed
that a possible solution would be to develop a system that would combine
multi-functional devices such as mobile phones with wireless communication
technology as this would negate the problems presented by the indoor use of
GPS alongside issues of cost and accessibility.
Research conducted in Chapter 2 regarding wayfinding and cognitive mapping
concluded that although the visually impaired are successful in creating
cognitive maps using other available sensory channels, this does not equate
the cognitive mapping capabilities of a visually able individual. As such, the
visually impaired are capable of navigating an indoor environment although
this capability is improved with the use of navigational aids. Further research
also suggested that these cognitive mapping abilities improved in a positive
correlation with the length of time spent in any one environment.
Chapter 3 continued to review wireless communication technology and its use
for orientation and wayfinding. It was ascertained that at present, many
systems are not wholly successful on account of cost efficiency and ergonomic
design. Furthermore, many are not suitable for indoor orientation of wayfinding
for the visually impaired. Having analysed the available wireless technology, it
was concluded that GPS and RFID infrastructures presented two possible
avenues of investigation on account of their accuracy, cost and compatibility
with mobile phones.
Research in Chapter 4 was used to identify which mobile phone would prove
the most conducive platform to design the ASOVI system for. The Apple iPhone
4 was highlighted as having high levels of multi-functionality, free screen
reading and an overlay already in existence to aid its use with the visually
impaired. Having identified the iPhone as the most suitable multi-function
device, testing was designed to establish if the AGPS functionality of the
iPhone would be sufficiently accurate for indoor orientation and wayfinding for
the visually impaired. Testing was conclusive that the iPhone AGPS was not
suitable for this purpose which lead to the use of RFID technology in the
subsequent design of the ASOVI system. Before the ASOVI system could be
designed, testing was undertaken to assess the functionality and application of
the RFID reader in order to ensure the system would function when
implemented in a real life situation. The read range of RFID system was
investigated with the receiver placed behind a variety of materials (to simulate
the potential flooring materials that may occur in the real life implementation
of the system)й The results of this investigation showed that the system’s read
range, although below that of the manufacturer’s specification, would be
suitable for the ASOVI system and its infrastructure.
Using the research from the previous chapters, the ASOVI system was then
designed and engineered. The ASOVI system was created using RFID to allow
users to locate points of interest and landmarks. This would aid in the
orientation and wayfinding of the user while further facilitating the creation of
cognitive maps. An RFID receiver was placed on the end of cane, enabling it
to receive and transmit data from points of interest into an iPhone. A customprogrammed piece of software then dealt with this data, translating it into
auditory and haptic feedback. The cane was chosen to work with RFID cards
placed under flooring in arcs in front of points of interest.
The investigation that followed was created to test the ability of the ASOVI
system to aid in navigation and orientation within indoor environments and its
reception with the target audience. Participants were asked to locate
landmarks or points of interest within a test environment using the ASOVI
system. It was concluded that the ASOVI system was successful in aiding the
visually impaired participants to wayfind and navigate in the test environment.
However, it was highlighted that the ASOVI system had problems with reading
some of the RFID cards placed in the environment due to the speed at which
the participants traversed the test area. It was established that a greater read
range is needed to compensate for this factor and as such, a different RFID
receiver may be needed. In addition, it was found that the configuration of the
card arc system caused reading problems and that as a result, straight lines of
cards protruding from objects or points of interest would be more efficient. A
focus group following the task produced significantly positive feedback,
concluding that the system has commercial value.
The final investigation was designed to ascertain whether the ASOVI system
has a positive effect on the ability of a visually impaired user to create
cognitive maps of their environment. Having gained experience of the test
environment with the ASOVI system in the previous investigation, participants
were then asked to describe routes from one of point of interest to another.
Overall, this investigation showed that the ASOVI system was successful in
aiding visually impaired users to create cognitive maps of the indoor
environment. Having evaluated anomalies in the investigation data, it was
concluded that with further development, real life installation of the ASOVI
system would mean that users may be exposed to indoor environments on
multiple occasions and for potentially extended periods of time resulting in
more uniform positive results.
8.2 – Future Research
Investigations undertaken as part of this thesis highlighted several areas of
further development to be implemented in order to make the ASOVI system
suitable for the UK market.
Although the ASOVI system tested successfully in its ability to aid indoor
navigation and wayfinding, the investigation discussed in section 6.3 indicated
that the read range of the RFID receiver was not suitable and an alternative
with a greater read range should be obtained. This would incur further
research into the RFID market or possible inquiry into the bespoke
manufacture of a model specifically engineered to the needs of the ASOVI
system. Furthermore, observations noted during this investigation marked the
positioning of the RFID receiver on the tip of the cane as preventing its use as
a method of obstacle avoidance. As a result, it may be possible to modify the
design of the cane and reposition the RFID receiver within its tip and
so avoiding the possibility of the receiver snagging on flooring materials.
Regarding RFID receiver positioning, it was also suggested during the focus
group session that the receivers could be developed to fit onto shoes or guide
dog collars, for example. This would widen its target market to include visually
impaired persons who do not rely on a cane as their primary method of
obstacle avoidance.
Testing revealed that infrastructure design using arcs of RFID cards placed
around a landmark or point of interest was somewhat flawed. This was
particularly highlighted when observing the different styles of orientation
utilised by some of the visually impaired participants. One possible solution to
this issue may be to use straight lines of RFID cards perpendicular to the
landmarks or points of interest. This would prevent users from inadvertently
missing points of interest and landmarks. Another advantage of using straight
lines of RFID cards would be that they would allow for the implementation of a
checkpoint system which could guide a user on a route from A to B. This
feature was a suggestion made by participants following the investigation. In
conjunction with a checkpoint system enable by using straight lines of RFID
cards, the recently released iPhone 4S’ voice recognition function, ‘Siri’ could
be utilised. This would allow the user to ask the ASOVI system to guide them
from A to B, using spoken commands in tandem with the aforementioned lines
of RFID cards as checkpoints.
Participants reported that the headphones would benefit from louder and more
clearly intelligible audio whilst still allowing the ambient noise from the
environment to infiltrate for safety reasons. This indicates further research
should be undertaken into methods of improving the headphones to this end.
It was also suggested that the ASOVI system could be developed to work on
different mobile smart phones to accommodate those users that already owned
different models. This development would be conducive to one of the prime
objectives of the system: as a solution to the commercial in-viability of other
currently available wayfinding systems on the market.
