Majestic Yosemite Hotel virtual tour application and indoor model

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Majestic Yosemite Hotel Virtual Tour Application and Indoor Model

by

Trevor James Denson

A Thesis Presented to the
Faculty of the USC Graduate School
University of Southern California
In Partial Fulfillment of the
Requirements for the Degree
Master of Science
(Geographic Information Science and Technology)

December 2017

Copyright © 2017 by Trevor Denson

To my friends and family.

iv

Table of Contents
List of Figures ............................................................................................................................... vii
List of Tables ................................................................................................................................. ix
Acknowledgements ......................................................................................................................... x
List of Abbreviations ..................................................................................................................... xi
Abstract ........................................................................................................................................ xiii
Chapter 1 Introduction .................................................................................................................... 1
1.1. Majestic Yosemite Hotel Background ................................................................................3
1.2. Project Scope ......................................................................................................................5
1.3. Motivation and General Objective ......................................................................................9
1.4. Thesis Organization ..........................................................................................................12
Chapter 2 Related Work ................................................................................................................ 13
2.1. Types of Positioning Systems and Technologies ..............................................................13
2.1.1. Global Positioning System .......................................................................................14
2.1.2. Wide-area Cellular Networks ..................................................................................17
2.1.3. In-building Positioning Networks ............................................................................17
2.2. Existing Tour Applications ...............................................................................................25
2.2.1. Virtual Tour Web Applications ...............................................................................26
2.2.2. Mobile Tour Applications ........................................................................................30
2.3. Indoor Modeling Techniques ............................................................................................33
Chapter 3 Methods ........................................................................................................................ 36
3.1. Choosing an Appropriate Indoor Positioning System Technology ..................................36
3.1.1. Cost Analysis ...........................................................................................................37
3.1.2. Time for Implementation .........................................................................................37
v

3.1.3. Long-Term Maintenance .........................................................................................38
3.2. Testing the Indoor Positioning System Methodology ......................................................40
3.2.1. Anyplace ..................................................................................................................43
3.2.2. IndoorAtlas and BlueCats Beacon Services ............................................................46
3.3. Developing with Unity ......................................................................................................59
3.3.1. Introduction to Unity Development .........................................................................60
3.3.2. Unity and Vuforia Object Recognition Service .......................................................62
3.3.3. Object Recognition Scripts ......................................................................................63
3.3.4. Testing Unity and Vuforia .......................................................................................64
3.3.5. Unity and Bluetooth .................................................................................................66
3.3.6. Educational Content .................................................................................................67
3.3.7. Designing the User Interface ...................................................................................68
Chapter 4 Model Development ..................................................................................................... 69
4.1. Scanning the Hotel’s South Wing .....................................................................................70
4.2. Scanning the Hotel’s Furniture .........................................................................................72
4.3. Building the Physical Model .............................................................................................73
Chapter 5 Results .......................................................................................................................... 75
5.1. Alternative Application Methodologies and Results ........................................................75
5.1.1. Anyplace ..................................................................................................................75
5.1.2. IndoorAtlas ..............................................................................................................76
5.1.3. BlueCats ...................................................................................................................76
5.1.4. Vuforia .....................................................................................................................77
5.2. The MVTA and Associated Project Requirements ...........................................................78
5.2.1. Testing and Debugging ............................................................................................84
5.3. The Physical Model ..........................................................................................................85
vi

Chapter 6 Conclusions .................................................................................................................. 87
6.1. Future Work ......................................................................................................................88
6.2. Technology Transfer .........................................................................................................89
References ..................................................................................................................................... 90
Appendix A User Guide ................................................................................................................ 96
Appendix B Draft Flowcharts ....................................................................................................... 97
Appendix C Audio Chapter List ................................................................................................... 99
Appendix D POI Photos ............................................................................................................. 100

vii

List of Figures
Figure 1 View from Tunnel View in Yosemite National Park along Highway 41 ......................... 1
Figure 2 View from the south side of the majestic Yosemite Hotel ............................................... 4
Figure 3 Location of the Majestic Yosemite Hotel in Yosemite National Park ............................. 6
Figure 4 Scanned Winter Club Room chair processed in the scanning software Skanect ............. 7
Figure 5 Catalogued Great Lounge American Indian Baskets ....................................................... 8
Figure 6 Majestic Yosemite Hotel Ground Floor used as the map in the MVTA .......................... 9
Figure 7 BlueCats BC313 beacon and coin beacon used in project ............................................. 19
Figure 8 IndoorAtlas Mobile application showing current location of user inside the hotel ....... 24
Figure 9 Virtual Tour of the National Museum of the American Indian ...................................... 27
Figure 10 Integrated interface of National Central University tour guide system ........................ 29
Figure 11 Quality report from IndoorAtlas guide ......................................................................... 39
Figure 12 3D Floorplanner model of the Majestic Yosemite Hotel and immediate outside
grounds .................................................................................................................................. 41
Figure 13 Update to back side of the bar area in 2017 ................................................................. 42
Figure 14 Anyplace Architect screenshot adding a building to Google Maps ............................. 44
Figure 15 Example of creating a location on the IndoorAtlas web API ...................................... 47
Figure 16 Zoomed in location of mapped paths (dark blue lines) and testing paths (light blue) in
IndoorAtlas MapCreator ........................................................................................................ 49
Figure 17 Portion of the Main.Storyboard for the IndoorAtlas version of the MVTA ................ 51
Figure 18 POIModel used in IndoorAtlas testing phase ............................................................... 53
Figure 19 BlueCats web API used for mapping individual beacons and scaling ......................... 54
viii

Figure 20 Cocoapods installation of BlueCats SDK ..................................................................... 55
Figure 21 Initial BlueCats testing and application structure ......................................................... 56
Figure 22 Sample for National Historic Landmark POI ............................................................... 58
Figure 23 A sample POI View Controller about the History of Yosemite and Stephen Mather
developed in Xcode ............................................................................................................... 59
Figure 24 Unity MVTA Game Scene on iPad resolution ............................................................. 61
Figure 25 Women's Restroom sign augmentability sample points and geometry ........................ 63
Figure 26 Restoration Stencil Work POI GameObject loading state ........................................... 64
Figure 27 Example of an initial Vuforia object recognition tests using Unity ............................. 66
Figure 28 Combined model view from north side of the hotel’s south wing ............................... 69
Figure 29 Scanned front side of the couch and adjacent wall in Winter Club Room ................... 70
Figure 30 Scanned portion of Winter Club Room ........................................................................ 71
Figure 31 Majestic Yosemite Hotel south wing ........................................................................... 72
Figure 32 Raw scanned data in Skanect of a Mural Room table (middle) ................................... 73
Figure 33 Birchwood model of south wing in Majestic Yosemite Hotel ..................................... 74
Figure 34 Splash Screen Image for MVTA .................................................................................. 81
Figure 35 MVTA map starting point and POI’s are shown as gold boxes ................................... 82
Figure 36 A sample POI in the MVTA ......................................................................................... 83
Figure 37 Majestic Yosemite Hotel Model ................................................................................... 86

ix

List of Tables
Table 1 Wireless-Based Indoor Positioning System and Solutions .............................................. 21
Table 2 Mobile Interpretation Applications in National Parks ..................................................... 32
Table 3 Possible Mapping Errors in IndoorAtlas ......................................................................... 40

x

Acknowledgements
I would like to thank my thesis chair Dr. Jennifer Swift for her help and guidance on this project
over the last ten months. I would also like to thank my committee members Dr. Yao-Yi Chiang
and Dr. An-Min Wu for their support during this project. I am grateful to my family, especially
my parents and grandmother, who have been there with me every step of the way during this
thesis process. I would also like to thank several people from Yosemite Hospitality for giving me
the opportunity to work on this project. These include Robert Concienne, Ashley McComb, Brett
Archer, Cassidy Nichols, Cory Goerhring, Zach Nagel, Dakota Snider, Tom Bertrand, Miles
Radin, Carlos Antunez, Breanne McNitt, Emily McIlvried, Bradley Olson, Erin Callanhan,
Majestic Yosemite Hotel maintenance and staff, and Dale Olander. I would also like to thank the
National Park Service for all that you have done, especially Jeffrey Trust and Virginia Sanchez. I
would like to thank Brandon Brooks for help building the model for this project and his constant
support throughout the program. I am also grateful for the Canvas Support team, especially Alex
for helping me with model development on this project. I would like to thank Harvard
University, Sierra Pacific University and the Muir-Hanna Trust, Alan Petersen and the Gunnar
Widforss Raisonne Project, Holly Cannan, Maynard Parker and the Huntington Library, Dakota
Snider, and Breanne McNitt for their photograph contributions in this project. A special thank
you to Jeffrey Dunk and Jennifer Tarlton for helping me these past couple years with advice and
support. It is with your assistance and guidance that I have been able to complete this project.

xi

List of Abbreviations
ADA American Disability Act
AP Access point
API Application Program Interface
AR Augmented Reality
FCC Federal Communications Commission
GIS Geographic information system
GPS Global Positioning System
IPS Indoor Positioning System
IR Infrared
LBS Location-Based Service
LTE Long-Term Evolution
MAU Monthly Active Users
MVTA Majestic Virtual Tour Application
PaaS Platform-As-A-Service
POI Point of Interest
POT Paths of Travel
RFID Radio-Frequency Identification
RSS Received Signal Strength
RSSI Received Signal Strength Indicator
SDK Software Developer Kit
TDOA Time Difference of Arrival
xii

TIMMS Trimble Indoor Mobile Mapping System
UHF Ultra High Frequency
USC University of Southern California
VR Virtual Reality
WWW World-Wide-Web

xiii

Abstract
The Majestic Yosemite Hotel, formerly known as the Ahwahnee, is a National Historic
Landmark located in Yosemite National Park, California, USA. Built in 1927, the hotel attracted
rich and wealthy individuals to help gain financial support for the National Park Service idea of
protecting wild spaces for future generations. To this day, the hotel stands as one of the National
Park Service’s most historic lodging units, providing luxury accommodations and services to
park visitors. In November of 2016 Yosemite Hospitality, Yosemite National Park’s
Concessionaire requested a mobile application to educate visitors on the cultural and historical
significance of the hotel to support the goals of the Long Range Interpretive Plan. Yosemite
Hospitality was the client for this project, and the application was developed in direct
consultation with Yosemite Hospitality’s Interpretive Services Department from November 2016
until August 2017. Several indoor positioning technologies and Augmented Reality services
were tested to deliver educational content based on user mobile device locations and camera
orientations. The processes tested the Anyplace indoor positioning service, IndoorAtlas indoor
positioning service, BlueCats beacon services, Vuforia Augmented Reality services, and the
gaming engine Unity. Testing and development occurred on both Android and iOS devices with
development in Javascript, C#, Swift, and Objective C. As part of this thesis work, a historical
model with digital furniture scans was also completed to preserve the current conditions of the
hotel’s original furniture. These scans were based on the Structure Sensor manufactured by
Occipital. This thesis documents the development and testing of the Majestic Virtual Tour
Application and the historic furnishings model built for the Majestic Yosemite Hotel in
fulfillment of the Yosemite Hospitality project.

1

Chapter 1 Introduction
Yosemite National Park, located in Central California, is one of the most well-known national
parks in the United States with over 5 million visitors in 2016 (Figure 1) (National Park Service
2017a). Since its inception, Yosemite has had its struggles with natural resource management,
particularly visitor encounters with wildlife. These include visitors feeding wildlife, car accidents
involving wildlife, and improper food storage, all things the National Park Service has been
trying to significantly reduce or eliminate through educational programs and park signage
(National Park Service 2016, 2017, n.d.a). Visitor education through interpretation continues to
be a top priority for preserving the park’s natural resources, and for the cultural and historical
resources in Yosemite National Park.

Figure 1 View from location “Tunnel View” in Yosemite National Park along Highway 41
To preserve some of the park’s cultural and historical resources, Yosemite Hospitality, a
subsidiary of Aramark and the new concessionaire for Yosemite as of May 2016, requested a
2

mobile application for the Majestic Yosemite Hotel that would showcase the history,
architectural design, and interior design elements (Yosemite Hospitality 2017a). Yosemite
Hospitality currently offers a Majestic Yosemite Hotel Historic walking tour available for park
visitors, but the addition of a mobile application will provide visitors with more freedom to
independently explore at their own pace to learn about the hotel. Another important reason why
Yosemite Hospitality requested a mobile application was to provide educational services to
visitors in eight different languages. Although these translations were not completed at the time
of this writing, this will be touched on in future work.
The Majestic Virtual Tour Application (MVTA), developed as part of this thesis project,
tells the story of the hotel and its creation using place-based interaction. Guests are now able to
check out Android tablets and iPads preloaded with the MVTA at the Concierge desk. This thesis
outlines several methodological approaches for spatial and augmented technologies to meet
client expectations and user needs. These include BlueCats Bluetooth Beacon services, Vuforia
Augmented Reality (AR) services, Indoor Atlas indoor positioning services, and the gaming
engine Unity (BlueCats n.d.; Vuforia n.d.; IndoorAtlas n.d.; Unity n.d.). All of the educational
content in this thesis was written for the project, and was approved by the National Park Service
and Yosemite Hospitality’s Interpretive Services Department.
In addition, several modeling technologies were looked at and used for this project. The
main modeling tool used was an infrared (IR) camera developed by Canvas.io. The IR camera is
an iPad attachment that can be used in conjunction with an application to scan rooms and
objects. Based off of these IR scans, a virtual model of the hotel’s first floor, a CAD model
featuring the Solarium, Winter Club Room, Mural Room, and Lounge, and a physical version of
the CAD model were all developed. To preserve the current conditions of the historical furniture,
3

each individual piece was scanned and documented, and will eventually be available in the
physical model display located inside the hotel.
1.1. Majestic Yosemite Hotel Background
Completed in 1927, the Majestic Yosemite Hotel, formerly known as the Ahwahnee, has
a rich and significant history coinciding with the formation of the National Park System and
National Park Service (Figure 2) (Yosemite Hospitality 2017b). The idea behind having a luxury
hotel in Yosemite National Park came from the National Park Service’s first director Steven T.
Mather. Steven T. Mather wanted a luxury hotel in Yosemite to attract wealthy park visitors in
order to acquire more interest and generate Congressional funding for the National Park System
(Sargent 1990).
The hotel’s stonework mimics the natural surroundings to support the goal of an
environmental architecture, as these designs fit in with the national park idea (Walklet 2004;
Sargent 1990). When the hotel was built, the developers did not use the park’s natural resources,
but instead, all the granite, stone, and other various materials were shipped in using railroads and
mules (Sargent 1990). Preserving and protecting the park’s resources set a precedent for other
National Parks for natural resources management and development.
Fire is a natural process in Yosemite’s ecosystem, and with fire suppression in
combination with the seven-year drought in California, there is a high fuel load in Yosemite
National Park. During the course of this thesis project, there were three separate fires within and
just outside of Yosemite National Park in July and August of 2017. As of August 15th, 2017, one
of the fires is currently threatening the Historic Wawona Hotel along Highway 41 in the park
(Rocha 2017).
4

Figure 2 View from the south side of the Majestic Yosemite Hotel
Since fire is common in Yosemite, the hotel needed to be fireproof. Sugar pine trees were
hollowed out and used as a mold for the concrete to appear as if it were made of wood, but closer
examination shows the same grain and appearance throughout the hotel. A combination of
concrete, granite, and steel were used in the construction of the hotel, making it completely
fireproof in its structure (Walklet 2004). At the time, the hotel not only stood out because of its
unique exterior design but was truly defined by the interior decorations and world-class
amenities (McComb 2017; Walklet 2004; and Sargent 1990).
Over the last ninety years, the Majestic Yosemite Hotel has provided luxury
accommodations to Yosemite National Park visitors including President John F. Kennedy, and
more recently, President Obama (Sargent 1990). It is one of the most famous National Park
5

Service’s lodging units, known for its elegance and significant history in helping shape protected
lands across the United States. A closed-door dinner by Stephen Mather in 1927 in the hotel’s
dining room inspired some of the wealthy guests to write and share their experiences in
Yosemite, further spreading the National Park Service idea and protecting these wild places into
what it is today (Walklet 2004; Sargent 1990). This project focused on developing an educational
mobile application to discuss the history and significance of the Majestic Yosemite Hotel.
1.2. Project Scope
The Majestic Yosemite Hotel is located in Yosemite National Park, as shown in Figure 3.
The hotel has 97 hotel rooms, a large dining hall, a bar, a sweet shop, and is home to many
different kinds of historical artifacts and objects. These objects represent cultures from all around
the world, but also more local influences from the seven different Native American tribes in the
greater Yosemite area (Farneth et al. 2011a, 2011b; Kollath McCann Creative Services 2011;
Walklet 2004; Sargent 1990).
6

Figure 3 Location of the Majestic Yosemite Hotel in Yosemite National Park
The educational content included in the application is separated into different sections
based on geographic location. To establish a flow, several chapters were created to guide users
into a certain order based on content, but each point of interest (POI) has stand-alone information
to allow flexibility for the end user to choose where they want to go and what topic they would
like to learn about. The content of the project focuses on topics including furniture, historical
rugs, American Indian baskets, architecture, gothic chandeliers, and the Firefall event that
occurred until 1968.
Much of the furniture pieces inside the hotel are original but have undergone several
restoration processes. When the Navy occupied the hotel during World War II, much of the
original furniture was taken out of the hotel to make room for the hospital beds. Several train
cars went into the Merced River, and some of the original chairs and tables were lost (Sargent
2000). In 2011, McCann Creative Services renovated the Great Lounge to help restore and
7

preserve the furniture and interior design based on design forensics (McCann Creative Services
2011). They looked at historical images, records, and documentation to try and replicate the
original conditions that were present when the hotel first opened. The model for this project
focused on the restored furniture in the south wing, consisting of the Mural Room, Lounge,
Solarium, and Winter Club Room of the hotel. The scans were completed to preserve a snapshot
as to what the original furniture looks like in the present day (Figure 4).

Figure 4 Scanned Winter Club Room chair processed in the scanning software Skanect
The rugs that can be seen hanging from the walls are known as “Kelim” carpets and were
originally used to protect furniture when shipped across overseas (McComb 2017). Now deemed
priceless and rare, the hotel now has the largest collection of Kelim carpets in the United States
(McComb 2017; Walklet 2004; Sargent 1990). The decorative and utilitarian American Indian
baskets that can be seen in the hotel’s Great Lounge have been restored or replaced over the
course of the hotel’s history (McCann Creative Services 2011). The current baskets have been
8

cataloged by National Park Service historians and are now on display in the glass cases inside the
great lounge (Figure 5).

Figure 5 Catalogued Great Lounge American Indian Baskets
The MVTA provides tour services for the first floor of the hotel only. The other five
floors of the hotel contain guest rooms and guest only facilities thus were not included within the
scope of this project. Many of the relics inside the hotel are kept on the first floor and are most
easily accessible to guests and park visitors. The floorplan and map that was used in this project
(Figure 6) show the extent of the MVTA and the MVTA’s educational stops or POIs as peach
colored dots. The hotel’s grounds include the back lawn at the south entrance and the grass patch
to the northeast.

9

Figure 6 Majestic Yosemite Hotel ground floor with locations of educational content (POIs are
indicated by peach colored dots)

1.3. Motivation and General Objective
First and foremost, the purpose of this project was to meet the desired user experience for
the Yosemite Hospitality. The desired user experience is defined as creating an overall engaging
10

mobile tour application through relevant information with the ability to select or receive
information based on where the MVTA user is inside the hotel, an audio component where the
user can choose to listen to the content or access a complete audio tour, a clean user interface
layout and design, and incorporating historical and present-day photographs to establish
relevancy to the educational content for each POI. All MVTA application design decisions were
based on client application specifications regarding functionality. Working with the client closely
facilitated achievement of those expectations.
To help meet the requirements of the thesis project, a second goal was set to test the
utility of an IPS with mobile integration for the indoor and outdoor transition. A third goal
included testing the applicability of different software architecture approaches for coding the
application to determine the system best suited to meet the client’s long-term needs and in-house
maintenance expertise. The fourth goal of this project was to build a historical furnishings 3D
model of the hotel for visual purposes to provide a current snapshot of the hotel and provide a
current inventory of the relics, artifacts, and furniture on the hotel’s first floor as part of the Long
Range Interpretive Plan.
From a spatial and technological standpoint, the motivation stemmed from facilitating the
indoor to outdoor transition using indoor positioning services. The natural outgrowth of
Geographic Information Systems (GIS) came from location storage, defense, and utilities
(Worboys 2011). The indoor to outdoor transition holds enormous potential (Goodchild 2011).
Using both an indoor and outdoor space to deliver educational content through proximity-based
location services can create an interactive and highly engaging virtual tour and help achieve the
desired user experience. This thesis looks at several technological approaches to achieve the
desired user experience for the client using different types of spatial technologies.
11

The intended users of the mobile application are guests and visitors of the Majestic
Yosemite Hotel in Yosemite National Park. Currently, there is only one historical in-person tour
available each day at the Majestic Yosemite Hotel. There is no limit as to how many guests can
attend this interpretive and historical tour to learn about the hotel’s history. The application was
designed to complement and provide an additional source of information for park visitors who
are curious to learn about the hotel.
The proposed methodologies of the application are focused on AR and the use of an
indoor positioning system (IPS). The AR feature employs object recognition services to identify
what the camera is looking at, and information about that POI will be displayed on the tablet
device. An advantage to this is if the object is eventually moved, it is not dependent on a place-
based information set, and the user can still use the camera function for object recognition and to
learn about the object.
Using an IPS is meant to inform users on where the POIs are located, what rooms are
restricted in the application, and help determine the current location of the user. The IndoorAtlas
indoor positioning service tested allows for geofences to be created. Geofences are virtual zones
that can be programmed into the application. When the user physically enters or exits these zones
by walking, the application can react to that specific action, whether it is based on if they entered
or exit, and display relevant information or take it away. Although this method was not
ultimately used due to final financial and logistical considerations, the client did like the
functionality of geofencing and the IndoorAtlas platform.
Over thirty POIs are available in the MVTA. From an educational standpoint based on
the client’s needs, the MVTA required high-quality content. The research was conducted at the
National Park Service Research Library, and the information was vigorously edited based on
12

content, flow, and clarity by the Interpretive Services Department and National Park Service
Interpretive staff.
1.4. Thesis Organization
This thesis outlines the processes and decisions made in the development of the MVTA
that demonstrate several positioning system utilities. The MVTA is intended to provide an
educational experience for visitors to learn about the park’s cultural and historical resources in
the Majestic Yosemite Hotel. Chapter 2 of this document provides a literature review of IPSs,
comparable mobile tour applications currently in use, and indoor modeling techniques. Chapter 3
describes the methodologies developed for this project, including an analysis of two low-cost IPS
services used in this project, testing BlueCats beacon services, Vuforia AR services, and building
an application in Unity (Anyplace Architect n.d.; IndoorAtlas n.d.; BlueCats n.d.; Vuforia n.d.;
Unity n.d). Chapter 4 outlines developing the model for the project. Chapter 5 describes the
results of the project and its applicability to the client. Chapter 6 includes a discussion and
closing thoughts on the project.

13

Chapter 2 Related Work
To understand the rationale for this project, it is important to understand the utility of an IPS.
Positioning technologies have always favored the outdoors, but the development of IPS
technologies have given established companies and start-ups a unique opportunity to fill the
market and define the space (Dodge 2013; Goodchild 2011). For example, companies can use
IPS data to better serve their customers by providing direct navigation to a specific POI, an IPS
can deliver content based on the user’s location within a specific location, or it can identify
shopping trends tailored towards a shopper’s experience. This attracted Yosemite Hospitality’s
desire for a unique user experience and overall engagement for place based interaction.
To achieve the desired user experience for the client, several approaches for content
delivery were tested, including object recognition services and IPSs. This chapter highlights
several positioning services available on the market at the time of this writing, and comparable
self-guided tour mobile applications and web applications. A movement in the GIS field focuses
on the indoor capacity and 3D indoor modeling, specifically in emergency and evacuation plans
(Xiong et al. 2017). This chapter also highlights indoor model development.
The first section of this chapter provides an overview of positioning technologies and
capabilities. The second section describes different applications for tour purposes, including
those that use Virtual Reality (VR) or AR tools. The third section focuses on indoor modeling
techniques and indoor scanning techniques.
2.1. Types of Positioning Systems and Technologies
There are many different types of positioning systems currently available including
Global Positioning Systems (GPS) and IPS. This section provides a brief overview of different
14

types of positioning systems available at the time of this writing and their respective
technologies.
2.1.1. Global Positioning System
GPS is a commonly used positioning technology that relies on satellites. GPS was
originally developed for military purposes in the United States and consists of twenty-four
satellites in six orbital planes to determine coordinates points that represent a location on the
Earth’s surface (Bahl and Padmanabhan 2000). The coordinate values are calculated by
comparing the timing of satellite signals in the line of sight of the user (Goodchild n.d). Trees,
clouds, and buildings are examples of obstacles that can disrupt the line of sight and the GPS
signal. Since GPS is not as effective indoors, it was not used for this project.
One utility GPS does provide though is a location-based service (LBS) in the form of
coordinates. LBS progressed in the late 1990s and early 2000s with pressure from the Federal
Communications Commission (FCC) to improve emergency positioning technologies (Rao and
Minakakis 2003). In an emergency, cellular carriers are required to identify an individual’s
location within a certain level of accuracy in order to identify and alert the correct entity. An
LBS usually involves the user, a positioning technology provider, and a service provider
(Gruteser and Liu 2004). Consumers continue to benefit from LBS with smartphone integration
for navigation or finding local goods and services, applications dependent on the context and
location of the device (Jiang and Yao 2006).
In an indoor space, the level of accuracy and precision performance is crucial to
delivering accurate information and services. In the Majestic Yosemite Hotel, there are very
narrow, roughly ~1m in width, paths of travel (POT). This could create potential issues when
entering and exiting the proposed geofences that may be in proximity to another POI in the hotel.
15

One solution, which will be explained in further detail later in this chapter, was to use object
recognition and AR services.
GPS in its pure form is generally not used in smartphone technology, but rather a
combination of technologies called A-GPS or Assisted GPS. Assisted GPS combines network-
based and handset-based technologies in an attempt to overcome the drawbacks of GPS (Fouskas
et al. 2002). These drawbacks include power consumption, speed, line-of-sight, and cost, as well
as keeping an estimated location based on the last recorded position (Fouskas et al. 2002).
Assisted GPS is a hybrid because it uses a combination of network services with available
wireless technologies (Wi-Fi). This combination of technologies gives LBS to the indoor/outdoor
transition, an integration that holds enormous potential according to Goodchild, but does not
necessarily provide location-based solutions for every situation (2011).
Giudice et al. mention that one fundamental research issue is the integration of an
outdoor to indoor space framework (2010). It is important to note that an ontology is specific to a
domain, and permits formal reasoning while using a computer (Zouggar et al. 2008). In order to
address this research integration, a generic ontology should be developed where O-space and I-
space are considered special cases for an OI- integration tool (Giudice et al. 2010). An OI-
ontology would provide seamless integration at the fundamental level. This project intended to
use an IPS for managing this outdoor to indoor integration on a small scale to demonstrate the
utility of an IPS, but more importantly to treat indoor and outdoor space as a single type of space
for smooth transitioning.
Although GPS and A-GPS are more commonly available positioning technologies, they
do not support or have the capabilities in a remote area such as Yosemite National Park. The
high amount of tree cover and granite cliffs in Yosemite Valley create obstacles for GPS to
16

acquire a signal and factored into the decision for which technology to use for an indoor/outdoor
virtual tour. GPS does offer fundamental characteristics of positioning technologies and is a
common technology available in most smart devices.
2.1.1.1. GPS Accuracy
An important thing to note about GPS is that its success depends on unit accuracy.
Separate from mobile devices, GPS units offer positioning solutions in outdoor settings. Outdoor
GPS technology has improved in covered areas or areas with some level of interference, but
these types of units can be very expensive to the average consumer. Lower cost GPS units may
be more sensitive to interference and obstacles (Abraham et al. 2012). The quality of the unit and
level of object interference can directly affect the position accuracy. Readings can range from a
few meters to below centimeter accuracy, depending on the quality of the GPS unit (Goodchild
n.d.).
Having a system that has sub-centimeter accuracy provides higher quality data, but can
also lead to security concerns and issues. An individual may approve certain smart phone
applications to track their location, whether it’s for navigation or emergency services, but may
not want their location revealed or tracked in every situation. This is known as situation-
dependent LBS (Gruteser and Liu 2004). Many phones now can turn off LBS for applications,
but it may be considered tedious to turn this setting on and off (Gruteser and Liu 2004).
One alternative for inherent privacy issues in an LBS can be found in the Empire State
Building in New York City, New York (The Travel Magazine 2017). Visitors can rent out iPads
with a preloaded application for a virtual tour of the New York skyline. Based on that example, it
was determined that the first 12-months would provide only rented tablets available with MVTA
17

preloaded from the concierge to address security concerns. This will also control the number of
people that can participate in the virtual tour simultaneously.
2.1.2. Wide-area Cellular Networks
Wide-area cellular networks provide another form of user location and tracking. This
form of tracking relies heavily on cellular network infrastructure and technologies (Moberg et al.
2009). Examples of a wide-area cellular network are long-term evolution (LTE) and 4G-LTE.
These provide the necessary infrastructure for GPS, navigation, and other LBS, common
applications in today’s mobile technologies, as previously stated.
Ideally, a wide-area cellular network would be the preferred method for a mobile
application. Wide-area cellular networks provide a commonly-used form of user location and
tracking. In 2015, an estimated 92% of adults owned a cell-phone, and 68% owned a smart
phone (Anderson 2015). Although accessibility to mobile technologies is high in the general
populous, lack of dedicated cellular service in Yosemite National Park contributed towards using
the proposed network-dependent IPS. In more developed areas, this could be a better fit solution
for a virtual tour if high-level indoor accuracy could be achieved.
2.1.3. In-building Positioning Networks
Although there are many emerging technologies for IPSs, there is no accepted standard
for their implementation (Davis et al. 2015). This project does not aim to define standards for
implementing an IPS. The project does provide a methodology of implementing a low-cost IPS
as an alternative to other positioning systems that may not have the optimal cellular network
infrastructure or environmental conditions to support GPS signal acquisition to demonstrate an
IPSs’ utility. Another approach considered was using an In-building Positioning Network.
Examples of these are Bluetooth, Wi-Fi, ultrasound, IR, or radio-frequency identification (RFID)
18