8.2 – Conclusion
In viewing the project as a whole, the ASOVI system was successful in its aims
of providing indoor orientation and wayfinding for the visually impaired using
an accessible, commercially viable platform. This was achieved by combining
the Apple iPhone 4, which is commercially viable on account of its multifunctionality and is also accessible to the visually impaired with the use of RFID
technologies, which proved conducive with an indoor environment at a
commercially viable cost. The ASOVI system tested positively with a group of
end-user participants, successfully aiding in navigation and wayfinding within
an indoor environment alongside supplementing the creation of cognitive
maps. Investigation processes undertaken throughout the project have also
highlighted several exciting areas of future research necessary to develop the
ASOVI system for release in the commercial market.
Appendix A
static struct termios gOriginalTTYAttrs;
//-------------------------------------------------------------void testApp::setup(){
ofBackground(0, 0, 0);
glEnableClientState( GL_VERTEX_ARRAY ); // this should be in OF
// initialize the accelerometer
// touch events will be sent to myTouchListener
getSerial = openSerialOther();
font.loadFont(ofToDataPath("verdana.ttf"),8, false, true);
nTimesRead = 0;
nBytesRead = 0;
readTime = 0;
memset(bytesReadString, 0, 4);
numBytes = 15;
lastBuffer = new unsigned char[numBytes]; // this buffer gets filled
memset(lastBuffer, 0, numBytes);
totalData ="";
//-------------------------------------------------------------bool testApp::openSerialOther() {
otherSerialFD = -1;
struct termios options;
printf("Initializing Serial Port... ");
otherSerialFD = open("/dev/tty.iap", O_RDWR | O_NOCTTY |
if (otherSerialFD != -1) {
if (fcntl(otherSerialFD, F_SETFL, 0) != -1) {
if (tcgetattr(otherSerialFD, &gOriginalTTYAttrs) != -1) {
options = gOriginalTTYAttrs;
if (tcsetattr(otherSerialFD, TCSANOW, &options) == 1)
printf("Error setting tty attributes %s %s(%d).\n",
"/dev/tty.iap", strerror(errno), errno);
return false;
return true;
} else {
printf("Error getting tty attributes %s - %s(%d).n",
"/dev/tty.iap", strerror(errno), errno);
if (otherSerialFD != -1) {
return false;
} else {
printf("Error clearing O_NONBLOCK %s - %s(%d).n",
"/dev/tty.iap", strerror(errno), errno);
if (otherSerialFD != -1) {
return false;
} else {
printf("Error setting TIOCEXCL on %s - %s(%d).n", "/dev/tty.iap",
strerror(errno), errno);
if (otherSerialFD != -1) {
return false;
return false;
void testApp::otherSerial() {
unsigned char tmpByte[15];
char tmpByte2[10];
memset(tmpByte, 0, 15);
memset(tmpByte2, 0, 10);
if (read(otherSerialFD, tmpByte, 15) == -1) {
printf("Read Error.rn");
printf("RFID Number: \n");
"%01x%01x%01x%01x%01x%01x%01x%01x%01x%01x\n", tmpByte[1],
tmpByte[2], tmpByte[3],
memcpy(lastBuffer,tmpByte2, 10);
string tstr;
tstr = (const char*)tmpByte2;
totalData += tstr + "\n";
NSArray *paths =
NSUserDomainMask, YES);
NSString *documentsDirectoryPath = [paths objectAtIndex:0];
NSString *filePath = [documentsDirectoryPath
NSString* result = [NSString stringWithUTF8String:totalData.c_str()];
NSData* settingsData;
settingsData = [result dataUsingEncoding: NSASCIIStringEncoding];
if ([settingsData writeToFile:filePath atomically:YES])
//---if statements for checking cards
if(whoBeIt != 1)
if(strncmp(tmpByte2,"34433030323045413045",20) == 0 ||
strncmp(tmpByte2,"34433030323045414639",20) == 0 ||
strncmp(tmpByte2,"34433030323041414537",20) == 0 ||
strncmp(tmpByte2,"34433030323130344134",20) == 0 ||
strncmp(tmpByte2,"34433030323043364135",20) == 0 ||
strncmp(tmpByte2,"34433030323044383136",20) == 0 ||
strncmp(tmpByte2,"34433030323044363442",20) == 0 ||
strncmp(tmpByte2,"34433030323041464442",20) == 0 ||
strncmp(tmpByte2,"34433030323043433834",20) == 0 ||
strncmp(tmpByte2,"34433030323130333743",20) == 0 ||
strncmp(tmpByte2,"34433030323130333841",20) == 0 ||
strncmp(tmpByte2,"34433030323046323130",20) == 0 ||
strncmp(tmpByte2,"34433030323045354431",20) == 0 ||
strncmp(tmpByte2,"34433030323042464335",20) == 0 ||//109
strncmp(tmpByte2,"34433030323045423744",20) == 0) //110
whoBeIt = 1;
if(whoBeIt != 2)
if(strncmp(tmpByte2,"34433030323045434345",20) == 0 ||
strncmp(tmpByte2,"34433030323130304533",20) == 0 ||
strncmp(tmpByte2,"34433030323042454642",20) == 0 ||
strncmp(tmpByte2,"34433030323044364634",20) == 0 ||
strncmp(tmpByte2,"34433030323043343139",20) == 0 ||
strncmp(tmpByte2,"34433030323045414634",20) == 0 ||
strncmp(tmpByte2,"34433030323046324331",20) == 0 ||
strncmp(tmpByte2,"34433030323044374630",20) == 0 ||
strncmp(tmpByte2,"34433030323046394438",20) == 0 ||
strncmp(tmpByte2,"34433030323046423541",20) == 0 ||
strncmp(tmpByte2,"34433030323043363041",20) == 0 ||
strncmp(tmpByte2,"34433030323045383138",20) == 0 ||
strncmp(tmpByte2,"34433030323042424538",20) == 0 ||
strncmp(tmpByte2,"34433030323130354639",20) == 0 || //111
strncmp(tmpByte2,"34433030323042433435",20) == 0) //112
whoBeIt = 2;
if(whoBeIt != 3)