(Fouskas et al. 2002). A system requiring Bluetooth beacons, ultrasound and IR sensors, or RFID
sensing capabilities would increase initial costs but may be cheaper for long-term maintenance.
The following sections cover a brief description of Bluetooth, ultrasound, and RFID location
services.
2.1.3.1. Bluetooth
Some services offered such as BlueCats Bluetooth beacons offer location services at the
cost of the technology itself, but also have the option to charge for their SDK for user tracking
and characteristics for a monthly fee or site license. For indoor positioning services, two of the
five states of Bluetooth technology, advertising, and scanning, help determine the location of the
devices (Conteras et al. 2017). In order to protect the privacy of the individuals in a Bluetooth
IPS, BlueCats uses the advertising ID instead of any personal information to reduce the risk of
information being leaked (BlueCats 2017).
Location errors have been known to occur, as with many different types of positioning
systems, but there are ways to reduce or eliminate errors in some cases. Smoothing algorithms
can be adapted to a specific space and is a key component to location accuracy in the proposed
methodologies for the MVTA. Each room required one or more smoothing algorithms to
determine the boundaries of each room. Conteras et al. (2017) recommend reducing cross-
correlation issues, for example, by removing stronger signals so that the weaker signals can be
favored. For BlueCats beacon services, the midrange beacons callback method available in their
software developer kit (SDK) can provide a list of beacons objects in the range where the
property “accuracy” is called (Dundee 2017). The beacons with the lowest accuracy represent the
closest in proximity and based on that number the scale can be set based on the needs of the
19

application (i.e., 1-2 meters). This allowed for the beacons to deliver content based on the
proximity of each beacon once the distance was set.
The beacons from BlueCats ranges between $15 USD for a coin beacon and $29 USD for
their second-generation BC 313 (Figure 7). Some projects require hundreds of beacons to be
deployed, but on this scale and methodology, it was determined that only thirty were required.
This was one of the costlier purchases for the MVTA, but it provides an easy and reliable IPS
service for the application.

Figure 7 BlueCats BC 313 beacon (76mm x 27.80mm) on left, and coin beacon (52mm x 27mm
x 7mm) on right used in project

20

2.1.3.2. Infrared Radiation IPS
One type of indoor positioning technology uses IR sensors. IR sensors use computing
devices for current location and allow the use of transmitters to automatically display
information about the POI within that proximity (Tseng et al. 2001). This technology is
appealing because it increases location-based interaction based on proximity to a POI, but
requires more technology investment up front. The National Park Service had concerns about
several hundred sensors installed near each POI as they would be noticeably visible inside the
hotel. From the client’s perspective, the cost was a large factor looking at IR IPS systems. Some
sensors can cost up to several hundred United States dollars (USD) apiece, so to reduce costs an
IR Radiation-based IPS was excluded.
2.1.3.3. Wireless Systems
Using a system that relies on already in-place infrastructure such as Wi-Fi is a simple
way to reduce costs. One of the main objectives of this project was to reduce the ultimate cost to
the client. Finding a low-cost IPS system with high performance and accuracy to demonstrate the
utility for virtual tour purposes in a small environment was vital. Two hardware components for
these types of systems are signal transmitters and measuring units (Liu et al. 2007). Many mobile
smart phones come with standardized Wi-Fi and can be used with these types of systems. Table
1 describes several Wireless-based IPSs, their specific technologies, and accuracies. Cost ranges
in the table represent the combination of money, time, space, weight, and energy for the
technologies. One disclaimer for Anyplace is that there is not an already-in-placed wireless
network in which the IPS could use.

21

Table 1 Wireless-Based Indoor Positioning System and Solutions (Adapted from Liu et al. 2007;
Georgiou et al. 2015)
System/
Solution
Wireless
Technologies
Positioning
Algorithms
Accuracy Precision Scalability Cost
Microsoft
Radar
WLAN,
Received Signal
Strength (RSS)
K NN,
Viterbi-like
algorithm
3-5m 50% within
around
2.5m and
90% within
5.9m
Good/2D, 3D Low
Horus WLAN RSS Probabilistic
method
2m 90% within
2.1m
Good/2D Low
DIT WLAN RSS MLP, SVM,
etc.
3m 90% within
5.12 m for
SVM; 90%
within
5.40m for
MLP
Good/2D, 3D Low
Ekahau WLAN Received
Signal Strength
Indicator (RSSI)
Probabilistic
method
(Tracking-
assistant)
1m 50% within
2m
Good/2D Low
SnapTrack Assisted GPS,
time difference of
arrival (TDOA)
5m –
50m
50% within
25m
Good/2D, 3D Med
Anyplace RSS Local IMU
and Wi-Fi
1.96m Good/2D Free
WhereNet Ultra-High
Frequency (UHF)
TDOA
Least
Square/
RWGH
2-3m 50% within
3m
Very good/
2D,3D
Low

The Anyplace indoor information service is one example of a Wi-Fi IPS that was
considered for this project. Anyplace is an indoor information service that combines network-
based service accuracies with a terminal-based service (Georgious et al. 2015). The five
components that comprise the Anyplace application are Server, Architect, Viewer, NoSQL data
store, and Logger/Navigator. This creates a full-service, scalable, open, modular, and extensible
architecture (Georgious et al. 2015). Anyplace stores buildings, floorplan data and POIs, and
uses Couchbase for database information retrieval (Petrou et al. 2014; Couchbase n.d.).
22

Information is continuously updated through crowdsourced information and data input for
location services, providing built-in long-term maintenance capabilities (Petrou et al. 2014).
The Anyplace Architect is a website feature service that allows the user to edit and
upload floorplans for each respective floor for any building (Anyplace Architect n.d.). Floorplans
are placed on a built-in Google Maps Application Programming Interface (API) for scaling and
resizing. Additional POIs inside the building can be uploaded, and feasible paths for indoor
navigation services, also known as POT, are created directly in the Anyplace Architect. The
Anyplace Logger uses Wi-Fi access points (APs) to track and record current locations and
contributes to the overall radio map (Petrou et al. 2014). The Anyplace Navigator gives users
their current location within the building based on exchanges from smartphone or device sensors
with Wi-Fi APs to obtain navigation direction from their current location to different POIs
(Petrou et al. 2014).
Researchers from the University of Cyprus in Greece developed the Anyplace technology
and implemented this type of IPS on the university campus (Zeinalipour-Yazti et al. 2017). The
Anyplace mobile application is currently available on Android for free, and the entire application
is available on their Github (Anyplace Github 2017a). The main functions of the Anyplace
application are for navigation purposes to different POIs and indoor positioning but could be
used for virtual tour purposes. Navigation using defined POT were considered for the user to find
the next POI.
Wireless technologies provide an easy solution for IPS services, and many public
buildings and outdoor areas offer Wi-Fi hotspots that customers can use. The Majestic Yosemite
Hotel already offers Wi-Fi access for hotel guests and park staff, but does not have a public
network available for non-guests. This proposed a fundamental problem as a separate network
23

would need to be provided for guests to use the MVTA, adding on additional infrastructure costs
to the client.
Another main issue with Anyplace is that there have been documented issues on the
Anyplace website with native mobile application integration (Anyplace Github 2017b). Since the
code is freely available, a developer can recreate the code to fit into their own application, but
other researchers have also had no success in implementation. Such efforts, including in this
thesis, have reported experiencing critical functionality errors and locational display errors. This
was considered a major drawback in using this type of system.
2.1.3.4. Geomagnetic Systems
Another type of IPS considered for this project that uses already in-place infrastructure is
based on the Earth’s magnetic field and is called a geomagnetic system. One example of a
geomagnetic platform-as-a-service (PaaS) leader is IndoorAtlas, an IPS service that provides
indoor location services using built-in magnetic sensors in mobile devices (IndoorAtlas 2016).
IndoorAtlas uses a similar web API to upload floorplans and building data, and the user
downloads a mobile application (Figure 8) to record the geomagnetic pulses inside the building
for POT and room information (IndoorAtlas 2016). One contingency is that IndoorAtlas requires
steel in the buildings because the interaction between the steel and the geomagnetic pulses
determines users’ locations within the building. The farther the user is from the building, the less
accurate the positioning is. Since the Majestic Yosemite Hotel is completely fireproof and made
out of only concrete and steel, it was a deemed a suitable location for testing the IndoorAtlas
service.

24

Figure 8 IndoorAtlas mobile application showing current location of user inside the hotel
25

The IndoorAtlas services are available to developers using a “Freemium” rate, so the
initial costs are free, then increase based on the amount of average monthly users. IndoorAtlas
services are provided in an SDK, which offers developers easy application integration.
IndoorAtlas’s patented technologies provide location accuracies around 2-3 meters, comparable
to other high-end IPS systems (IndoorAtlas 2017a). IndoorAtlas provides one of the only
geomagnetic positioning systems currently available and was the IPS that produced the most
successful prototype while testing in this project.
Since the application is only available on a set number of devices, knowingly using
IndoorAtlas’ services under 100 users in the “freemium” model in perpetuity would raise terms
of service concerns. IndoorAtlas does offer a flat rate service fee based on specific project needs,
but the estimated costs for the service would be around $100 a month. Although IndoorAtlas
provided the most user-friendly experience when implementing an IPS, it does require a data
signal of some sort, and therefore could not be guaranteed throughout the tour for guests should
the application launch for public devices.
2.2. Existing Tour Applications
The field of interpretation forges connections between the audience and the inherent
meanings in the resource (NAI n.d.). Non-personal interpretation fosters the connection between
the visitor and the resource using signs or mobile applications, instances where an interpreter is
not presenting the message or theme directly. Understanding why protected areas exist through
interpretation can reduce negative impacts of tourism and improves visitor enjoyment
(Danwandee et al. 2015).
New technologies have made progress in the virtual tour industry, including more
technological approaches to deliver information in new and exciting ways. Traditionally,
26

museum tours may have pointed to objects with signage placed in front as the main
communication method. In this setting, the user looks at only text and the actual image or object,
with limited engagement and interaction. The communication only flows one direction, and the
information presented is limited to the text provided. This section explores tour and virtual tour
technology progression using the web and mobile applications.
2.2.1. Virtual Tour Web Applications
In the early 2000s, web application technologies represented a major focus point for
virtual tours. Users could connect to the World Wide Web (www) on their personal computers
and devices to explore different museums that had this technology available. Visitors could learn
about the artifacts and objects in the particular location even though they were not physically
present. One web application example is the Virtual Tour of the National Museum of the
American Indian (Figure 9).
27

Figure 9 Virtual Tour of the National Museum of the American Indian (Jones and Christal 2002)
The National Museum of the American Indian virtual tour provides an interactive map,
object images, and a text description. Created as a collaboration project between the National
Museum of the American Indian, the Four Directions project, and the Bureau of Indian Affairs,
this tour includes web links to other resources where the “virtual visitors” can learn more about
the items in the museum and the Native American cultures represented, without being physically
present in the museum.
The MVTA contains several common elements used in the National Museum of the
American Indian virtual tour. These include text, photos, and a map of the hotel and grounds,
which are all basic components for a tour. Another common element between the MVTA and the
National Museum of the American Indian virtual tour is the goal behind it. The Four Directions
project and the National Museum of the American Indian shared the desire to offer meaningful
learning experiences in museums and schools (NMAI n.d.). These meaningful learning
28

experiences are similar as to what Yosemite Hospitality wanted to achieve with the desired user
experience for the MVTA.
Another early 2000’s web-based tour application was developed for the fourteen
Smithsonian Museums. Users were not required to be physically present to learn about the 360
artifacts highlighted in the Virtual Smithsonian web application. This tour included high-
resolution images, videos, audio clips, and 3D rendered artifacts featured throughout the
museums (Jones and Christal 2002). The limitations of having 3D rendered artifacts in a web
application required the end user to have a faster internet connection speed. In the early 2000’s, a
faster internet connection was much more expensive and less accessible to the average consumer.
This resulted in end user access issues and did not necessarily provide the best user experience.
In 2013, a more updated web application featuring the Smithsonian National Museum of
Natural History was developed and released by Loren Ybarrondo (Smithsonian Institution n.d.).
Here, the user can select a virtual tour of past exhibits, virtual tours of permanent exhibits, and
other Smithsonian exhibits that may be added in the future. The user has the freedom to access
the virtual tour using either a desktop computer or on a mobile device. In addition, using the
high-resolution panoramic photos taken for each room can help navigate the user through the
virtual exhibit, providing educational information and content anywhere there is an internet
connection.
An even more recent update to these virtual tours includes Google Cardboard VR
capabilities. Users can download the Google Cardboard VR application plugin to view the
Smithsonian presentation in VR (SNMNH n.d.). To navigate from room to room, the user
focuses on the navigation arrows within the application. The user can also zoom in and out, as
29

well as move around the scene, regardless of where they are. This is a potential direction that the
MVTA could eventually reach using similar VR technologies.
One example of a place-based web application is the National Central University Tour
Guide system (Tseng et al. 2001). Researchers at the National Central University in Taiwan
prototyped a location-aware and context-aware tour guide system. The prototype included a 2D
map, showing the current position of the user, a 3D virtual world where users can track their
movements and find different POIs, and a web feature that provided additional information
through a web browser (Tseng et al. 2001). This system has both an educational feature and an
IO- location feature. These represent two major components of this project. Figure 10 displays
the National Central University tour interface.