if(strncmp(tmpByte2,"34433030323046353332",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
strncmp(tmpByte2,"34433030323046383638",20) == 0 || //113
strncmp(tmpByte2,"34433030323044454539",20) == 0)//114
whoBeIt = 3;
== 0
== 0
== 0
== 0
if(whoBeIt != 4)
if(strncmp(tmpByte2,"34433030323045393146",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 || //115
== 0) //116
whoBeIt = 4;
== 0
== 0
== 0
== 0
== 0
if(whoBeIt != 5)
if(strncmp(tmpByte2,"34433030323130453231",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 || //117
== 0)//118
whoBeIt = 5;
== 0
== 0
== 0
== 0
== 0
== 0
if(whoBeIt != 6)
if(strncmp(tmpByte2,"34433030323044334430",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 || //119
== 0)//120
whoBeIt = 6;
== 0
== 0
== 0
== 0
== 0
== 0
if(whoBeIt != 7)
if(strncmp(tmpByte2,"34433030323130313037",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 )
whoBeIt = 7;
== 0
== 0
== 0
== 0
== 0
== 0
if(whoBeIt != 8)
if(strncmp(tmpByte2,"34433030323045364330",20) == 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 ||
== 0 )
whoBeIt = 8;
if(whoBeIt != 9)
if(strncmp(tmpByte2,"34433030323130423537",20) == 0 ||
strncmp(tmpByte2,"34433030323042333832",20) == 0 ||
strncmp(tmpByte2,"34433030323045343236",20) == 0 ||
strncmp(tmpByte2,"34433030323043324432",20) == 0 ||
strncmp(tmpByte2,"34433030323044433644",20) == 0 ||
strncmp(tmpByte2,"34433030323044384345",20) == 0 ||
strncmp(tmpByte2,"34433030323130433239",20) == 0 ||
strncmp(tmpByte2,"34433030323130314342",20) == 0 ||
strncmp(tmpByte2,"34433030323130384642",20) == 0 ||
strncmp(tmpByte2,"34433030323043363438",20) == 0 ||
strncmp(tmpByte2,"34433030323046343037",20) == 0 ||
strncmp(tmpByte2,"34433030323044323241",20) == 0 ||
strncmp(tmpByte2,"34433030323130434331",20) == 0 ||
strncmp(tmpByte2,"34433030323044323235",20) == 0 ||
strncmp(tmpByte2,"34443030374135454435",20) == 0 )
whoBeIt = 9;
if(whoBeIt != 10)
if(strncmp(tmpByte2,"34433030323130423037",20) == 0 ||
strncmp(tmpByte2,"34433030323130314132",20) == 0 ||
strncmp(tmpByte2,"34433030323044304134",20) == 0 ||
strncmp(tmpByte2,"34433030323043463236",20) == 0 ||
strncmp(tmpByte2,"34433030323045303339",20) == 0
strncmp(tmpByte2,"34433030323044323833",20) == 0 ||
strncmp(tmpByte2,"34433030323044324439",20) == 0 ||
strncmp(tmpByte2,"34433030323044423836",20) == 0 ||
strncmp(tmpByte2,"34433030323043424343",20) == 0 ||
strncmp(tmpByte2,"34433030323045383436",20) == 0 ||
strncmp(tmpByte2,"34433030323041423539",20) == 0 ||
strncmp(tmpByte2,"34433030323043373446",20) == 0 ||
strncmp(tmpByte2,"34433030323042363936",20) == 0 ||
strncmp(tmpByte2,"34433030323043303531",20) == 0 ||
strncmp(tmpByte2,"34433030323044353543",20) == 0)
whoBeIt = 10;
if(whoBeIt != 11)
if(strncmp(tmpByte2,"34433030323041423438",20) == 0 || //151
strncmp(tmpByte2,"34433030323043313736",20) == 0 ||
strncmp(tmpByte2,"34433030323046304641",20) == 0 ||
strncmp(tmpByte2,"34433030323043394336",20) == 0 ||
strncmp(tmpByte2,"34433030323043463433",20) == 0 ||
strncmp(tmpByte2,"34433030323041423343",20) == 0 ||
strncmp(tmpByte2,"34433030323046304239",20) == 0 ||
) //165
whoBeIt = 11;
if(Loader != 1)
{if(strncmp(tmpByte2,"34433030323043424641",20) ==
strncmp(tmpByte2,"34433030323043333243",20) == 0
strncmp(tmpByte2,"34433030323046423637",20) == 0
strncmp(tmpByte2,"34433030323130423939",20) == 0
strncmp(tmpByte2,"34433030323043413532",20) == 0
strncmp(tmpByte2,"34353030423842323845",20) == 0
strncmp(tmpByte2,"34433030323044313241",20) == 0
strncmp(tmpByte2,"34433030323044424631",20) == 0
strncmp(tmpByte2,"34433030323043344532",20) == 0
strncmp(tmpByte2,"34433030323041394134",20) == 0
strncmp(tmpByte2,"34353030424534384531",20) == 0
strncmp(tmpByte2,"34433030323045413443",20) == 0
strncmp(tmpByte2,"34433030323041433046",20) == 0
strncmp(tmpByte2,"34433030323130303742",20) == 0
Loader = 1;
0 ||
if(whoBeIt != 12 && Loader == 1)
if(strncmp(tmpByte2,"34433030323045353533",20) == 0
strncmp(tmpByte2,"34433030323042354238",20) == 0 ||
strncmp(tmpByte2,"34433030323042343130",20) == 0 ||
strncmp(tmpByte2,"34433030323046313136",20) == 0 )
whoBeIt = 12;
Loader = 0;
if(Loader != 2)
if(strncmp(tmpByte2,"34433030323045323030",20) == 0 ||
strncmp(tmpByte2,"34433030323044434439",20) == 0 ||
strncmp(tmpByte2,"34433030323130364431",20) == 0 ||
strncmp(tmpByte2,"34433030323046434142",20) == 0 ||
strncmp(tmpByte2,"34433030323043303045",20) == 0 ||
strncmp(tmpByte2,"34433030323045454435",20) == 0 ||
strncmp(tmpByte2,"34433030323043414138",20) == 0 ||//(200)
strncmp(tmpByte2,"34433030323131314443",20) == 0 ||
strncmp(tmpByte2,"34433030323046313842",20) == 0 ||
strncmp(tmpByte2,"34433030323046454530",20) == 0 ||
strncmp(tmpByte2,"34433030323043374434",20) == 0 ||
strncmp(tmpByte2,"34433030323130323945",20) == 0 ||
strncmp(tmpByte2,"34433030323045463033",20) == 0 ||
strncmp(tmpByte2,"34433030323046383245",20) == 0 )
Loader = 2;
if(whoBeIt != 13 && Loader == 2)
if(strncmp(tmpByte2,"34433030323043323241",20) == 0 ||
strncmp(tmpByte2,"34433030323041374136",20) == 0 ||