Figure 10 Integrated interface of National Central University tour guide system (Tseng et al.
2001)

30

Tour participants used a laptop that was connected to a GPS receiver to determine the
current location. Participants were also able to communicate with other users (Tseng et al. 2001).
An advantage of this type of system is that the system is location-aware and context-aware,
meaning if the user is inside or outside, it can determine location based on either the GPS signal
for outdoor use and indoor based on contextual information (Tseng et al. 2001). This type of
system can provide users with directions and information on how to get to specific locations and
information about each area.
2.2.2. Mobile Tour Applications
Mobile capabilities and technologies have proven to be more capable than ever before.
Mobile application utility and pervasiveness continues to increase as more services become
available (New Media Consortium 2012). Recent mobile application tours provide users with
navigation, IPS integration, and LBSs indoors. These technologies increase place-based
interaction with the resource or objects of interest in places such as universities, museums,
hospitals, and conferences. The flexibility and utility of a mobile tour application using IPS
services gave Yosemite Hospitality different options to achieve the desired user experience,
which was a major driving factor for this mobile application. This section explores virtual tour
mobile applications, including those available in National Parks.
An example of an IPS project includes the one used by the Smithsonian museum and
National Zoo located in Washington D.C. A collaborative project with Google, the Smithsonian
mapped out over 2.7 million square feet for indoor navigation purposes (Smithsonian 2012). The
IPS was built right into Google Maps for Android and allowed users to navigate to specific
exhibits at the museum and National Zoo. An advantage of building an application right into
Google Maps is that it is more widely available and accessible for users. Users do not have to
31

download any additional applications or plugins to use it, and can directly access the content
based on their location.
2.2.2.1. Mobile Interpretive Tour Applications in National Parks
The National Park Service works with partners, universities, non-profits, and third-party
developers to develop mobile applications for several national parks. The MVTA was developed
to educate park visitors about the hotel. According to the National Park Service, most of the apps
previously developed have focused less on education and interpretation aspects, and more on
directions and planning (National Park Service n.d.b). Blaser (2015) looked at several mobile
interpretive applications that varied in geographic area, content, capabilities, and level of
engagement. The Chesapeake Explorer Application provides a basic informational mobile
application but does not provide any map services. It offers textual directions to the sites and
visitor fee information.
The National Park Service National Mall and Memorial Parks application are more
similar to a virtual tour. It provides self-guided trail capabilities and audio tour services for park
visitors, as well as Google Maps integration and web integration where users can learn about
each of the 70 cultural sites through an interactive map (Blaser 2015). National Park Service
Independence National Historical Park also includes self-guided tours, program information, and
general site information. The “Park Lens” feature uses AR services and a mobile device’s
camera to find out what is around the user (National Park Service n.d.c). This includes labels of
buildings, memorials, and other POIs at the National Park Service National Mall and Memorial
Parks.
The Boston National Historic Park and Boston African American National Historic Site
has a joint mobile application that visitors can download ahead of time to learn about the “Trails
32

to Freedom” (iTunes 2017). In addition to providing map and location services, the application
comes equipped with navigation features, audio tours about different POIs, and a general
overview of the area (National Park Service n.d.d).
There are currently a few available applications for Yosemite National Park. Many focus
on the climbing community and mapping over 2,000 of the routes in the park. A lot of these
applications are developed by third-party creators, and those entities are explicitly stated in the
respective app stores. The National Park Service Yosemite application provides simple
information about the park, focusing on the amenities, including the Majestic Yosemite Hotel.
Table 2 lists the currently active mobile applications relevant to the National Parks.
Table 2 Mobile Interpretive Applications in National Parks (Blaser 2015)
Application

Description
Chesapeake Explorer Contains directions for finding park entrances, fees, and
program information for over 400 National Park Service sites
in the Chesapeake region.
National Park Service
National Mall and
Memorial Parks
Contains map tours, site information, and the ability to create
digital postcards that visitors can customize while on their
trip. Application covers over 70 cultural sites and provides
interpretive information for each.
National Park Service
Independence National
Historical Park
Provides a park map, interpretive information about the site,
current events and program information, and self-guided
tours.
National Park Service
Yosemite
Features 50 landmarks in the park and amenity information.
Park programs and event schedules can also be found.

2.2.2.2. Yosemite Hospitality and Aramark Applications
There are currently no other applications owned and operated by Yosemite Hospitality.
Yosemite Hospitality is a subsidiary of Aramark. Aramark owns a few applications in the
hospitality business. The CampusDish mobile application allows college students to look up
33

nutritional information for food choices on their respective campuses (Food-Management 2012).
Aramark also partnered with Major League Baseball to provide food delivery service directly to
your seat at MLB games (MLB 2010). The MVTA is a first of its kind application for Yosemite
Hospitality, creating an engaging and educational mobile application for one of its biggest assets.
2.3. Indoor Modeling Techniques
There are several Mobile scanning technologies that are currently available and were
considered for this project. The more expensive mobile mapping units generally provided a faster
scanning and customizable output formats compared to cheaper solutions, but some units can
cost as much as $15,000 USD (Viametris n.d.; Applanix 2015; Geoslam n.d.). Rental services are
available for some units and can help reduce the cost to the client.
The lower cost Structure Sensor unit used in this project required more upfront scanning
time but processing time was reduced by their paid “Scan to CAD” service (Structure sensor
n.d.). This service converted the point-cloud scans to a CAD file. The goal with mobile mapping
and scanning for this project was to incorporate newer spatial technologies for the indoor area to
preserve the architecture and furniture in the hotel. Many of these units allow for easy GIS
integration and spatial modeling and would meet this project’s requirements, but costs and
resources heavily influenced the decision to go with a cheaper scanning unit.
Several high powered mobile scanning units are offered by Viametris, GeoSlam, and
Trimble. Although there are many others, these units all offer customization, user-friendly
framework, and high-level of accuracy (Viametris n.d.; Applanix 2015; GeoSlam n.d.).
Viametris is a continuous indoor mobile mapping scanner and includes full building
environments including furniture. This unit can be used in the outdoor capacity to scan exterior
building conditions, and produces average accuracies of ± 30 mm (Viametris n.d.).
34

Geoslam is a 100Hz handheld mobile laser scanner with a simple interface and operating
system (Geoslam n.d.). This scanning solution is time-efficient, it takes roughly 30 minutes to
scan a 3-story building, and an accuracy of around 15mm (Geoslam n.d.). Geoslam offers the
high accuracy and efficiency desired for this project, but the costs for the unit exceeded the
budget for this project and were not used.
The Trimble Indoor Mobile Mapping System (TIMMS) by Applanix can produce highly
accurate scans and models. Several case studies documented their accuracies, processing time,
and results. One case study reported over 1.8 million square feet of the Los Angeles International
Airport (LAX) in under 30 hours (Applanix 2015). In comparison, a traditional LiDAR method
would have taken over 3 weeks to complete the same amount of space (Applanix 2015). All
interior components were included in the model, and the post-processing time took around 100
hours for this project (Applanix 2015).
Another case study featuring TIMMS was coordinated with the mapping and survey
company Pictometry. The Pictometry project merged both aerial and indoor georeferenced data
to create Critical360, a technology used for querying building resources and structural details
(Applanix 2016). This type of unit was the preferred method for Pictometry because of the time
it took to scan 91 separate buildings and the level of accuracy the unit produced.
Since the project’s scope focused only on the first floor of the hotel due to time
constraints and money, a more realistic and hands-on approach was used to complete the indoor
model of the hotel. The Structure Sensor built by Occiptal is an iPad attachment for scanning and
measuring environments (O’Kane 2016). Each room was scanned individually and uploaded to
the “Scan to CAD” feature in the Canvas.io mobile application. This scanning technique reduced
processing time and overall costs to the client and produced the desired CAD files.
35

The Canvas iOS application with the Structure Sensor allows digitization for any room
instantly, including measurements of the scanned space. This modeling technology is cost
effective at $399, besides the cost of the iPad. The Structure Sensor also works for Skanect, a
scanning environment and software used to create 3D printed objects and rooms. This system
offered the necessary means to scan and produce the furniture models for this project.

36

Chapter 3 Methods
This project compared four different approaches to deliver educational content for MVTA users
to determine the technology best suited to fulfill the client’s project requirements. Each
methodology was chosen based on customer review and product functionality capabilities. This
chapter explains the rationale for each approach and the chosen methodology based on Yosemite
Hospitality’s needs. The initial approach to complete this project was based on testing two
separate IPSs for the indoor and outdoor spaces, one based on Wi-Fi provided by Anyplace, and
the second based on IndoorAtlas. The second approach to this project relied on the BlueCats
beacon and SDK services, which can deliver content based on proximity to the beacons. A third
approach tested for this project focused on the use of object recognition and beacons in Unity.
This methodology resulted in a shift from Android to iOS platform development, based on the
client’s requirements. The fourth methodology tested, and the most basic approach deployed
Unity without any additional software or infrastructure and provides services for both iOS and
Android devices. Long term maintenance, user-friendly application architecture, and costs to the
client ultimately decided the direction that was chosen for this project. This chapter provides an
outline to technologies that were ultimately used.
3.1. Choosing an Appropriate Indoor Positioning System Technology
Cost, time for implementation, accuracy, and long-term maintenance were determining
factors when selecting the appropriate IPS for this project. Chapter 2 explored different types of
IPSs. As previously stated, two low-cost IPSs that were tested as part of this project were
Anyplace and IndoorAtlas.
37

3.1.1. Cost Analysis
When looking at different factors that influence cost, infrastructure stood out as one
determining factor. Using an already in-place infrastructure such as a Wireless system is one
cost-effective approach for an IPS (Ott et al. 2011). Microsoft Radar, Horus, DIT, Ekahau, and
WhereNet are wireless IPSs with lower cost infrastructure (Liu et al. 2007; Georgiou et al. 2015).
The differences between these systems occur with the algorithms, and each has a different price
point based on the particular technologies used. Under normal circumstances, the availability of
a wireless IPS is higher because it is a common technology. However, there is limited internet
access only available for Majestic Yosemite Hotel guests, and an additional installation of a
wireless system dedicated to this project application was not in the budget. It was also
determined that all of the application information would be stored locally on the device since
dedicated wireless access would not be available for the tablets.
Another low-cost system that was analyzed for this project was IndoorAtlas. IndoorAtlas
provides a “freemium” service based on the monthly active users (MAUs). IPS services would
be available through IndoorAtlas’s Android and iOS SDKs. Initial tests do not require any
commitment up front which allowed for the service to be tested during this project. No financial
loss was suffered to the client during the testing of these IPS systems.
3.1.2. Time for Implementation
The initial start-up time for IPS implementation in a native application heavily favored
IndoorAtlas. Anyplace required the building to be “added” to a Google Map, and once the
floorplans were uploaded and resized, an indoor model was required for the POI components that
could be exported to JSON format. IndoorAtlas had a similar web API where floorplans were
uploaded, but this required the developer to go immediately into the physical space using
38

IndoorAtlas’ MapCreator application and collect measurement points for the geomagnetic
readings. The collection of these geomagnetic signals instantly gave the user a location on a map.
These readings were also available through the IndoorAtlas Cloud server, and the locations could
be added into a native app in either Google Maps for Android or Apple’s MapKit for iOS.
With regards to Anyplace, several factors prevented it from working in a native
application. Android source code was readily available through the Anyplace application and
was originally considered for use as a guidance tool for implementation of the project as a native
Android application (Anyplace Github 2017a). At the time of this writing, it was also possible to
develop an Anyplace implementation as a native iOS application.
3.1.3. Long-Term Maintenance
Another key factor for this project was to have a system with ease of long-term
maintenance capabilities. Anyplace uses an algorithm to identify data outliers from
crowdsourced heat maps and removes them to maintain accurate positions. In this case, the user
is responsible for maintaining accurate indoor positioning since it is crowdsourced data.
Compared to IndoorAtlas, Anyplace did not offer sufficient guides or technical support for
implementing and maintaining the services. When updates were implemented for Anyplace,
maintenance went well-beyond a single update to an SDK, but rather required an update to each
of the Anyplace Server, Anyplace Architect, Anyplace Viewer, NoSQL data store,
Logger/Navigator, and the model.
IndoorAtlas uses updates in the SDK to reduce overall updates to application code if the
errors or updates occur on the server side. If positioning needs to be corrected, a quality report
can be generated on the IndoorAtlas website where it outlines several reasons for inaccurate
information (Peltola 2015). These quality reports let the client know where the problem areas
39

are. If there is a need to go back into the field to collect data, the report can be automatically
updated using the IndoorAtlas application and uploaded directly to the cloud service (Figure 11).

Figure 11 Quality report from IndoorAtlas guide (Peltola 2015)
Table 3 lists several reasons why the errors may have occurred in the IndoorAtlas
mapping service. The table also provides solutions to fixing the different types of errors.
Other errors can occur caused by uneven walking speed and misplacement of landmarks (Petola
2015). Nevertheless, the solutions for long-term maintenance, consistent server updates on both
the cloud and SDK side, and overall technical support and guides initially heavily favored
IndoorAtlas for use in this project.

40

Table 3 Possible mapping errors in IndoorAtlas (Peltola 2015)
Problem Way to fix
Lack of nearby test paths: red/blue Record a test path through the
affected area
Lack of nearby mapping paths:
“hole” in the coverage
Record a mapping path through
the affected area
Problem in recorded test path data:
red through the test path
Delete the affected test path and
re-record the test path
Problem in recorded mapping
data: red through the test path
Delete the mapping path passing
the area and re-map
Problem persists after remapping You may have a problem with the
mapping device. Check that your
device meets the mapping device
requirements, reboot the device to
make sure any sensor errors are
cleared

3.2. Testing the Indoor Positioning System Methodology
Since both services provided no up-front costs and long-term maintenance solutions, both
were considered for use in the final MVTA. A digital indoor model was created using
Floorplanner as both a model and a map within the application (Figure 12) (Floorplanner n.d.).
This map was overlaid on both the Anyplace API and IndoorAtlas’s API. Specifically, the model
was created to define the spatial scale and extent of the hotel’s immediate grounds.
41

Figure 12 3D Floorplanner model of the Majestic Yosemite Hotel and immediate outside
grounds
Floorplanner was used to develop one of the 3D models of the hotel. The 2D version of
the model was used as a Floorplan and map in the final version of the MVTA. An overlay feature
was used in Floorplanner to maintain an accurate scale from historical maps provided by the
maintenance staff at the Majestic Yosemite Hotel (Figure 13). During the course of the thesis,
construction on a set of stairs on the Northeastern side of the bar was completed, and the
floorplan needed to be updated to include the area highlighted in Figure 13.

42

Figure 13 Update to back side of the bar area in 2017 (Ahwahnee Rehab Plan Component
2014)

The tour of the Majestic Yosemite Hotel takes place on the first floor, in addition to the
hotel’s immediate grounds because these locations are where the relics and artifacts are kept. It
was important to have the “Architecture” POI on the hotel’s back lawn because it was originally
going to be the front of the hotel (Sargent 1990). Floors two, three, four, five, and six have only
43

guest rooms and as previously mentioned were not a focus of this project. The methodology
supported the capability of producing an indoor model for the entire hotel but was not necessary
because of the application’s tour objective for the first floor only.
3.2.1. Anyplace
Anyplace Architect was the first tool tested to develop the Majestic Yosemite Hotel’s
indoor model including the location of wireless routers and location of POIs. Anyplace Architect
uses a Gmail associated email account to store building information data on the Anyplace Cloud
Storage. This allows the developer to access, modify, edit, or delete building data created in
Anyplace Architect. The user can also set buildings to private mode so only specific individuals
can have access to its contents, floorplans, and POIs.
Developers search for a building in the search bar where they want to “add” a building.
When an app developer “adds” a building, an icon appears where the user inputs specific
information such as the building’s name, and the user marks on the map the building’s position
on Earth. Once the building is added to the map, building information such as floorplans can be
uploaded to the Anyplace Architect server (Figure 14). Figure 14 shows the floorplan as an
image accurately georeferenced and scaled to the map interface integrated into the Anyplace
Architect API.
44

Figure 14 Anyplace Architect screenshot adding a building to Google Maps
3.2.1.1. POIs Toolbox
The POI toolbox in Anyplace Architect consisted of adding a POI, adding a new
connector for navigation purposes, and toggling the edge mode. The “Add POI” tool allowed the
user to click and drag a POI icon to a specific location within the hotel. The “Add Connector”
tool was used for adding connection points for POT. The toggling edge tool created links
between connection points and POIs.
The add POI feature in Anyplace Architect allowed the user to define the location of each
POI within the hotel, the name of the object, and a small description. Each POI was selected
based on its historical or cultural significance, if it was protected in a display case, or if it offered
a good view of a particular design feature. The majority of the POIs were rugs, hand-woven
baskets, murals, stained-glass windows, desks, old winter apparel, and notable architectural
designs of the hotel. The type of POI was also noted in the description.
45

Anyplace Architect has many different types of POIs that can be added to the map,
depending on the type of building and its contents. Predetermined types of POIs provided by
Anyplace were categorized into Disabled Toilets, Elevator, Entrance, First Aid, Office, Ramp,
Room, Stair, Telephone, and Toilets. Each POI added for the MVTA based on the educational
significance was categorized into a POI type, and were confirmed by ground truthing.
The Add Connector tool in the POI toolbox was used for creating connection points of
travel. These connection points were strategically placed to avoid furniture, pianos, and other
obstacles within the hotel. Occasionally, the furniture is moved for events such as the Chef’s
holiday and weddings, so navigation was ultimately excluded in MVTA. The add connector tool
was complemented with the toggle edge mode tool.
The toggle edge mode provided the direct route between the connection points and the
POIs in the hotel. When a user selects a POI in the mobile application, the route is automatically
generated using the created toggle edge lines. The toggle edge mode was also used to connect to
different floors within the hotel through stair cases or elevators.
3.2.1.2. Anyplace Logger
The Anyplace mobile application that is available on Apple’s App store and Google Play
store has Anyplace Logger built in. Anyplace Logger was downloaded and installed on a Nexus
5 Tablet to record the positioning and readings around each room on the hotel’s grounds. The
readings were based off a “heading” signal, similar to an x, y coordinate system, and represented
a calculated location in the hotel. The boundary of each room was “traced” using Anyplace
Logger, and each reading was collected and stored on the Anyplace Cloud server. Once all the
rooms were recorded, a database was created to store the locations of the wireless routers.
46

The database was stored on the Anyplace Cloud server for the locations of POIs, routers,
and navigation routes. This, coupled with Couchbase SQL, allowed for fast POI information
retrieval. The Anyplace Cloud service stores much of the backend information for retrieval. At
this point in time during testing, only nine POIs were approved for their educational content and
were included during testing phase.
3.2.2. IndoorAtlas and BlueCats Beacon Services
IndoorAtlas was the second IPS tested in this thesis project. During this testing phase, a
large-scale shift in programming from Android to iOS development occurred due to client
requirements based on device reliability. The programming shift used Swift 3.0 and XCode 8 to
program the application from scratch for IndoorAtlas. Beacon integration was also tested during
this phase of the project. A beacon was purchased for the original nine POIs to calibrate and help
correct positioning in the IndoorAtlas IPS. Additional beacons were purchased for the thirty
POIs in the final version of the MVTA as the educational content was approved. The beacon
integration capabilities were heavily favored by Yosemite Hospitality for improved accuracy.
3.2.2.1. Creating the Mapping Area on IndoorAtlas API
The first step for using IndoorAtlas was to create an account on their website. Using their
web API, a location was created for the Majestic Yosemite Hotel in Yosemite National Park
(Figure 15). In Figure 15, the building is shown as a green polygon and marked with a classic
Google marker. A floorplan was uploaded and scaled to overlay on to the Google Maps tool on
the web API. Once the map was uploaded to the API, the mobile map tool was used to record the
geomagnetic signals used for determining on-site location.

47

Figure 15 Example of creating a location using the IndoorAtlas web API
3.2.2.2. Recording Geomagnetic Signals
For best practices, the “Mapping Quick Start Guide with MapCreator2” was consulted to
map the first floor of the Majestic Yosemite Hotel. Once the MapCreator2 Application was
installed on an iPhone 6, the device was calibrated by moving the device in all directions and
orientations to set the magnetometer, accelerometer, and gyroscope sensors in the device. It was
noted in the Quick Start Guide that some Samsung Devices have sensor filtering and may impact
48

mapping, so the original Samsung tablet purchased for testing phases in the project was not used
(IndoorAtlas 2017b).
The first floor of the hotel was mapped over the course of several days between 11 P.M.
and 2 A.M. to avoid disturbing the guests inside the hotel. Once the first-floor map was
complete, the hotel’s immediate grounds were recorded, focusing mainly on the hotel’s back
lawn and near the entrance to the hotel. Figure 16 shows the two different types of paths that
were recorded during the mapping phase, the test paths and the mapped paths. The test paths
provide a summary analysis in the map’s quality report (Peltola 2015). The mapped paths feature
is the final output for the geomagnetic recording that is used to generate locations in the app.
49

Figure 16 Zoomed in location of mapped paths (dark blue lines) and testing paths (light blue) in
IndoorAtlas MapCreator

50

3.2.2.3. Generating and Saving the Map
The map was saved when all the recordings were completed and uploaded to the cloud
server. Immediate indoor positioning testing began using the IndoorAtlas Service. The map data
was managed on the IndoorAtlas web API once the map was generated on the MapCreator
mobile application.
3.2.2.4. XCode and MVTA Structure
The structure of iOS applications relies on the Model-View-Controller. The model
represents what the application is, the controller is how the model is presented, and the view is
what sends the action when things happen in the UI (Hegarty 2017). In XCode 8, the design for
the application can be either done in the Main.Storyboard (Figure 17) or programmatically in
each individual class or Cocoatouch file that makes up the app.

51

Figure 17 Portion of the Main.Storyboard for the IndoorAtlas version of MVTA
52

The screenshot (Figure 17) shows several of the View Controllers used in this testing
phase for the MVTA. When the user selects a button, the segue transitions to another View
Controller based on the user’s input. The bottom left view controller is a table view controller
which houses the list of POIs that the user can access the information without having to move
around the hotel. This was a requirement of the project to create a backup methodology should
the positioning service no longer work, and to provide accessibility to users who may not be able
to walk or navigate through the hotel (National Park Service 2017b).
The top right ViewController shows two separate buttons, one of which that segues into
the bottom right View Controller. This view controller is called the POIViewController and was
the most important one in the project because depending on the user’s location; an array was
used to identify which zone, and later which beacon was closest in proximity to the user (Figure
18). The “struct” model was created in Swift and was used to rotate through the text, images, and
descriptions used in each POI based on where they were. All thirty POIs were included in the
struct and were assigned a reference id in the form of a beacon String value.

53

Figure 18 POIModel used in IndoorAtlas testing phase
Additional beacons were purchased for each individual POI for this project to deliver
content based on proximity to each POI. These beacons were provided by BlueCats and were
mentioned in Chapter 2. An update to the BlueCats web API required remapping beacon
locations if an object needed to be moved (Figure 19). The peach colored dots shown in Figure
19 represent the POI locations, and the white dots represent the locations of the beacons. Once
all the beacons were purchased, a beacon table was created using Serial Numbers and POI
location. These were used to monitor and manage each individual beacon, and make sure they
were placed in the correct location.

54

Figure 19 BlueCats web API used for mapping individual beacons and scaling
3.2.2.5. Adding the IndoorAtlas and BlueCats SDK to an XCode Project
Cocoapods was used to install both the IndoorAtlas SDK and the BlueCats SDK
(Cocoapods n.d.). This was the preferred method to install the SDKs because of its user-friendly
update methods and the ability to control the project files from a terminal program. Cocoapods
was installed using iTerm, and the directory was changed to the local file of the MVTA
application. A podfile was created and both SDKs were installed (Figure 20).
55

Figure 20 Cocoapods installation of BlueCats SDK
The podfile created was directed to the source files for both the SDKs for installation on
their respective Github pages. This established the connection required to access the IndoorAtlas
web API and the BlueCats web API services in the MVTATests application. In June 2017, an
additional Geofence Beta API was released to trigger callbacks based on entering or exiting the
geofenced borders (IndoorAtlas 2017c). The SDK could also be installed to the MVTATests
project workspace using the Cocoapods dependency manager.
The podfile created an .xcworkspace file and was used from that moment on instead of
the original .xcodeproj created (Figure 21). Since the application testing was done using Swift, a
bridging header between the BlueCats Objective C files and the Swift files needed to be created
56

in order to access the BlueCats beacons. To test the interaction between the application and the
beacons, a print statement was used for beacons that were in the range of the iPhone 6 device
used during testing. Successful interaction confirmed that the BlueCats beacons were working in
the application, then the struct file used was set up in the MapViewController.swift file.

Figure 21 Initial BlueCats testing and application structure
A separate test was conducted using sample code for the IndoorAtlas SDK. Both the
Quick Start Guide and iOS SDK guide were used to add the indoor positioning services to the
MapViewController in MVTATests. The spatial components were then accessible in the
application, and the callbacks needed to be managed using the POIModel for the
TableViewController.Swift and MapViewController.Swift files. This methodology was the most
spatially relevant used in the project. However key issues were encountered in testing with Swift
programming, described in Chapter 5.1.2 and 5.1.3.
57

3.2.2.6. Programming and Testing IPS Services and BlueCats in XCode
The POIViewController needed to have a high level of customization and reliability to be
capable of managing all of the hotel model data. Each POI had a different set of texts, number of
images, and a video. In addition, the views for each POI needed to be adjusted individually
based. The POIViewController was set up to handle either one or two paragraphs and one
photograph (Figure 22). Other customizations were included in separate view controllers for
longer POIs with more content and more photographs. Storing POI content in separate
ViewControllers made the programming interface confusing, and risked a higher chance of user
device memory failure. An array in a single ViewController.Swift file was the preferred solution
for this project, set up to rotate based on the location ID or AudioChapter ID. The high level of
customization for the ViewController.Swift file required this to be done programmatically.

58

Figure 22 A sample view for a National Historic Landmark POI
In initial testing, a separate ViewController was created for the Stephen Mather and
History of Yosemite POI which is the longest POI in the MVTA (Figure 23). This tested the
scroll feature and an alternative approach to triggering the beacon interaction. The high level of
customization for the scroll menus built in the Main.Storyboard file met the requirements of the
project, but issues for scaling across all the devices were encountered which was one of the
client’s requirements.
Each user interface (UI) element needed to be “pinned” and anchored for scaling on
alternative devices (i.e., iPhone or other iPads). This divided the screen into different sections
based on points. Since different devices have different resolutions, points are used to set
distances based on the percent of space from surrounding elements. For example, in Figure 23,
59

the content box was set 20 points from the “History of Yosemite and Stephen Mather” title. The
results for this section are described in Chapter 5.1.3.

Figure 23 A sample POI View Controller about the History of Yosemite and Stephen Mather
developed in Xcode

3.3. Developing with Unity
After the testing of the applications described above was completed, it was determined by
the client there would need to be an iOS version of the application because of the reliability of
Apple products and the device’s capabilities for camera initialization using Vuforia Object
Recognition (Vuforia AR Discussion, February 7
th
, 2017, Forum post communication). The iOS
devices currently available offer fewer differences in resolutions compared to Android devices,
60

which is anticipated to reduce the number of updates to the application required over the long-
term. Thus, an iPad pro was selected as the base screen resolution that would be used for
developing the application and the ideal device for guests to check out in the hotel to go on the
tour.
Since many of the POIs were close to one another, an attempt was made to use both the
beacons and object recognition/AR services was made. The idea behind this methodology was to
use the user’s position in reference to the object to deliver educational content based on what the
camera was seeing. This was made available using both Unity 5.5.2f1 and Vuforia.
3.3.1. Introduction to Unity Development
Unity is available on both iOS and Android, Mac and PC. Since the target device was an
iPad pro with a resolution of 2056x1536, this was set as the default screen size for the Unity
project, as a game scene. Unity is highly flexible in its design elements and functionality and
would meet the user design requirements of the client. Overall maintenance for the MVTA in
Unity was also preferred based on Unity’s functionality and components. Figure 24 shows a
sample introduction menu of the MVTA developed as part of this thesis project in the Unity
Game Scene.
61

Figure 24 Unity MVTA Game Scene on iPad resolution
Once the target resolution was set, the camera in Unity was set to display what was on the
canvas, or within the frame of the application. Since this application is rendered in 2D based on
specifications of the client, the 3D components that Unity provide were not required. The first
step was to design each of the POI panels, as in XCode. Each POI was required by the client to
have at least one image of an object in the hotel and one text description, along with an audio file
that would satisfy the requirements of the American Disability Act (ADA) and National Park
Service accessibility guidelines (National Park Service 2017b).
62

3.3.2. Unity and Vuforia Object Recognition Service
The Vuforia asset package was downloaded from their developer website and was used to
test out the object recognition features in the MVTA. The scripts used for object recognition
features were written in C# and were modified for use in the MVTA based on target images.
Once the Vuforia assets were downloaded and added to the Unity project, a Vuforia web API
key was created to initialize the camera services in Unity. Object target photos were taken of
each individual POI to be included in the object recognition database created on Vuforia’s size
and in gray scale per Vuforia’s database requirements. Once all of the POIs had a target image,
the target image database was uploaded into the Unity application.
Next, each POI was assigned an id number for reference. In case the client required more
POIs, sixty-five additional potential POIs were documented and processed for object recognition
services. With the timeframe of this project, the final amount was reduced to thirty. Each target
image was rated on a 0 – 5-star scale based on “augmentability.” This calculation is based on
Vuforia’s algorithm where geometric points are defined within each photo (Figure 25). The more
geometric points used, the higher the augmentability rating and the greater chance for success in
object recognition.
63

Figure 25 Women's Restroom sign augmentability sample points and geometry
Multiple image databases were included in this project for different times of day and
from different angles to reduce user error for object recognition. Ninety-three target images were
used in the Vuforia and Unity application for the thirty final POIs. The images ranged on
augmentability, but the “0”, “1”, and “2” star-rated photographs were removed because of lack
of a reliable detection. The higher augmentability-rated target images for specific POIs detected
objects 100% of the time and the “backup” photographs were removed to reduce application
space.
3.3.3. Object Recognition Scripts
Each individual POI had its own scripts built into the application based on image target
recognition. Each script would SetActive each appropriate GameObject based on what the
camera device was looking at. The GameObject was the main component used in the project.
64

Each individual POI had its own GameObject appropriately named based on the POI and would
activate whenever triggered or identified in the application. For consistency, each GameObject
consisted of all the photos, videos, text, titles, and buttons in the application. An advantage to
using the camera was when the RestorationStencilWork GameObject was identified, information
about the object would display on the screen. The GameObject would be loaded during the
application startup but would start in a “false” or inactive state (Figure 26).