strncmp(tmpByte2,"34433030323130424230",20) == 0
strncmp(tmpByte2,"34433030323043424338",20) == 0 ||
strncmp(tmpByte2,"34433030323046463638",20) == 0 ||
strncmp(tmpByte2,"34433030323044374342",20) == 0 ||
strncmp(tmpByte2,"34433030323043324235",20) == 0 ||
strncmp(tmpByte2,"34433030323046363132",20) == 0 ||
strncmp(tmpByte2,"34433030323045414638",20) == 0 ||
strncmp(tmpByte2,"34433030323046463646",20) == 0 ||
strncmp(tmpByte2,"34433030323130393630",20) == 0 ||
strncmp(tmpByte2,"34433030323130343131",20) == 0 ||
strncmp(tmpByte2,"34433030323041463239",20) == 0
strncmp(tmpByte2,"34433030323130443734",20) == 0 )
whoBeIt = 13;
Loader = 0;
if(whoBeIt != 14)
if(strncmp(tmpByte2,"34433030323130353341",20) == 0 ||//222
strncmp(tmpByte2,"34433030323046304335",20) == 0 ||
strncmp(tmpByte2,"34433030323045373842",20) == 0 ||
strncmp(tmpByte2,"34433030323041434133",20) == 0 ||
strncmp(tmpByte2,"34433030323043393036",20) == 0 ||
strncmp(tmpByte2,"34433030323042363332",20) == 0 ||
strncmp(tmpByte2,"34433030323130434539",20) == 0 ||
strncmp(tmpByte2,"34433030323044443234",20) == 0 ||
strncmp(tmpByte2,"34433030323044353235",20) == 0
strncmp(tmpByte2,"34433030323043314530",20) == 0 ||
strncmp(tmpByte2,"34433030323130363833",20) == 0 ||
strncmp(tmpByte2,"34433030323046433444",20) == 0 ||
strncmp(tmpByte2,"34433030323045344332",20) == 0 ||
strncmp(tmpByte2,"34433030323043464546",20) == 0 ||
strncmp(tmpByte2,"34433030323045374346",20) == 0)//236
whoBeIt = 14;
if(whoBeIt != 15)
if(strncmp(tmpByte2,"34433030323043413730",20) == 0 ||//237
strncmp(tmpByte2,"34433030323042433032",20) == 0 ||
strncmp(tmpByte2,"34433030323045394342",20) == 0 ||
strncmp(tmpByte2,"34433030323046454235",20) == 0
strncmp(tmpByte2,"34433030323041453846",20) == 0 ||
strncmp(tmpByte2,"34433030323044434430",20) == 0 ||
strncmp(tmpByte2,"34433030323045413537",20) == 0 ||
strncmp(tmpByte2,"34433030323045414434",20) == 0 ||
strncmp(tmpByte2,"34433030323044364641",20) == 0 ||
strncmp(tmpByte2,"34433030323043464343",20) == 0 ||
strncmp(tmpByte2,"34433030323042343438",20) == 0 ||
strncmp(tmpByte2,"34433030323131303541",20) == 0 ||
strncmp(tmpByte2,"34433030323042344233",20) == 0 ||
strncmp(tmpByte2,"34433030323041383835",20) == 0
strncmp(tmpByte2,"34433030323042434638",20) == 0)//251
whoBeIt = 15;
if(whoBeIt != 16)
if( strncmp(tmpByte2,"34433030323043363146",20) == 0 ||//252
strncmp(tmpByte2,"34433030323041374343",20) == 0 ||
strncmp(tmpByte2,"34433030323045463334",20) == 0 ||
strncmp(tmpByte2,"34433030323042373143",20) == 0 ||
strncmp(tmpByte2,"34353030424535373242",20) == 0 ||//256
strncmp(tmpByte2,"34433030323046383543",20) == 0 ||
strncmp(tmpByte2,"34433030323130383541",20) == 0 ||
strncmp(tmpByte2,"34433030323045463642",20) == 0 ||
strncmp(tmpByte2,"34433030323130343942",20) == 0
strncmp(tmpByte2,"34433030323041363734",20) == 0 ||
strncmp(tmpByte2,"34433030323046413031",20) == 0 ||
strncmp(tmpByte2,"34433030323130314436",20) == 0)//273
whoBeIt = 16;
if(whoBeIt != 17)
if(strncmp(tmpByte2,"34433030323044444335",20) == 0 ||//274
strncmp(tmpByte2,"34433030323041374235",20) == 0 ||
strncmp(tmpByte2,"34433030323046353541",20) == 0 ||
strncmp(tmpByte2,"34433030323130463242",20) == 0 ||
strncmp(tmpByte2,"34433030323042374241",20) == 0 ||
strncmp(tmpByte2,"34433030323130413135",20) == 0 ||
strncmp(tmpByte2,"34433030323046443042",20) == 0
strncmp(tmpByte2,"34433030323045463543",20) == 0 ||
strncmp(tmpByte2,"34433030323041454432",20) == 0 ||
0 ||
0 ||
0 ||//285
0 ||
0 ||
whoBeIt = 17;
if(whoBeIt != 18)
if(strncmp(tmpByte2,"34433030323042343546",20) == 0 ||//289
strncmp(tmpByte2,"34433030323042373237",20) == 0
strncmp(tmpByte2,"34433030323130453338",20) == 0 ||
strncmp(tmpByte2,"34433030323045314330",20) == 0 ||
strncmp(tmpByte2,"34433030323046443243",20) == 0)//303
whoBeIt = 18;
if(whoBeIt != 19)
if(strncmp(tmpByte2,"34433030323044423834",20) == 0 ||//304
strncmp(tmpByte2,"34433030323041393134",20) == 0 ||
strncmp(tmpByte2,"34433030323045443444",20) == 0 ||
strncmp(tmpByte2,"34433030323046454444",20) == 0 ||
strncmp(tmpByte2,"34433030323045413830",20) == 0 ||
strncmp(tmpByte2,"34433030323045463941",20) == 0 ||
strncmp(tmpByte2,"34433030323130364231",20) == 0 ||
strncmp(tmpByte2,"34433030323045363445",20) == 0 ||
strncmp(tmpByte2,"34433030323042364445",20) == 0 ||
strncmp(tmpByte2,"34433030323043304436",20) == 0 ||
strncmp(tmpByte2,"34433030323045434346",20) == 0 ||
strncmp(tmpByte2,"34433030323045394536",20) == 0 ||//315
strncmp(tmpByte2,"34433030323044343746",20) == 0 ||
strncmp(tmpByte2,"34433030323046373241",20) == 0 ||
strncmp(tmpByte2,"34433030323044423641",20) == 0 ||
strncmp(tmpByte2,"34433030323046333735",20) == 0)//319
whoBeIt = 19;
if(whoBeIt != 20)
if(strncmp(tmpByte2,"34433030323046353639",20) == 0
strncmp(tmpByte2,"34433030323046453934",20) == 0 ||
strncmp(tmpByte2,"34433030323045394646",20) == 0 ||
strncmp(tmpByte2,"34433030323043304130",20) == 0 ||
strncmp(tmpByte2,"34433030323041384230",20) == 0 ||
strncmp(tmpByte2,"34433030323042364444",20) == 0 ||//325