Figure 26 Restoration Stencil Work POI GameObject loading state
In the OnTrackableFound method, this is where the GameObject would be set to true and
would display each GameObject in the application. In the OnTrackableLost method, the
GameObject would disappear when the camera no longer recognized the object and the device
had moved away. Setting the GameObject to “false” when the object was lost created a problem
of the GameObject from deactivating and reactivating numerous times. This was later adjusted
by creating a close button within each GameObject to set the GameObject back to false.
3.3.4. Testing Unity and Vuforia
The initial object recognition tests and layouts were designed to wrap around each
individual object present to orient itself in virtual space (Figure 27). The example shown in
65

Figure 27 shows the text floating in mid-air, but the content is not easily read on the device.
Depending on the background, the text would augment itself around the object. The first test for
augmented services was successful. The test produced an application that could recognize
objects in the real-world environment, and augment the content based on where the user’s device
was oriented. It also challenged the current interface design and its ability to clearly provide the
user’s information.

Figure 27 Example of an initial Vuforia object recognition tests using Unity
66

3.3.5. Unity and Bluetooth
Assets provide code packages to help developers run certain tasks or provide assistance
in application development for Unity projects. There is an Asset Store where developers can read
and test out different types of assets based on their project’s needs. For this project, a Bluetooth
asset would be required in order to communicate between the beacons provided by BlueCats and
the Unity project. The following discussion highlights two separate approaches that were
considered.
3.3.5.1. Unity and Bluetooth Assets
There are several scripts to establish a connection from the Unity gaming engine to
Bluetooth technologies in Unity’s Asset Store. These scripts, in the form of assets, were written
by Unity developers, and provide their services for sale on the market that other developers can
use. Most of the Bluetooth Assets on the Unity Store at the time of this writing were Android
only. There was one Bluetooth Asset that was taken into consideration, but because of its high
cost and lack of reviews, it was ultimately not used.
3.3.5.2. Unity as an XCode Project
The final attempt to incorporate Bluetooth technology into the Unity application occurred
when exporting the Unity application as an XCode Project. The idea behind this tested the ability
to manipulate the Unity project to incorporate the BlueCats SDK, but in fear of going against the
terms of service for the Unity project and failure to manipulate game objects based on Bluetooth
proximity, it did not work out.
Manipulating the Unity XCode project did not have much documentation on forums or
YouTube. Much of the code was converted from C# into Objective C using Unity’s scripts, and
67

the structure of the Unity application in XCode did not allow for easy manipulation of sample
code within the given time frame. The experience in Objective C projects was also limiting.
3.3.6. Educational Content
The educational content in the application was developed specifically for this project.
The information was co-authored by Yosemite Hospitality’s Interpretive Services Department,
and final revisions were completed in August 2017 after approval by the National Park Service.
All of the POIs were determined during the first few walkthroughs of the hotel based on
questions asked by park visitors and common information already discussed during the Historic
Hotel tour. The educational content was designed to go into further detail about each POI and
cover more about the rooms that aren’t frequented during the Historic Hotel tour. The rooms in
which the content went into greater detail included the Solarium and Winter Club Room as these
rooms aren’t frequented as much during the historic hotel tour. Each POI was designed to have
varying amounts of information based on the subject matter and was broken up accordingly.
Each set of text was recorded by Cory Goehring of Yosemite Hospitality’s Interpretive
Services and was edited to fit into the application to meet ADA standards. Each audio recording
was recording using a Yeti Stereo Microphone and GarageBand. The audio recordings were
exported as a low-quality MP3 to save space in the final app. An Audio Tour menu was also
included where the user could click one button and learn about all thirty POIs to meet ADA
requirements. This did not require the use of navigating between different menus in case of
vision impairments. Table 4 lists all of the Audio Chapter Titles and POI topics in the MVTA.
Most of the information used in the application was based on historical documents, books
by Shirley Sargent, Keith Walklet, Yosemite Nature Notes, and historical newspaper clippings
from the Yosemite National Park Research Library (National Archives 2017; National Park
68

Service 2017a, 2017b, 2017c; Lund 2014; National Park Service 2012; National Park Service
2009a, 2009b; Walklet 2004; Sargent 1990; National Park Service 1986; Century Magazine
1890; Sierra Club n.d.a, n.d.b, n.d.c). All photographs used in the application required Copyright
permissions and one-time use agreements for Commercial Use should the client end up selling
the application in the future. Copyright was provided by the Huntington Library, Alan Petersen
and the Gunnar Widforss Raisonne’ Project, the John Muir Trust, Holly Cannan, Ansel Adams,
Dakota Snider, and Breanne McNitt.
3.3.7. Designing the User Interface
Four menus were proposed and tested in the MVTA. The first and introductory menu is a
welcome screen where it explains how the application works to the user. Once the user closes the
introduction menu, the POI Scroll Menu is displayed, and the user can navigate between the Map
feature or select a POI in which they want to learn about. The final menu proposed consisted of
an Audio Only menu with a play and pause button.
Each user interface object such as buttons, photographs, or text needed to be able to scale
across different screen resolutions. Unity allows anchor points to be set on each object for
scaling across all resolutions and devices. To avoid distortion in portrait mode, a landscape only
orientation was tested. The font proposed in the MVTA was Century Gothic for ADA standards,
and the size minimum was equivalent to a size 16. The final User Interface is shown in Chapter
5.2.

69

Chapter 4 Model Development
A historical model of the south wing of the hotel (Figure 28) was created as part of this project
for the purposes of preserving the architectural components, and with the intent to preserve
conditions of the historic furniture within the south wing of the hotel. The model is comprised of
two separate components, a digital and physical model. Each room was digitally scanned using
the Canvas.io application and Structure Sensor device previously mentioned. The scans were
converted using the “Scan to CAD” feature in the Structure application and uploaded to a
SketchUp file (SketchUp n.d.). A SketchUp Pro license was purchased in case the client wanted
to use it for commercial purposes.

Figure 28 Combined model view from north side of the hotel’s south wing
The furniture scans used the same Structure Sensor iPad connection with Skanect.
Skanect is a scanning environment that uses a wireless network to transmit the Structure sensor
70

signals and creates a virtual scan of an object, room, or space (Skanect n.d.). Each piece of
furniture was scanned and documented based on its location and assigned an ID number. Figure
29 shows one of the furniture pieces in the Winter Club Room being scanned. This chapter
highlights the processes to create both the digital and physical model of the Majestic Yosemite
Hotel’s south wing.

Figure 29 Scanned front side of the couch and adjacent wall in Winter Club Room

4.1. Scanning the Hotel’s South Wing
Each room was scanned individually up until the 2GB ram capacity on the iPad Pro 9.7-
inch model. The optimal scanning environment required a well-lit area and a lack of moving
objects. Scans were completed between 10 P.M. and 2 A.M. in the hotel when many of the
guests were asleep to avoid conflicts. The goal was to scan the whole first floor of the hotel, but
due to limitations experienced while scanning, specifically the time frame and the range of the
scanner, this goal was ultimately not met. Several of the rooms on the first floor have very high
71

ceilings, and the range of the Structure Scanner could not reach the tops of the Solarium, Great
Lounge, and the Dining Room. A number of rooms on the first floor are not publicly accessible
and were not scanned because of privacy concerns. In addition, most rooms required two or more
scans, and each scan needed to be stitched together using the Scan to CAD feature. A scanned
portion of the Winter Club Room can be seen in Figure 30.

Figure 30 Scanned portion of Winter Club Room
Both models focus on the south wing of the Majestic Yosemite Hotel, which includes the
Lounge, the Solarium, the Winter Club Room, and the Mural Room (Figure 31). The peach
colored dots in Figure 31 represent the POIs in the hotel’s south wing. Additional measurements
for each of these rooms were taken using a Bosch GLM 35 laser measurement tool and a
measuring tape. The measurements were taken to compare the accuracy of the Structure Sensor
72

scans to more classical measuring and surveying techniques. After the scans were complete, they
were uploaded on the Structure App and processed into a CAD file.

Figure 31 Majestic Yosemite Hotel South Wing
4.2. Scanning the Hotel’s Furniture
The hotel’s original furniture in the Lounge, Solarium, Winter Club Room, and Mural
Room were all scanned as part of this project. Due to time constraints, one piece of each
furniture was scanned and labeled. Types of furniture included tables, chairs, the Solarium
fountain, and desks. The American Indian baskets on the fireplace’s mantle were not scanned to
avoid possible damage from IR sensors. The lights in the hotel were not scanned because of their
narrow structure and the inability to get a complete scan from all angles, from above each
lighting fixture.
Each piece was scanned using Skanect. Skanect can use either the Structure Sensor or the
Kinect by Microsoft as its input source and transmits data streamed from the iPad sensor to the
73

computer. The scan on the iPad is also displayed on the computer in Scanect after the scanning
environment is set up on the computer. If there was not enough “geometry” or reference objects
detected in the sensor’s view, the scan would need to be reset to the last known location.
Raw data recorded in Scanect for one of the Mural Room tables is shown in Figure 32.
The width of the scanning environment was recorded at 6 feet. Each scan took several attempts
to make sure each angle and side were completed for each piece. This scan shows the initial data
from one angle previously scanned. The scans were then processed in Scanect using the crop and
fill holes tool to complete each object.

Figure 32 Raw scanned data in Skanect of a Mural Room table (middle)

4.3. Building the Physical Model
The physical version of the CAD model was built using 1/8
th
-inch birch wood and a laser
cutter. Every panel in the CAD model was exported as a PDF and scaled down using calculations
derived from the SketchUp model. Every inch in the model represents 3 feet in the hotel (Figure
74

33). Initial testing of the laser engraver used a higher setting for each individual face to confirm
that it would cut through the 1/8
th
-inch piece of wood. There were forty sheets of 8in x 10in birch
wood used in the making of the physical model. Several panels such as the model’s doors
required a “cut” file and a “draw” file. The cut file used the higher laser setting to remove excess
wood, and the draw file used to burn a thinner line into the wood for accent. Once each face was
cut out, the model was assembled on top of a 48in x 36in wooden base, glued, and laser aligned.
Figure 33 shows the assembled model.

Figure 33 Birchwood model of south wing in Majestic Yosemite Hotel
The next chapter details the results of the MVTA and the physical and digital south wing
model.
75

Chapter 5 Results
Although several of the methodologies tested in the development of the application architecture
were rejected upon consideration by the client, these tests contributed to the overall design of the
final project. The first section of this chapter highlights the decisions based on testing results and
working with the client to determine the final product. This chapter also highlights the final
version of the application submitted to and accepted by the client and the current version of the
indoor model at the time of this writing. The final section of this chapter discusses the lessons
learned while working on this project.
5.1. Alternative Application Methodologies and Results
This thesis documented several approaches for developing the MVTA, including use of
Anyplace, IndoorAtlas, BlueCats, Vuforia, and Unity. This section highlights the results of each
application methodology.
5.1.1. Anyplace
Anyplace on the Android platform was the first IPS tested in this project because of its
free and open-source code. It uses a wireless network and RSS values to determine the location
within a given space. The Anyplace Architect was used to overlay the Majestic Yosemite Hotel’s
floorplan in Google Maps, and to attempt to create the initial location services. Since the guest
wireless network was ultimately deemed not usable for this project and because there is no other
accessible wireless infrastructure in the hotel, the Anyplace IPS was eliminated from
consideration early on. Additional location display errors and critical errors when incorporating
Anyplace into a native application would also require heavy debugging and complex application
76

updates as needed, both of which would not meet Yosemite Hospitality’s goal in maintaining the
application effectively in-house. In normal circumstances with readily available wireless
systems, this IPS solution may be more applicable, but further development and improvements
were required.
5.1.2. IndoorAtlas
The results from IndoorAtlas produced LBS inside the hotel in a native application on
iOS with just a few lines of code. The cloud-based features relied heavily on a data source to
initialize the location and to keep an accurate location. Testing was completed using a Verizon
Wireless cellular mobile device, but reception could not be guaranteed across all devices because
of the lack of cellular support infrastructure. During testing on the mobile device, location
display errors occurred due to the lack of a dedicated service. The resulting errors, as well as the
service’s costs for the large venue, ultimately did not meet the goals of the client. This service
may be considered in future versions of the application once more infrastructure is readily
available and cellular service is more reliable in Yosemite National Park.
5.1.3. BlueCats
BlueCats beacon services was the third methodology tested during this project, also
tested in combination with IndoorAtlas. Using the BlueCats SDK allowed for easy integration
into the MVTA. Many of the costs were up-front on the technology side, and the BlueCats SDK
long-term service costs would have been based on the client’s use. Programming using the SDK
allowed information to be displayed based on the categories set for each beacon or based on their
proximity. These were the two methodologies tested in the MVTA. Issues programming the
MVTA’s interface challenged the backend structure of the application. The POIViewController
required varying amounts of information such as multiple text boxes and photographs and
77

needed to be extremely adaptive in the live version of the application. Issues with the
POIViewController excluded this methodology from consideration, and ultimately developing a
native application in either XCode or Android Studio within the given time frame permitted for
project completion.
5.1.4. Vuforia
Vuforia was tested using Unity, and although this component was not ultimately chosen,
rejection was due to the particular conditions encountered within the physical space of the hotel
and overloading the device’s processing power. Each target POI generally needed to have a well-
lit area that could be used for object recognition. Many of the POIs considered did provide well-
lit areas, but at different times of day, the object recognition services would not work for certain
POIs due to the amounts of light inside the hotel. In the evening, the chandeliers in the hotel do
not provide adequate lighting for the POIs and led to object recognition issues. Even different
weather patterns during the winter such as outside snow produced unreliable user experiences.
During Vuforia testing, 27 out of 30 POIs tested successfully. Testing with Interpretive
Services staff, the department of Yosemite Hospitality that will maintain the MVTA, produced
22 out of 30 POIs tested successfully. Unsuccessful tests resulted because it wasn’t clear how far
away the user needed to stand from each POI in order for object recognition to be successful. As
a workaround for this issue, additional target images were collected for those POIs but ended up
requiring so much memory in the application that the MVTA crashed often and the power usage
drained the battery in 45 minutes of use. Appendix D contains the POI image targets collected
for the MVTA. These difficulties as well as the additional cost for the Vuforia software
ultimately influenced the final decision of choosing Unity by itself.
78