strncmp(tmpByte2,"34433030323044434344",20) == 0 ||
strncmp(tmpByte2,"34433030323043433330",20) == 0 ||
strncmp(tmpByte2,"34433030323042394139",20) == 0 ||
strncmp(tmpByte2,"34433030323130353631",20) == 0 ||
strncmp(tmpByte2,"34433030323044374230",20) == 0
0 ||
0 ||
0 ||
0 ||
whoBeIt = 20;
if(whoBeIt != 21)
if(strncmp(tmpByte2,"34433030323041453637",20) == 0 ||//336
strncmp(tmpByte2,"34433030323042323334",20) == 0 ||
strncmp(tmpByte2,"34433030323046363235",20) == 0 ||
strncmp(tmpByte2,"34433030323041433445",20) == 0 ||
strncmp(tmpByte2,"34433030323130393142",20) == 0
strncmp(tmpByte2,"34433030323046373436",20) == 0 ||
strncmp(tmpByte2,"34433030323046393944",20) == 0 ||
strncmp(tmpByte2,"34433030323043464536",20) == 0 ||
strncmp(tmpByte2,"34353030424532464132",20) == 0 ||
strncmp(tmpByte2,"34353030423845463435",20) == 0 ||//345
strncmp(tmpByte2,"34353030424532364433",20) == 0 ||
strncmp(tmpByte2,"34353030424534373141",20) == 0 ||
strncmp(tmpByte2,"34433030323045464333",20) == 0 ||
strncmp(tmpByte2,"34433030323041414430",20) == 0 ||
strncmp(tmpByte2,"34433030323046334642",20) == 0
strncmp(tmpByte2,"34433030323045394131",20) == 0 ||
strncmp(tmpByte2,"34433030323046364536",20) == 0 ||
strncmp(tmpByte2,"34433030323043354336",20) == 0 ||
strncmp(tmpByte2,"34433030323131303236",20) == 0 ||
strncmp(tmpByte2,"34433030323046303432",20) == 0 ||//355
strncmp(tmpByte2,"34433030323044464234",20) == 0 ||
strncmp(tmpByte2,"34433030323043453441",20) == 0 ||
strncmp(tmpByte2,"34433030323042384444",20) == 0 ||
strncmp(tmpByte2,"34433030323131314534",20) == 0 ||
strncmp(tmpByte2,"34433030323043454536",20) == 0
0 ||
0 ||
0 ||
0 ||//364
whoBeIt = 21;
if(whoBeIt != 22)
if(strncmp(tmpByte2,"34433030323043393933",20) == 0 ||//366
strncmp(tmpByte2,"34433030323041353339",20) == 0 ||
strncmp(tmpByte2,"34433030323130413544",20) == 0 ||
strncmp(tmpByte2,"34433030323044303133",20) == 0 ||
strncmp(tmpByte2,"34433030323045353830",20) == 0
strncmp(tmpByte2,"34433030323046414436",20) == 0 ||
strncmp(tmpByte2,"34433030323042344142",20) == 0 ||
strncmp(tmpByte2,"34433030323046454443",20) == 0 ||
strncmp(tmpByte2,"34433030323042313633",20) == 0 ||
strncmp(tmpByte2,"34433030323046343935",20) == 0 ||//375
strncmp(tmpByte2,"34433030323130433338",20) == 0 ||
strncmp(tmpByte2,"34433030323130374344",20) == 0 ||
strncmp(tmpByte2,"34433030323043354241",20) == 0 ||
strncmp(tmpByte2,"34433030323046354333",20) == 0 ||
strncmp(tmpByte2,"34433030323044353136",20) == 0
strncmp(tmpByte2,"34433030323130313545",20) == 0)//381
whoBeIt = 22;
if(whoBeIt != 23)
if( strncmp(tmpByte2,"34433030323043413438",20) == 0 ||//382
strncmp(tmpByte2,"34433030323042303236",20) == 0 ||
strncmp(tmpByte2,"34433030323042313730",20) == 0 ||
strncmp(tmpByte2,"34433030323130313736",20) == 0 ||//385
strncmp(tmpByte2,"34433030323044344546",20) == 0 ||
strncmp(tmpByte2,"34433030323046443033",20) == 0 ||
strncmp(tmpByte2,"34433030323046384332",20) == 0 ||
strncmp(tmpByte2,"34433030323041354344",20) == 0 ||
strncmp(tmpByte2,"34433030323131313246",20) == 0
strncmp(tmpByte2,"34433030323043374632",20) == 0 ||
strncmp(tmpByte2,"34433030323043443343",20) == 0 ||
strncmp(tmpByte2,"34433030323130463645",20) == 0 ||
strncmp(tmpByte2,"34433030323045333830",20) == 0 ||
strncmp(tmpByte2,"34433030323041463246",20) == 0 ||//395
strncmp(tmpByte2,"34433030323041393137",20) == 0 ||
strncmp(tmpByte2,"34433030323045373239",20) == 0)//397v
whoBeIt = 23;
if(whoBeIt != 24)
if( strncmp(tmpByte2,"34433030323042394539",20) == 0 ||//398
strncmp(tmpByte2,"34433030323130344436",20) == 0 ||
strncmp(tmpByte2,"34433030323043344138",20) == 0 ||
0 ||
0 ||
0 ||
0 ||
0 ||//415
0 ||
0 ||
whoBeIt = 24;
if(whoBeIt != 25)
if( strncmp(tmpByte2,"34433030323042423039",20) == 0 ||//419
strncmp(tmpByte2,"34433030323044413743",20) == 0
strncmp(tmpByte2,"34433030323043353330",20) == 0 ||
strncmp(tmpByte2,"34433030323045393441",20) == 0 ||
strncmp(tmpByte2,"34433030323045383031",20) == 0 ||
strncmp(tmpByte2,"34433030323046363838",20) == 0 ||
strncmp(tmpByte2,"34433030323044333842",20) == 0 ||//425
strncmp(tmpByte2,"34433030323041443945",20) == 0 ||
strncmp(tmpByte2,"34433030323043414332",20) == 0 ||
strncmp(tmpByte2,"34433030323044443046",20) == 0 ||
strncmp(tmpByte2,"34433030323043463045",20) == 0 ||
strncmp(tmpByte2,"34353030423845333337",20) == 0
strncmp(tmpByte2,"34433030323041373539",20) == 0 ||
strncmp(tmpByte2,"34433030323044433133",20) == 0 ||
strncmp(tmpByte2,"34433030323046333139",20) == 0)//433
whoBeIt = 25;
if(whoBeIt != 26)
if( strncmp(tmpByte2,"34433030323130423232",20) == 0 ||//434
strncmp(tmpByte2,"34433030323043394430",20) == 0 ||//435
strncmp(tmpByte2,"34433030323044334433",20) == 0 ||
strncmp(tmpByte2,"34433030323043394331",20) == 0 ||
strncmp(tmpByte2,"34433030323046424131",20) == 0 ||
strncmp(tmpByte2,"34433030323044313935",20) == 0 ||
strncmp(tmpByte2,"34433030323042393931",20) == 0
strncmp(tmpByte2,"34433030323045373136",20) == 0 ||
strncmp(tmpByte2,"34433030323045443230",20) == 0 ||
strncmp(tmpByte2,"34433030323046354636",20) == 0 ||
strncmp(tmpByte2,"34433030323042374243",20) == 0 ||
strncmp(tmpByte2,"34433030323045344232",20) == 0 ||//445
strncmp(tmpByte2,"34433030323043463535",20) == 0 ||
strncmp(tmpByte2,"34433030323044453532",20) == 0 ||
strncmp(tmpByte2,"34433030323130333143",20) == 0 ||
strncmp(tmpByte2,"34433030323042324335",20) == 0 ||
strncmp(tmpByte2,"34423030373646453946",20) == 0)//(450)
whoBeIt = 26;
if(whoBeIt != 27)
if(strncmp(tmpByte2,"34433030323045323239",20) == 0 ||//451
strncmp(tmpByte2,"34433030323041373943",20) == 0 ||
strncmp(tmpByte2,"34433030323044383236",20) == 0 ||
strncmp(tmpByte2,"34433030323130463333",20) == 0 ||
strncmp(tmpByte2,"34433030323042384641",20) == 0 ||//455
strncmp(tmpByte2,"34433030323130343631",20) == 0 ||
strncmp(tmpByte2,"34433030323130393233",20) == 0 ||
strncmp(tmpByte2,"34433030323044443645",20) == 0 ||
strncmp(tmpByte2,"34433030323131304246",20) == 0 ||
strncmp(tmpByte2,"34433030323043313446",20) == 0
strncmp(tmpByte2,"34433030323130344643",20) == 0 ||
strncmp(tmpByte2,"34433030323042454343",20) == 0 ||
strncmp(tmpByte2,"34433030323044443733",20) == 0 ||
strncmp(tmpByte2,"34433030323045313731",20) == 0 ||
strncmp(tmpByte2,"34433030323046443339",20) == 0)//465
whoBeIt = 27;
if(whoBeIt != 28)
if( strncmp(tmpByte2,"34433030323042353633",20) == 0 ||//466
0 ||
0 ||
0 ||
0 ||
0 ||
0 ||
0 ||
0 ||//475
0 ||
0 ||
0 ||
0 ||
whoBeIt = 28;
if(whoBeIt != 29)
if(strncmp(tmpByte2,"34433030323045354633",20) == 0 || //481
strncmp(tmpByte2,"34433030323131303044",20) == 0 ||
strncmp(tmpByte2,"34433030323043383438",20) == 0 ||
strncmp(tmpByte2,"34433030323045414238",20) == 0 ||
strncmp(tmpByte2,"34433030323044453044",20) == 0 ||//485
strncmp(tmpByte2,"34433030323044303934",20) == 0 ||
strncmp(tmpByte2,"34433030323042414544",20) == 0 ||
strncmp(tmpByte2,"34433030323045424544",20) == 0 ||
strncmp(tmpByte2,"34433030323046353634",20) == 0 ||
strncmp(tmpByte2,"34433030323044343545",20) == 0
strncmp(tmpByte2,"34433030323042433937",20) == 0 ||
strncmp(tmpByte2,"34433030323045363436",20) == 0 ||
strncmp(tmpByte2,"34433030323045423130",20) == 0 ||
0 ||
0 ||//495
0 ||
whoBeIt = 29;
//-------------------------------------------------------------void testApp::update(){
if (getSerial){
//-------------------------------------------------------------void testApp::draw(){
// Serial
font.drawString("Cerberus 0.1" , 120, 20);
font.drawString("Scanned Tag:\n" , 20, 70);
font.drawString(totalData, 80, 90);
void testApp::exit() {
//-------------------------------------------------------------void testApp::mouseMoved(int x, int y ){
// this will never get called
//-------------------------------------------------------------void testApp::mouseDragged(int x, int y, int button){
//-------------------------------------------------------------void testApp::mousePressed(int x, int y, int button){
//-------------------------------------------------------------void testApp::mouseReleased(){
printf("frameRate: %.3f, frameNum: %i\n", ofGetFrameRate(),
//-------------------------------------------------------------void testApp::mouseReleased(int x, int y, int button){
//-------------------------------------------------------------void testApp::touchDown(float x, float y, int touchId,
ofxMultiTouchCustomData *data){
printf("touch %i down at (%i,%i)\n", touchId, x,y);
//-------------------------------------------------------------void testApp::touchMoved(float x, float y, int touchId,
ofxMultiTouchCustomData *data){
printf("touch %i moved at (%i,%i)\n", touchId, x,y);
//-------------------------------------------------------------void testApp::touchUp(float x, float y, int touchId, ofxMultiTouchCustomData
printf("touch %i up at (%i,%i)\n", touchId, x,y);
//-------------------------------------------------------------void testApp::touchDoubleTap(float x, float y, int touchId,
ofxMultiTouchCustomData *data){
printf("touch %i double tap at (%i,%i)\n", touchId, x,y);
Appendix B
Tables of investigation data pertaining to the “RFID Device Read Range Experiment”й
Appendix C
Participant 1: Some card arcs are not being scanned due to the speed at which the
user is scanning with the cane.
Some information is lasting too long and causing the participant to stop
and wait while the information has finished being conveyed.
Participant seems confident while using the system and traversing at a
good speed.
Participant finds the target location without difficulty.
Participant missing some card arcs due to using opposite wall for walking
in a straight line.
The above-mentioned walking technique causes participant to miss the
target location on first traversal causing them to double back.
Participant gaining in confidence with the ASOVI system.
Participant struggling with RFID receiver attached to the ball on the end
of the cane due to the fact the ball needs to rotate and is causing
problems with the wires and movement.
Participant hitting multiple arcs causing them to slow down.