5.2. The MVTA and Associated Project Requirements
This thesis outlined several methodologies that were tested in the context of this project
for the development of the MVTA, both with successes and failures incorporating spatial
technologies and object recognition technologies. IndoorAtlas and BlueCats produced favorable
results for this project, but lack of programming experience using XCode and the Model-View-
Controller scheme, as well as long term in-house maintenance considerations and associated
costs eliminated these options. In a location with more Wi-Fi and cellular infrastructure, these
systems may be more reliable and preferred. One lesson learned in regard to infrastructure was
not to assume that these infrastructure resources are readily available in every context. Costs,
long-term and short-term maintenance, and programming experience of in-house staff led to the
final decision for using Unity.
Unity was the preferred architecture for building the application in the end because of the
problems that occurred with the other technologies during testing phases. One of the biggest
issues faced was during the upgrade to XCode 8 and iOS 10. Many of the iOS programming
classes that were used became deprecated and needed to be modified for improved functionality.
Updates to the SDKs were not readily available after the noted iOS update and provided a
valuable lesson in long-term application maintenance. Relying on multiple software and web
service companies for different services and updates requires time, and towards the end of this
project, it was deemed most critical to the success of the project to complete the application in a
timely manner.
Maintenance costs on the client side were heavily considered in the final MVTA
implementation. In order to avoid having to hire a full-time developer, an application built from
scratch was not preferred. A Unity Pro license was purchased for this project in exchange for the
79

vendor’s services and support. Enhanced capabilities in Unity also provided the necessary tools
for the Interpretive Services to maintain the application over the long-term.
Unity was proposed because it did not require a full-time developer to maintain the
application since many of the components could be edited directly within the interface. Since
part of the future work involves training Interpretive Services staff on maintenance, limiting
development to only Objective C and C# in Unity was preferred, as opposed to implementing all
four programming languages used and tested during development. Unity also allows each project
to be exported to work on a variety of platforms including Mac, PC, Linux, iOS, and Android.
These different platforms give the client flexibility as to which devices can be used with the
application.
In addition, the ability to modify and create the user interface to the specifications of the
client was ideal in Unity. The Unity structure allowed for a high level of customization without
relying on any outside services. Many of the interface elements can be built directly into the
view. This gives the client the ability to edit or create content directly into the interface, rather
than doing everything programmatically in multiple languages. One additional requirement was
to make the application commercial ready. Each photo used in the application was either in-
house or required a one-time Copyright use form. Unity gives the client the ability to monitor
sales and usage on their website as part of the Unity Pro license.
The Unity Pro license is $1,500 per seat, per year. The costs for hiring a full-time
developer to maintain the application would be significantly higher. The license gives Yosemite
Hospitality the ability to make as many applications as they would like in the future, as well as
maintain the MVTA. One of the original ideas was to have an application “game” in the Mural
Room of the hotel. This could become its own application in the future and may incorporate
80

Vuforia’s AR services that were tested during this project. It is for these reasons that Unity was
the preferred methodology to go with for this project. The goals of reaching the desired user
experience that meets ADA standards, an easy to use interface and map feature, and having an
audio tour were all met on the client side. Although the costs were high in initial startup, the
Unity solution provides services for long-term maintenance and gives the client flexibility for
developing other applications.
Working with all of the architectures listed in this thesis produced unexpected results.
Building a “commercially ready” application requires a more involved approach compared to a
theoretical approach, especially while looking at the legal requirements of many of the terms of
use of the software components involved. In addition, working with staff in a large company on a
day-to-day basis requires a certain level of professionalism. One of the challenges was the
amount of time that this project was to be completed in. Application development is deceptively
long, and there needs to be adequate time allocated for application testing and debugging. The
final product was completed 10 months after starting, almost double the proposed time.
Nevertheless, the end result met the satisfaction and requirements of the client listed above.
As a result, both an Android and iOS version using Unity have been created for the
MVTA and were submitted to the client. The final version of the application was tested on the
iPad Pro 9.7 inch and Samsung Galaxy S2 Tab and is owned by the client Yosemite Hospitality.
The Splash Image when the Application is first opened (Figure 34) contains the Yosemite
Hospitality Logo and the Unity Logo. After the Splash Screen Menu, the Introduction menu is
displayed. A large image of the hotel’s south wing and granite backdrop immediately brings the
user into the setting. Once the user closes out the introduction page, the audio chapter scroll
81

menu is displayed with all 30 POIs. The end user is able to select the POI based on topic or
navigate to the map function in the application.

Figure 34 Splash Screen Image for MVTA

The map function contains buttons to each POI game object similar to the audio menu but
is based on where the user is located. When the user selects the map function, a “You are here”
box pops up at the location of the concierge desk in the Majestic Yosemite Hotel (Figure 35).
The user can then dismiss this box and begin the tour. Each stop is in a publicly accessible area.
The buttons were originally color coded by room and geographic section to suggest a flow for
the application, but ADA standards required a single color since communication cannot only
occur based on color. The user can walk out the doors and learn about National Historic
Landmark POI which is listed as Chapter 1.1 in the MVTA, or the user can select any other
location-based on where they want to go.
82

Figure 35 MVTA map starting point and POI’s are shown as gold boxes
The map contains a scroll feature which allows users to readily access other POIs and
information. The close button remains stationary so the user can exit and segue between different
menus. Once the user selects a POI, they are able to learn about that specific location or object.
An example of a POI is the Women’s Restroom POI, or Chapter 5.1 within the MVTA (Figure
36).
83

Figure 36 A sample POI in the MVTA
The client required at least one photograph for each POI, and an audio version of the
content. Consistent formatting throughout the POIs was also required to keep the MVTA
uniform in structure or user experience. Each title font was Century Gothic 80, and the body
paragraphs were Century Gothic 50. The play and pause buttons were created in Adobe
Illustrator, and the close buttons were created in Unity. Each menu in the MVTA contains an off-
screen scroll-bar to avoid clutter. All images used in the MVTA were saved as a texture in order
to overlay it onto the Unity user interface canvas.
84

There is one video in the application which was also saved as a texture. The video is from
the 1950s, a recording of the historic Firefall event. The video required a separate C# script to be
written to access it from the StreamingAssets in the project’s files loaded onto each device.
The user-friendly navigation and options met the needs of the client. Although it did not
incorporate the place-based interaction originally desired using an IPS, beacons, or object
recognition, it does provide a consistent method for content delivery and a professional-grade
application without draining the device’s battery. The final version of the application also met
the client’s needs of providing content based on where the user is inside the hotel, while also
providing access to each POI without the use of LBS for ADA standards.
5.2.1. Testing and Debugging
Initial tests involved checking memory and energy consumption on the iPad Pro device.
Memory tests using XCode’s debugging feature produced energy and memory consumption data.
Based on those results, modifications were made to the Unity project. The amount of data in the
application exceeded 500mb and consumed a high level of the device’s energy. Each file
included in the Unity project was compressed to reduce energy power and memory consumption.
This lowered the initial MVTA loading time from thirty seconds to twenty-two seconds.
Interface elements during initial testing failed to produce intended results. Several buttons
were not working because of missing scripts and functions. These were easily noted and fixed.
Testing on an iPhone 7 resulted in a much smaller font compared to the iPad Pro. The “best fit”
option for each text string was used and would scale the font based on screen size. Once all the
interface elements were fixed, the application was ready for submission to the National Park
Service.
85

The National Park Service and Yosemite Hospitality’s Interpretive Services Department
required staff testing of the final version of the MVTA before public release. The content was
edited based on accuracy and clarity, and the audio components were analyzed for quality
standards. After the final round of testing and revisions requested by both entities, the application
was ready to submit to Yosemite Hospitality.
5.3. The Physical Model
This project’s physical model will be on display in the Majestic Yosemite Hotel later this
year. At the time of this writing, some of the paint needs to be redone, and the 3D printed
furniture needs to be added. Initial furniture printing resulted in a morphed and inaccurate scaled
chair which was eventually discarded. The physical model in its current form is shown in Figure
37. The purpose of the model was to display some of the architectural components and updates
in a smaller form. The added furniture will provide the historical context originally envisaged for
the project.

86

Figure 37 Majestic Yosemite Hotel Model

87

Chapter 6 Conclusions
The purpose of this project was to meet the desired user experience for the client. The desired
user experience was defined as creating an overall engaging mobile tour application through
relevant information about POIs within the hotel with the ability to select or receive information
based on where the MVTA user is inside the hotel, an audio component where the user can
choose to listen to the content or access a complete audio tour, a clean user interface layout and
design, and incorporating historical and present-day photographs to establish relevancy to the
educational content for each POI. All MVTA application design decisions were based on client
application specifications regarding functionality. The main component of this project was the
mobile application and much of the thesis project time was focused on its implementation. The
end result of this thesis produced a “living” virtual tour application that can easily be modified
and updated in the coming years by the client. The ability to reach more park visitors and
increase the visitor experience supported the client’s desire for building such an application.
The testing of different kinds of positioning technologies such as several Indoor
Positioning Services, Bluetooth beacon interaction, and AR services presented many challenges
along the way. From programming a native Android application to making a switch to iOS for
development, new technologies were experimented with and tested to try and meet the desired
user experience. The limiting cellular infrastructure and wireless access in Yosemite National
Park ultimately eliminated the applicability of some of the methodologies, whereas required
programming experience prevented success with others.
There are however, many positive outcomes that came from this project. In terms of
professional growth, the author was fortunate to work with many excellent colleagues across
88

disciplines in getting this application to the best state that it could be in the time frame required
by the client. The author was part of a team to help write interpretive text, worked in a
professional environment and in meetings with executives and managers, and coordinated with
hotel staff on project updates and methodologies. Working on both the Android and iOS
platforms, learning how to write applications in Swift and scripts in C#, working with
dependency managers, working with outside developers, students, professional networking, and
building an application to be proud of were all positives experiences and thus invaluable
outcomes from this project.
6.1. Future Work
The MVTA is a living document, as many more POIs will be added in the future. Thirty
were chosen for this project based on qualities previously noted. The Interpretive Services staff
at Yosemite Hospitality will continue to improve and update the application for both iOS and
Android devices. Future versions of the MVTA will include audio and text in multiple human
languages to provide educational services to more international visitors who come to Yosemite.
This was one of the main appeals for Yosemite Hospitality, and the type of system that is in
place for the application’s architecture will allow for this to happen.
With regards to the physical model, this will also continue to be a living document. The
goal is to provide visual, tactile information about each piece of historic furniture, and further
research will be required about each piece within the model. This is just the start of preserving
the cultural and historical resources inside the hotel and will provide a snapshot into the hotel’s
interior design elements for future generations.
89

6.2. Technology Transfer
The final MTVA files will be transferred to Yosemite Hospitality’s Interpretive Services
Department in Yosemite National Park for Android and iOS release. Interpretive Services staff
will be trained on how to update, debug, and maintain the application using Unity, XCode, and
Android Studio. Training will involve designing the application’s architecture and interface for
consistency across all devices, scaling, photo processing, audio processing, script updates and
resources, Unity resources, and a user guide manual.
The Unity Pro license seat will be transferred to one member of Yosemite Hospitality’s
staff for long-term maintenance of the MVTA. All of the tested applications created as part of
this project will also be included should the client choose to go in another direction in the future.
All model resources created will also be transferred to Yosemite Hospitality Interpretive Services
staff. The technology transfer will be completed in September 2017.

90

References
Abraham, Pierre, Bénédicte Noury-Desvaux, Marie Gernigon, Guillaume Mahé, Thomas
Sauvaget, Georges Leftheriotis, and Alexis Le Faucheur. 2012. The inter- and
intra-unit variability of a low-cost GPS data logger/receiver to study human
outdoor walking in view of health and clinical studies. PloS One 7 (2): e31338.

Anyplace Architect. n.d. API. Accessed January 7
th
, 2017. Available at
https://anyplace.cs.ucy.ac.cy/architect/.

Anyplace Github. 2017a. Anyplace Application. Accessed July 15
th
, 2017. Available at
https://github.com/dmsl/anyplace/.

Anyplace Github 2017b. Documented Issues. Accessed July 15
th
, 2017. Available at
https://github.com/dmsl/anyplace/issues/44.

Applanix. “Pictometry Case Study.” 2016. Accessed August 13
th
, 2017. Available at
https://www.applanix.com/downloads/case_studies/timms/pictometry_casestudy_web.pd
f. Accessed July 15th, 2017.

Applanix. 2015. “Applanix LAX.” Accessed July 15
th
, 2017. Available at
https://www.applanix.com/downloads/case_studies/timms/Applanix_LAX_One_Sheet-
r4web_(1).pdf.

Bahl, Paramvir, and Venkata N. Padmanabhan. 2000. RADAR: An in-building RF-based User
Location and Tracking System. INFOCOM 2000. 19
th
Annual Joint Conference of the
IEEE Computer and Communications Societies. Accessed November 8
th
, 2016. Available
at http://ieeexplore.ieee.org/document/832252/

Blaser, Monica. 2015. “Opportunities of an Interpretive Application for Self-Guided Tourism
within the National Park System. Professional Project: University of Nebraska – Lincoln.

BlueCats. n.d. “BlueCats Developer Portal.” Accessed July 15
th
, 2017. Available at
https://developer.bluecats.com/

Century Magazine. “Features of the Proposed National Park.” 1890. Vol XL., No. 5. Accessed
July 15
th
, 2017. Available at http://vault.sierraclub.org/john_muir_exhibit/writings/
features_of_the_proposed_yosemite_national_park/default.aspx. Accessed July 15th,
2017.
Chen, Yiqiang, Qiang Yang, Jie Yin, and Xiaoyong Chai. 2006. Power-efficient access point
selection for indoor location estimation. IEEE Transactions on Knowledge and Data
Engineering 18 (7): 877-88.

Cocoapods. n.d. Accessed July 15
th
, 2017. Available at https://cocoapods.org/.
91

Couchbase. “NoSQL Engagement Database.” Accessed February 18
th
, 2017. Available at
https://www.couchbase.com/

Dundee, Mick. 2017. “Setting Up Indoor Positioning for Android Application.” Accessed July
15
th
, 2017. Available at http://support.bluecats.com/ customer/portal/questions/16879397-
setting-up-indoor-positioning-for-android-application

Farneth, S.J., D. Wessel, K.T. Petrin, S.E. Watson, and H.Y. Granke. 2011a. “The Ahwahnee:
Historic Furnishings Report 95% Draft – March 2011.” Architectural Resources Group,
Inc.

Farneth, S.J., D. Wessel, K. Petrin, K. Vieth, M. Slater, S.E. Watson, H.Y. Granke, K. Untch, K.
Wong, and M. Lovato. 2011b. “The Ahwahnee: Historic Structures Report – January
2011.” Architectural Resources Group, Inc.

Floorplanner. “Create floorplans the easy way.” Accessed March 20
th
, 2017. Available at
https://floorplanner.com/

Food-Management. “Aramark Launches Mobile App for Campus Dining.” 2012. Accessed July
15
th
, 2017. Available at http://www.food-management.com/contractors/aramark-
launches-mobile-app-campus-dining

Fouskas, Kostas, George Giaglis, Panos Kourouthanassis, Adamantia Pateli, and Argiris
Tsamakos. 2002. "On the potential use of mobile positioning technologies in indoor
environments." BLED 2002 Proceedings (2002): 33.
Georgious, Kyriakos, Timotheos Constambeys, Christos Laoudias, Lambros Petrou, Georgios
Chatzimilioudis, and Demetrios Zeinalipour-Yazti. 2015. Anyplace: A Crowdsourced
Indoor Information Service. IEEE Press (2): 291-294.

Giudice, Nicholas, Lisa Walton, and Michael Worboys. 2010. The informatics of indoor and
outdoor space: A research agenda. Indoor Spatial Awareness Conference 2010. DOI:
10.1145/1865885.1865897.

Goodchild, Michael F. 2011. “Looking Forward: Five Thoughts on the Future of GIS.”
ArcWatch, February. Accessed September 17, 2016. Available at
http://www.esri.com/news/arcwatch/0211/future-of-gis.html

Goodchild, Michael F. n.d. “Location-Based Services.” National Center for Geographic
Information and Analysis, and Department of Geography, University of Southern
California, Santa Barbara. Available at http://www.geog.ucsb.edu/~good/papers/471.pdf

Gruteser, Marco, and Xuan Liu. 2004. “Protecting Privacy in Continuous Location-Tracking
Applications.” IEEE Security and Privacy.