Participant has slight eyesight (enough to make out showed objects) and
due to the cards being placed on top of the flooring as the modification of
the environment is not possible this is causing distraction.
Participant 2: Participant has trouble with the RFID receiver being attached to the
bottom of the cane. Receiver keeps snagging on the floor and when the
end of the cane rolls as designed the wires wrap around the cane and on
one occasion this causes a short circuit and the cane to cease
Participant cautious at first when using the ASOVI system however
gaining confidence as time passes.
As the cards are placed on top of the environment the cane snags some
of the cards and moves them from their intended arc position.
Participant find target location however due to the afore mentioned
problems this takes some time.
Participant missing some card arcs due to following the wall as a mode of
orientation during traversal of the environment.
Some card arcs not being scanned efficiently due to the speed and
method of scanning with the cane.
Some information is lasting too long and causing the participant to stop
and wait while the information has finished being conveyed.
Participant 3: Participant has trouble with reading the cards due to speed of movement
of the cane.
Participant has trouble scanning some cards due to speed of the scan.
Participant misses some card arcs due to following the wall for
orientation. (Participant verbally mentions that they are a qualified cane
teacher and that using a wall is commonplace)
Participant successfully finds target location without any difficulty.
The receiver snags on the floor repeatedly (participant verbally
communicates that it is snagging and thus hinders the scan).
Participant reacting positively to feedback and taking note of warnings
(end of test area / stairs approaching).
As the cards are placed on top of the environment the cane snags some
of the cards and moves them from their intended arc position.
Participant 4: Travelling at very high speed with a huge amount of confidence.
Due to speed of travel a large amount of card arcs being missed.
The receiver snags on the floor repeatedly.
As the cards are placed on top of the environment the cane snags some
of the cards and moves them from their intended arc position.
The participant is having trouble locating the specified area due to lack of
information, as traversal speed is that of a slow jog not a walk.
Appendix D
CW4/01 to CW4/04.
CW4/01 to Office Of Damien De Luca.
Office Of Damien De Luca to Modern Artwork – Dots.
CW4/07 to Male Toilets.
Lifts to Security Phone.
Water Fountain to Seating Area.
First Aid Box to Corridor Left Turn.
Disabled Toilet to CW4/01.
Apple. (2011). Audiobooks by Cross Forward Consulting. Retrieved 2011
8-July from itunes.Apple:
Apple Inc. (2011). Apple iPhone4 Specs. Retrieved 2011
л 5-April from
Apple Inc:
Apple. (2011). oMoby By IQ Engines. Retrieved 2011
л 8-July from
Arslan, H., Chen, Z. N., & Benedetto, M.-G. D. (2006). Ultra Wideband
Wireless Communication. John Wiley & Sons.
Brabyn, J. A. (1982). New Developments in Mobility and Orientation Aid for the
Blind. IEEE Transaction on biomedical engineering , 285-289.
Chipworksй (2011
л 23-June). Recent Teardowns. Retrieved 2010 from
Chivers, M. (2006). Differential GPS Explained . Trimble Navigation.
CNC Router Source. (2011). Breakout Boards. Retrieved 2011
л 16-July
from CNC Router Source:
Diggelen, F. V. (2009). A-GPS Asissted GPS, GNSS, and SBAS. Boston /
London: Artech House.
Dodds, A. G., Howarth, C. I., & Carter, D. C. (1982). The mental maps of the
blind: the role of previous experience. Journal of Visual Impairment and
Blindnes , 5-12.
Espinosa, M. A., & Ochaita, E. (1998). Using tactile maps to improve the
practival spatial knowledge of adults who are blind. Journal of visual
impairment and blindness , 338-345.
Esquirol, Mй (2011
л 11-March). iPhone for the viually impaired. Retrieved
л 8-July from SoftCity:
Etienne, A. S., & Jeffery, K. J. (2004). Path Integration in Mammals.
Hippocampus , 180-192.
Fortin, M., Voss, P., Lord, C., Lassonde, M., Pruessner, J., Saint-Amour, D., et
al. (2008). Wayfinding in the blind: larger hippocampal volume and
supranormal spatial navigation. Brain , 2995-3005.
Fosi, Bй (2008
л 13-November). Invisual - A tactile siicone case for the apple
iphoneй Retrieved 2011
л 8-July from Bruno Fosi:
Gill, J. (1996). An Orientation and Navigation System for Blind Pedestrians.
Retrieved 2011
л 18-April from Magdeburg University:
Golledge, R. G. Schematic of the PGS System. Schematic of the PGS System.
University Of California, Santa Barbara, Santa Barbara, California,.
Golledge, R. G., Klatzky, R. L., & Loomis, J. M. (1996). Cognitive Mapping and
Wayfinding by Adults Without Vision. The Construction of Cognitive Maps ,
Google. (2011). Google Mapsй Retrieved 2011
л 10 from Google:
Greenemeier, L. (2008). A Positioning System That Goes Where GPS Can't.
Scientific American.
Helal, A., Moore, S. E., & Ramachandran, B. (2001). Drishit: An Integrated
Navigation System for Visually Impaired and Disabled. Florida: University Of
Hightower, J., Borriello, G., & Want, R. (2000). SpotON: An Indoor 3D Location
Sensing Technology Based on RF Signal Strength. Washington: University Of
Washington .
Hoovers. (2011). Hoovers. Retrieved 2011
л 14-March from Hoovers:
HumanWare. (2011). Oratio for BlackBerry Smartphonesй Retrieved 2011
8-July from HumanWare:
HumanWare. (2010). Trekker Breeze handheld talking GPS. Retrieved 2011
л 6-April from HumanWare:
Humanware. (2011). Trekker Breeze Handheld Talking GPS. Retrieved 2011
л 23-April from Humanware:
Hunt, D. V., Puglia, A., & Puglia, M. (2007). RFID : a guide to radio frequency
identification. New Jersey: Wiley-Interscience .
Ingram, S. (2006). UltraWideBand Indoor Positioning (White Paper). Reading:
Thales Group.
Jacobson, W. H. (1993). The art and science of teaching orientation and
mobility to persons with visual impairments. New York: AFB Press.
Jones, G. (2005). Echolocation (Primer). Current Biology , 484-485.
Kirkpatrick, S. (2011). Loadstone GPS - Satellite Navigation for blind mobile
phone usersй Retrieved 2011
л 23-April from Loadstone GPS:
Kotsakis, E., Caignault, A., Woehler, W., & Ketselidis, M. (2001). Integrating
Differential GPS data into an Embedded GIS and its Application to Infomobility
and Navigation. Joint Research Centre CCR , n/a.