92

iTunes. 2017. “Freedom Trail Walking Tour.” Accessed August 20
th
, 2017. Available at
https://itunes.apple.com/us/app/freedom-trail-walking-tour/id355207897?mt=8)

Jiang, Bin, and Xiaobai Yao. 2006. “Location-based services and GIS in perspective.”
Computers, Environment and Urban Systems 30: 712-25.

Jones, Greg, and Mark Christal. 2002. “The Future of Virtual Museums: On-Line, Immersive,
3D Environments.” Created Realities Group. Accessed November 9
th
, 2016. Available at
http://w.created-realities.com/pdf/Virtual_Museums.pdf
Lund, Morten. 2014. “How the Olympics came to a sleepy Adirondack Village.” Accessed July
15th, 2017. Available at https://www.skiinghistory.org/news/lake-placid-1932.

Geoslam. n.d. “The ZEB-REVO Solution.” Accessed July 15th, 2017. Available at
https://geoslam.com/wpcontent/uploads/2017/08/GeoSLAM-ZEB-REVO-Solution-
v7.pdf

Hegarty, Paul. 2017. “Developing iOS 10 Apps with Swift.” Stanford University CS193p Winter
2017. Accessed July 3
rd
, 2017. Available at https://itunes.apple.com/us/course/developing
-ios-10-apps-with-swift/id1198467120. Accessed July 15th, 2017.

IndoorAtlas. 2017a. “What are Indoor Positioning Systems.” Accessed July 15th, 2017.
Available at http://www.indooratlas.com/

IndoorAtlas. 2017b. “Mapping Quick Start Guide with MapCreator 2.” Accessed July 15th,
2017. Available at http://docs.indooratlas.com/app/

IndoorAtlas. 2017c. “IALocationManager Class Reference.” Accessed July 15th, 2017.
Available at http://docs.indooratlas.com/ios/2.5.0-beta/Classes/IALocationManager

Kollath McCann Creative Services. 2011. “Ahwahnee Hotel Basket Acquisition.”

Liu, Hui, H. Darabi, P. Banerjee, and Jing Liu. 2007. Survey of wireless indoor positioning
techniques and systems. IEEE Transactions on Systems, Man, and Cybernetics, Part C
(Applications and Reviews) 37 (6): 1067-80

McComb, Ashley. 2017. “Historic Majestic Yosemite Hotel Tour.” Majestic Yosemite Hotel,
Yosemite National Park, CA.

MLB. 2010. “MLBAM, Philadelphia Phillies and Aramark join to debut Mobile Food Ordering
App at Citizens Bank Park.” Accessed July 15th, 2017. Available at
http://mlb.mlb.com/news/press_releases/press_release.jsp?ymd=20100923&content_id
=14992180&vkey=pr_mlbcom&fext=.jsp&c_id=mlb
Moberg, Peter, Afif Osseiran, and Per Skillermark. 2009. Cost comparison between SISO and
MIMO deployments in future wide area cellular networks.
93

National Archives. 2017. “Hetch Hetchy Environmental Debates.” Accessed July 15th, 2017.
Available at https://www.archives.gov/legislative/features/hetch-hetchy. Accessed July
15th, 2017.

National Park Service. 2017a. “Yosemite National Park Statistics.” Accessed June 9
th
, 2017.
Available at https://www.nps.gov/yose/learn/management/statistics.htm

National Park Service. 2017b. “Accessibility.” Accessed August 26
th
, 2017. Available at
https://www.nps.gov/hfc/accessibility/

National Park Service. 2017c. “National Historic Landmarks Program: Listing of National
Historic Landmarks by State.” Accessed August 15th, 2017. Available at
https://www.nps.gov/nhl/find/statelists/ca/CA.pdf

National Park Service. 2017d. “Birth of a National Park.” Accessed July 21st, 2017. Available at
https://www.nps.gov/yell/learn/historyculture/yellowstoneestablishment.htm

National Park Service. 2016. “Bear Facts.” Accessed November 9
th
, 2016. Available at
https://www.nps.gov/yose/planyourvisit/bearfacts.htm

National Park Service. 2013. “Management of Habituation and Food Conditioning in the
National Parks.” Natural Resources Report NPS/BRMD/NRR- 2013/626.

National Park Service. 2012. “Sense of Place: Design Guidelines for Yosemite.” ISBN: 978-0-
16-090412-7

National Park Service. 2009a. “The First Directors: Mather and Albright.” Accessed July 15th,
2017. Available at https://www.nps.gov/cham/learn/nature/upload/nps_first_directors
_6_17_09.pdf

National Park Service. 2009b. “Badger Pass Ski Area.” Accessed August 21st, 2017. Available
at https://www.nps.gov/yose/learn/historyculture/upload/Badger-Pass-low-res.pdf
Accessed July 15
th
, 2017.

National Park Service. 1986. “Architecture in the Parks.” Accessed August 11
th
, 2017. Available
at http://npshistory.com/publications/architecture-in-the-parks.pdf

National Park Service. n.d.a. “Yosemite National Park: Your Safety - Special Protection for
Special Places.” Accessed June 9
th
, 2017. Available at
https://www.nps.gov/yose/planyourvisit/safety.htm

National Park Service. n.d.b. “Harpers Ferry Center: Mobile Apps.” Accessed 1.31.17.
Available at https://home.nps.gov/hfc/products/digitalmedia/mobileapps/

94

National Park Service. n.d.c. “National Mall and Memorial Parks District of Columbia Mobile
App Page.” Accessed 1.31.17. Available at
https://www.nps.gov/nama/learn/photosmultimedia/app-page.htm

National Park Service. n.d.d. “Boston National Historical Park Massachusetts Mobile App.”
Accessed 1.31.17. Available at https://www.nps.gov/bost/planyourvisit/app.htm

New Media Consortium. 2012. “NMC Horizon Project Preview: 2012 Higher Education
Edition.” Accessed July 15
th
, 2017. Available at http://www.nmc.org/pdf/2012-horizon-
HE-preview.pdf.

NMAI. n.d. The 4D/NMAI Virtual Museums Project Story. Accessed 1.31.17. Available at
http://www.nmai.si.edu/exhibitions/all_roads_are_good/VTStory.htm

Ott, A. T., M. Shalaby, U. Siart, E. Kaliyaperumal, T. F. Eibert, J. Engelbrecht, and R.
Collmann. 2011. Performance analysis of a low-cost wireless indoor positioning system
with distributed antennas. Advances in Radio Science: ARS 9: 79-84.

Petrou, Lambros, Georgios Larkou, Christos Laoudias, Demetrios Zeinalipour-Yazti, and
Christos G. Panayiotou. 2014. “Demonstration Abstract: Crowdsourced Indoor
Localization and Navigation with Anyplace”. IEEE Press: 331-332.

Peltola, Elina. 2015. “Map Accuracy Analysis with MapCreator 1.5.” Accessed July 15th, 2017.
Available at https://www.indooratlas.com/2015/06/26/map-accuracy-analysis/

Rao, Bharat, and Louis Minakakis. 2003. Evolution of mobile location-based services. Vol. 46.
New York: ACM.

Rocha, Veronica. 2017. “Wildfire in Yosemite National Park swells to 1,600 acres near
Wawona, historic hotel.” Los Angeles Times. Accessed August 24
th
, 2017.
http://www.latimes.com/local/lanow/la-me-ln-yosemite-fire-wawona-20170815-
story.html

Sargent, Shirley. 1990. The Ahwahnee Hotel. Yosemite Park and Curry Co. Santa Barbara, CA.
ISBN: 0-917859-38-3

Sierra Club n.d.a. “Robert Underwood Johnson.” Accessed July 15
th
, 2017.
Available at http://vault.sierraclub.org/john_muir_exhibit/people/johnson.aspx

Sierra Club n.d.b. “Theodore Roosevelt: 1958 – 1919.” Accessed July 15
th
, 2017.
Available at http://vault.sierraclub.org/john_muir_exhibit/people/roosevelt.aspx

Sierra Club n.d.c. “Hetch Hetchy.” Accessed July 15
th
, 2017.
Available at http://vault.sierraclub.org/ca/hetchhetchy/history.asp

95

Skanect. n.d. “3D Scanning: Fast, Easy, and Low-Cost.” Accessed July 15
th
, 2017.
Available at http://skanect.occipital.com/

SketchUp. “The easiest way to draw in 3D.” Accessed June 20
th
, 2017. Available at
https://www.sketchup.com/

Smithsonian. 2012. Smithsonian museums now mapped from the inside out. US Fed News
Service, Including US State News 2012. Press Release.

SNMNH. n.d. Smithsonian National Museum of Natural History. “About These Tours”.
Accessed June 20
th
, 2017. Available at http://naturalhistory.si.edu/VT3/about.html

Structure Sensor. n.d. “3D scanning, augmented reality, and more for mobile devices.” Accessed
June 20
th
, 2017. Available at https://structure.io/

The Travel Magazine. 2017. “New app launched for the Empire State Building
Observatory.” Accessed July 15
th
, 2017. Available at
http://www.thetravelmagazine.net/new-app-for-the-empire-state-building-
replaces-self-guide-device.html

Tseng, Yu-Chee, Shih-Lin Wu, Wen-Hwa Liao, and Chih-Min Chao. 2001. Location awareness
in ad hoc wireless mobile networks. Computer 34 (6): 46-52.

Unity. n.d. Gaming Engine. Accessed January 28
th
, 2017. Available at https://unity3d.com/

Viametris. n.d. “As simple as traditional indoor mapping techniques.” Accessed July
15
th
, 2017. Available at http://viametris.info/resources/public/iMMS-2/EN/VIAMETRIS-
iMMS-2-Short-Leaflet-A-EN.pdf

Vuforia. n.d. “Developer Portal.” Accessed July 15
th
, 2017. Available at
https://developer.vuforia.com/

Walklet, Keith. 2004. “The Ahwahnee: Yosemite’s Grand Hotel.” DNC Parks and Resorts at
Yosemite, Inc.

Worboys, Michael. 2011. Modeling Indoor Space. Proceedings of the 3rd ACM SIGSPATIAL
International Workshop on indoor spatial awareness, 11/2011.

Xiong, Qing, Qing Zhu, Zhiqiang Du, Xinyan Zhu, Yeting Zhang, Lei Niu, Yun Li, and Yan
Zhou. 2017. A dynamic indoor field model for emergency evacuation simulation.
International Journal of Geo-Information: 6, 104.

Zouggar, N., D. Chen, and B. Vallespir. 2008. Enterprise modelling and ontology. IFAC
Proceedings Volumes 41 (2): 11907-12.

96

Appendix A User Guide

MVTA Sample User Guide

97

Appendix B Draft Flowcharts

MVTA proposed flowchart using Unity and Vuforia with IPS services
98

MVTA proposed Architecture for Unity and BlueCats
99

Appendix C Audio Chapter List
Audio Chapter Title
1.1 National Historic Landmark
1.2 Cornerstone
2.1 Bar
2.2 Floor Mosaics
3.1 History of Yosemite and Stephen Mather
3.2 Organic Act
3.3 Architecture
4.1 Sweet Shop
4.2 Gunnar Widforss
4.3 Gothic Chandeliers
5.1 Women’s Restroom
5.2 Kelim Carpet
5.3 Basket Swirl Mural
5.4 Kitchen
5.5 Dining Room
6.1 Covered Up Mural Above Fireplace
6.2 Navy Occupied the Hotel
6.3 Baskets
6.4 Stained Glass
6.5 Interior Designers
6.6 Restoration Stencil Work
7.1 Winter Club Room
7.2 Donald Tresidder and Skier’s Ten Commandments
7.3 Della Taylor Hoss and Peter Hoss
7.4 Frank Givens and Luggi Foeger
7.5 Leroy Rust and Hannes Schroll
8.1 Fountain
8.2 Firefall
9.1 Mural Room
9.2 Secretary Desks

100

Appendix D POI Photos

Dining Room

Solarium Fountain
101

Kelim Carpet

Winter Club Room Panel 1
102

Sweet Shop

Floor Mosaics
103

Old Mural in Great Lounge

Gothic Chandeliers
104

Organic Act

American Indian Baskets
105

Kitchen

Gunnar Widforss
106

National Historic Landmark

Women’s Restroom
107

Secretary Desks

Cornerstone
108

Stained Glass

Winter Club Room Panel 3
109

History of Yosemite and Stephen Mather

Restoration Stencil Work
110

Mural

Firefall (film)

111

Great Lounge

Bar
112

Basket Swirl Mural and Jeanette Dyer Spencer

Navy Occupied the Hotel
113

Winter Club Room Panel 4

Winter Club Room

Asset Metadata

Creator Denson, Trevor James (author)

Core Title Majestic Yosemite Hotel virtual tour application and indoor model

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 09/26/2017

Defense Date 09/01/2017

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Android,application,augmented reality,bluetooth,indoor modeling,indoor positioning system,iOS,Majestic Yosemite Hotel,mobile application,Model,OAI-PMH Harvest,Tour,Unity,virtual reality,virtual tour,wireless,Yosemite,Yosemite National Park

Language English

Advisor Swift, Jennifer (committee chair), Chiang, Yao-Yi (committee member), Wu, An-Min (committee member)

Creator Email tdenson@usc.edu,trevorjdenson@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-435803

Unique identifier UC11263811

Identifier etd-DensonTrev-5774.pdf (filename),usctheses-c40-435803 (legacy record id)

Legacy Identifier etd-DensonTrev-5774.pdf

Dmrecord 435803

Document Type Thesis

Rights Denson, Trevor James

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract The Majestic Yosemite Hotel, formerly known as the Ahwahnee, is a National Historic Landmark located in Yosemite National Park, California, USA. Built in 1927, the hotel attracted rich and wealthy individuals to help gain financial support for the National Park Service idea of protecting wild spaces for future generations. To this day, the hotel stands as one of the National Park Service’s most historic lodging units, providing luxury accommodations and services to park visitors. In November of 2016 Yosemite Hospitality, Yosemite National Park’s Concessionaire requested a mobile application to educate visitors on the cultural and historical significance of the hotel to support the goals of the Long Range Interpretive Plan. Yosemite Hospitality was the client for this project, and the application was developed in direct consultation with Yosemite Hospitality’s Interpretive Services Department from November 2016 until August 2017. Several indoor positioning technologies and Augmented Reality services were tested to deliver educational content based on user mobile device locations and camera orientations. The processes tested the Anyplace indoor positioning service, IndoorAtlas indoor positioning service, BlueCats beacon services, Vuforia Augmented Reality services, and the gaming engine Unity. Testing and development occurred on both Android and iOS devices with development in Javascript, C#, Swift, and Objective C. As part of this thesis work, a historical model with digital furniture scans was also completed to preserve the current conditions of the hotel’s original furniture. These scans were based on the Structure Sensor manufactured by Occipital. This thesis documents the development and testing of the Majestic Virtual Tour Application and the historic furnishings model built for the Majestic Yosemite Hotel in fulfillment of the Yosemite Hospitality project.