Lahav, O., & Mioduser, D. (2008). Haptic-feedback support for cognitive
mapping of unknown spaces by people who are blind. International Journal of
Human-Computer Studies , 23-35.
Liu, H., Darabi, H., Banerjee, P., & Liu, J. (2007). Survey of Wireless Indoor
Positioning Techniques and Systems. IEEE Transactions on systems, man, and
cybernetics-part c : Applications and reviews , 1067-1080.
Lorenz, A., Schmitt, C., Oppermann, R., Eisenhauer, M., & Zimmermann, A.
(2001). Location and Tracking in Mobile Guides. Fraunhofer FIT (Institute for
Applied Information Technology) .
Lynch, K. (1960). The Image of the City. Cambridge : MIT Press.
Millar, S. (1994). Understanding and Representing Space. Oxford: Oxford
University Press.
Morrongiello, B. A., Timney, B., Humphrey, K., Anderson, S., & Skory, C.
(1995). Spatial Knowledge in Blind and Sighted Children. Journal Of
Experimental Child Psychology , 211-233.
Ni, L. M., Liu, Y., Lau, Y. C., & Patil, A. P. (2004). LANDMARC: Indoor Location
Sensing Using Active RFID. Wireless Networks , 701-710.
Noyce, Dй Aй, & Barlow, Jй Mй (2003
л April)й Interfacing Accessible
Pedestrian Signals (APS) with Traffic Signal Control Equipment. Retrieved 2012
Pinouts.RU. (2010). Apple iPod, iPhone (2g, 3g), iPad Dock connector pinout.
Retrieved 2011
л 18-March from Pinouts.RU:
Pitman, M. (1994). Bat sonar points to creation. Creation , 219–220.
Popular Mechanicsй (2011
л 20-April)й Retrieved 2011
л 8-July from Field
Guide: What's the Best Smartphone Operating System?:
Raferty), L. (. Rough diagram of sound frequency scale, showing ultrasound
and some applications. Derivative of Heath Raftery's own creation. Rough
diagram of sound frequency scale, showing ultrasound and some applications.
at University of Newcastle, Mayfield.
Ran, L., Helal, S., & Moore, S. (2003). Drishti: An integrated Indoor/Outdoor
Blind Navigation System and Service. Florida: University Of Florida.
RFID Journal LLC. (2011). General RFID Informationй Retrieved 2011
л 1April from Aim Global:
Rieser, J., Ashmed, C., Taylor, C., & Youngquist, G. A. (1990). Visual
perception and the guidance of locomotion without vision to previously seen
targets. Perception , 675-689.
RNIB Business Development Team / SFX Technologies. (2007). RNIB RF FOB
RNIB Business Development Team. (2008). RNIB React Advice Sheet. London:
RNIB. (2011). Guide Canesй Retrieved 2011
л 19-March from RNIB:
RNIB Research. (2011). Key Information and statisticsй Retrieved 2011
л 6April from RNIB:
Robot Electronics. (2011). Robot Electronics Ultrasound Rangers. Retrieved
л 22-March from Robot Electronics:
Rodriguez, M., Pece, J. P., & Escudero, C. J. (2010). In-building location using
Bluetooth. Universiadade da Coruna , n/a.
Ross, D. A. (2004). Cyber Crumbs for Successfull Aging with Vision Loss.
Pervasive Compting , 30-35.
Ross, D. A., Lightman, A., & Henderson, V. L. (2005). Cyber Crumbs: An
Indoor Orientation and Wayfinding Infrastructure.
Schroeder, Pй (2010
л 9-June). Accessibility for the visually impaired has
come a long way, thanks to Apple, but not far enough. Retrieved 2011
л 8July from Smart Planet:
Seraphin, Aй (2010
л 12-June). My First Week With The iPhone. Retrieved
л 8-July from BehindTheCurtain:
SparkFun Electronics. (2011). 5V DC to DC Step Up - 2xAA. Retrieved 2011
л 18-March from SparkFun Electronics:
SparkFun Electronics. (2011). Logic Level Converter. Retrieved 2011
л 18March from SparkFun Electronics:
SparkFun Electronics. (2011). RFID Reader ID-12 (125kHz). Retrieved 2011
л 18-March from SparkFun Electronics:
T - Mobileй (2011
л March)й T- Mobile LGA140й Retrieved 2011
л 18March from T- Mobile:
TalkingSigns. Helping Visually Impaird People Find Their Way. Helping Visually
Impaird People Find Their Way. Talking Signs, Inc., Los Angeles .
The Modern Greenй (2008
л 17-October). Light Spectrum. Retrieved 2011
л 25-March from The Modern Green:
The Times, N. (2007). Blind Independance. Times Newspapers Limited .
Three. (2011). BlackBerry Torch 9800. Retrieved 2011
л 8-July from Three:
Three. (2011). HTC Sensationй Retrieved 2011
л 8-July from Three:
Three. (2011). iPhone on Pay as You Go. Retrieved 2011
л 8-July from
Three. (2011). Three Samsung Galaxy S II Blackй Retrieved 2011
л 8-July
from ThreeStore:
uBlox. (2011). NEO-6P GPS Moduleй Retrieved 2011
л 14-08 from Ublox:
U-Bloxй (2011
л March)й Online-Shopй Retrieved 2011
л 29-March from UBlox Shop:
uBlox. (2011). SuperSense. Retrieved 2011 from uBlox:
UK Government. (2011). Disability Living Allowance - rates and how to claim.
Retrieved 2011
л 6-April from Direct Gov:
Ungar, S., Blades, M., & Spencer, C. (1995b). Mental Rotation of a Tactile
Layout by Young Visually Impaired Children. Perception , 891-900.
Unger, S. (2000). Cognitive Mapping Without Visual Experience. Cognitive
Mapping: Past Present and Future. , 1-16.
VanHemert, Kй (2010
л 20-September). How the iPhone helped a blind
person see color. Retrieved 2011
л 8-July from Gizmodo:
Vodafone. (2011). BlackBerry Price Plans. Retrieved 2011
л 8-July from
Vodafone. (2011). iPhone on Pay Monthly. Retrieved 2011
л 8-July from
Vodafone UK:
Vodafone. (2011). Price Plans for HTC Sensationй Retrieved 2011
л 8-July
from Vodafone:
Vodafone. (2011). Price Plans For Samsung Galaxy S IIй Retrieved 2011
л 8July from Vodafone:
XE Corporation. (2011). XE The Worlds Favourite Currency Site. Retrieved
л 29-March from XE:
ZigBee Alliance. (2011). ZigBee Specificationsй Retrieved 2011
л 14-March
from ZigBee:
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF