Professional Documents
Culture Documents
Workspaces
Michael Vernik
LIST OF FIGURES..............................................................................................................................III
LIST OF TABLES..................................................................................................................................V
ABSTRACT...........................................................................................................................................VI
DECLARATION.................................................................................................................................VII
ACKNOWLEDGEMENTS...............................................................................................................VIII
1 INTRODUCTION.....................................................................................................................- 1 -
1.1 AIMS...................................................................................................................................- 2 -
1.2 THESIS CONTRIBUTIONS.....................................................................................................- 2 -
1.3 OTHER RELATED CONTRIBUTIONS.....................................................................................- 3 -
1.4 STRUCTURE.........................................................................................................................- 4 -
2 UBIQUITOUS WORKSPACES..............................................................................................- 5 -
2.3 LIVESPACES......................................................................................................................- 16 -
2.3.1 LiveSpaces Reference Architecture............................................................................- 16 -
2.3.2 Meta Applications.......................................................................................................- 20 -
i Michael J Vernik
2.3.3 LiveSpaces Environments...........................................................................................- 21 -
2.3.3.1 University of South Australia e-World Lab....................................................................- 21 -
2.3.3.2 DSTO Intense Collaboration Space................................................................................- 23 -
2.4 AUSPLANS.........................................................................................................................- 27 -
3.4 PLEXUS.............................................................................................................................- 40 -
3.5 SESSION MANAGEMENT...................................................................................................- 43 -
4.1 IGNITE...............................................................................................................................- 45 -
4.2 PLEXUS.............................................................................................................................- 48 -
4.3 SESSION MANAGER..........................................................................................................- 51 -
5 EVALUATION........................................................................................................................- 53 -
ii Michael J Vernik
5.4.2.1 Overall Responses..........................................................................................................- 69 -
5.4.3 PointRight...................................................................................................................- 75 -
7 CONCLUSIONS......................................................................................................................- 85 -
8 REFERENCES........................................................................................................................- 87 -
iv Michael J Vernik
Figure 5-5. Ignite Results for Coordination Mechanic Questions. in TE2.1.........- 68 -
Figure 5-6. Plexus Overall response graph of participants in TechEx 1.2............- 70 -
Figure 5-7. Plexus Overall response graph of participants in TechEx 2.1............- 70 -
Figure 5-8. Plexus Results for Communication Mechanic Questions. in TE1.2...- 71 -
Figure 5-9. Plexus Results for Coordination Mechanic questions in TechEx 2.1.- 73 -
Figure 5-10. PointRigh Results for Coordination Mechanic questions in TE 2.1.- 76 -
Figure 5-11. TE2.1 results based on each question for PointRight.......................- 77 -
v Michael J Vernik
List of Tables
Table 5-1. Participants during TechEx 1.2 and TecEx 2.1. Note: participants in grey
were present at TechEx 1.2 and 2.1.................................................................- 64 -
Table F-1. Plexus 1.1 Data...................................................................................- 118 -
Table G-1. Ignite TE 2.1 Data – focus group.......................................................- 119 -
Table G-2. Ignite TE 2.1 Data – control group....................................................- 119 -
Table G-3. Plexus TE 2.1 Data – focus group......................................................- 120 -
Table G-4. Plexus TE 2.1 Data – control group...................................................- 120 -
Table G-5. PointRight TE 2.1 Data – focus group...............................................- 121 -
Table G-6. PointRight TE 2.1 Data – control group............................................- 121 -
vi Michael J Vernik
Abstract
This thesis reports on research undertaken into session interfaces for ubiquitous
workspaces. These future workspaces are media-rich environments that employ new
types of operating systems and services to coordinate and manage interactions
between people, multiple display surfaces, information, personal devices, and
workspace applications. Session interfaces are a new class of application which will
be of critical importance to the success of teams operating in these future work
environments. They are aimed at reducing the cognitive load required by users in
understanding and interacting through a vast array of application and device interfaces
and making these spaces more accessible and useable by the general community,
much as WIMP interfaces have done for personal computers. Several contributions
have resulted from the research reported in this thesis. The research area for session
interfaces has been scoped through extensive literature review, initial definitions and
formalism that help provide the foundations for future research in this area have been
developed, prototype session interfaces have been developed and integrated into two
LiveSpaces environments and initial evaluations have been conducted as part of the
DSTC AUSPLANS project using the new Orchestrated Evaluation approach.
I certify that the thesis written by Michael John Vernik entitled Session Interfaces for
Ubiquitous Workspaces and submitted for the degree of Advanced Computer and
Information Science (honours) is the result of my own research except where
otherwise acknowledged and that this thesis or any part of the same has not been
submitted for a degree to any other university or institution.
Michael Vernik
15-11-2004
I would like to start with my supervisors Bruce Thomas and Damien Bright for the
guidance you have given me this year. Without your guidance, it would be very
difficult to undertake this honours thesis.
I would also like to thank my Dad, Rudi Vernik for his support with this thesis.
Sometimes during an honours year, it is easy to get off track. He has been an amazing
influence in keeping me in the right direction and his advice is always valuable.
The fellow students I have worked with this year, in particular Steven Johnson,
Andrew Cunningham, Dennis Hooijmaijers, Benjamin Avery and Vivian Nguyen.
Discussing and learning concepts together has been a fun and mind-expanding
experience.
The colleagues I have worked with over the past year. Rudi and Peter Evdokiou for
their development of LiveSpaces concepts and infrastructure. Eugene Kharabash for
his technical support. David Karunaratne and Tom Roberts for their ongoing
development work on the orchestrated evaluation framework, meta applications and
the DSTO LiveSpace environment.
The support and encouragement given by my friends Stewart Itzstein, Antony Brewer,
Garrie Bubicich is really appreciated.
ix Michael J Vernik
My entire family, Rudi, Sue, Maree and my partner Amanda. Their support this year
has been incredible and has made the year so much smoother. Each have gone
through major events themselves this year, and the support they have offered during
these times has been very special.
x Michael J Vernik
1 Introduction
Ubiquitous workspaces are future media-rich environments that employ new forms of
operating systems and services to coordinate and manage interactions between people,
multiple display surfaces, information, personal devices, and workspace applications.
Researchers in this area have has been focusing their attention on the technology
augmentation of physical collaborative workspaces such as meeting rooms. Examples
of research initiatives in this area include Stanford’s Interactive Workspaces [STA],
GMD’s i-Land [AMB], MIT’s Intelligent Room[MIT], and Active Spaces at the
University of Illinois Urbana Champaign (UIUC)[ACT].
Much of the research undertaken in the area of ubiquitous workspaces has focused on
the infrastructure aspects of providing the operating environments for integrating and
coordinating various workspace technologies and applications. LiveSpaces draws
from and extends this research to focus on the application of ubiquitous workspaces
for intense collaborative activities. Work is underway to define and evaluate a range
of workspace services and applications to support teams involved in activities such as
defence planning, software review, and decision-making in health domains. One
1 Michael J Vernik
class of application of critical importance to the success of these teams will be
session interfaces. These applications will provide a common interface through
which users can control and interact with the host of devices, services, media and
information available in the workspace. They are aimed at reducing the cognitive
load required by users in understanding and interacting through a vast array of
application and device interfaces. They will support team collaboration by providing
ready access to required information and by providing access to a range of workspace
knowledge and support services such as instant messaging, speech transcription, and
context awareness. Session interfaces have the potential to make future ubiquitous
workspaces accessible and usable by a broad range of users, much as WIMP
(window, icon, menu, pointer) interfaces have made personal computers accessible to
the general public.
The key research question that has driven the research discussed in this thesis is:
What impact could new forms of session interfaces have on the effectiveness of
teams in ubiquitous workspaces?
1.1 Aims
The aims of this thesis project were to:
2 Michael J Vernik
Extensive review and consolidation of literature to help scope and define the
research area for session interfaces. (Chapter 2).
Use initial evaluation approach to evaluate session interfaces. This has helped
test the evaluation approach and allowed for future development work to help
evaluate LiveSpace components more effectively.
3 Michael J Vernik
Initial research on workspace simulation. This was conducted as part of a
research methods subject in 2003 and resulted in research publication
[VER04c] and an honours project which extended and implements the
concepts.
1.4 Structure
This thesis is structured as follows; Chapter 2 sets the context of the thesis by
providing an overview of ubiquitous workspaces and a review of relevant literature.
Chapter 3 provides definitions, formalisms and characteristics of session interfaces.
The session interfaces used for experimentation are also discussed. Chapter 4
discusses the implementation details of the session interfaces described in Section 3.
Chapter 5 covers the evaluation approach and results obtained during the evaluation.
Chapter 6 discusses the results and highlights areas of future work. Chapter 7
provides a conclusion to the thesis. Following this, there are a number of appendixes
which supply support material to the thesis.
4 Michael J Vernik
2 Ubiquitous Workspaces
Ubiquitous computing relates to a future where computers will become embedded in
our natural movements and interactions with our environments (both physical and
social). In his seminal article on “the computer of the 21st Century” Weisser
[WEI91] introduced the concept of ubiquitous computing. He described a future
where computers will become largely invisible to users, being “woven into the
backdrop of natural human interactions”.
There are a number of terms that describe and contribute to this future trend in
computing. These include ubiquitous, pervasive, mobile, wearable, environmental
computing. Lyytinen and Yoo [LYY02] argue that ubiquitous computing in some
sense encompasses both pervasive and mobile computing in that it deals with large-
scale mobility (the focus of mobile computing) and the concept of making the
computer invisible through high levels of embeddedness.
5 Michael J Vernik
2.1 Related projects
2.1.1 Using Project/War Rooms for Extreme Collaboration
Several research programmes are focusing on defining the processes and facilities to
support extreme (or intense collaboration). This work is referred to as project
[COV98] or war rooms. Mark [MAR02a] describes extreme collaboration as
“Working within an electronic and social environment that maximizes communication
and information flow.” She did a study of a 16-person team that designs space
missions NASA team at the Jet Propulsion Lab (JPL). Her study showed that the
amount of time required to create a space mission design was reduced from 3-6
months to about 9 hours by switching to a technologically rich “war room”
environment. The JPL war room was set up with use several public displays, database
programs that allow access to data on past space missions, satellite orbit visualization
programs, and a custom publish-subscribe system that allowed individuals to push and
pull spreadsheet data to and from other workstations.
A war room [MAR02b] is described as a team collaboration room where all work
takes place in the room. It allows team members to continually move back and forth
between individual work, small group work, and orchestrated teamwork that results in
the development of some product. Mark distinguishes a war room to be different
from a control room. She argues that control rooms are event driven where workers
react to a continuous stream of events in a desired way (e.g. in an air traffic control
center).
The study showed the importance of peripheral awareness, the importance of having
direct access and communication with others (to, for example, discuss alternatives and
question assumptions), and the ability to quickly identify and correct errors. The war
room approach requires: experts to be always available, use of dynamic electonic
artifacts, support for human and electronic networking, and team relationships made
visible.
6 Michael J Vernik
sliders, and wands [BAL03a]. An infrastructure called the interactive Room
Operating System (iROS) [JOH02] has been developed to support ubiquitous control
of applications across a workplace, the movement of data across applications and
devices, and dynamic application coordination. The underlying architectural metaphor
for iROS is a blackboard with two levels of data abstraction implemented using an
extended tuple space approach. iROS also implements a persistent context memory
that allows any process to store and retrieve data that will be relevant across
applications and sessions [WIN01]. iROS includes several other components
including iCrafter [PON01], a system that provides support for service advertisement
and invocation and for the generation of service user interfaces, and PatchPanel
[BAL03b], a mechanism for integrating heterogeneous software and hardware
components into a ubiquitous workspace.
1
Much of the work on workspace interfaces was demonstrated during interactions with the Interactive
Workspaces team but has not been widely published.
7 Michael J Vernik
agents and device controllers. A primary focus of the project is in the area of context
awareness to allow environments to be aware of, and respond automatically to users’
needs (designated a type of intelligence). The work also focuses on natural modes of
human-computer interactions including the use of speech and gestures.
The Intelligent Room project has developed new approaches to resource management
in ubiquitous workspaces (see discussion on Rascal in Section 2.2.2.2) and
applications such as Meeting Manager which aids the planning, facilitation and
recording of meetings though multi-modal interaction and multi-media capture
(Section 2.2.3.3).
8 Michael J Vernik
context, presence, a space repository (using CORBA Trader), and a context file
system. The concept of a Unified Object Bus (UOB) is a key underlying abstraction.
Gaia OS uses an extended version of the model-view-controller approach (MVC)
called MPACC (Model, Presentation, Adapter, Controller, Coordinator) to guide
application development. The Active Spaces work takes on more of an enterprise
computing perspective than the other projects discussed in that the focus is on
extending the concepts of traditional operating systems to support heterogeneous
distributed ubiquitous spaces, where a space could range from a single mobile user
through to collaborative environments such as meeting rooms and classrooms.
Most of the ubiquitous workspaces projects focus on the control and use of devices
within physical workspaces. Virtual and augmented reality approaches provide new
dimensions for how humans can interact in work settings. The Livespaces project has
developed an infrastructure based on the AR Toolkit [KAT99], Passive Detection
Framework[SLA03], and the use of Fiducial Marker Cards to experiment with these
concepts.
9 Michael J Vernik
integration of various devices into a workspace and to support the generation of
common interfaces. This work is discussed in Section 2.2.2. Various applications
have been developed to support human activities in ubiquitous workspaces. These
include applications that support the control of a cursor across multiple display
surfaces using a single pointing device, applications that support the recording of
workspace activities and new applications that support the automation of sessions
such as briefings and presentations. These types of applications are discussed in 2.2.3.
10 Michael J Vernik
To detect the rotation motion, accelerometers have been placed in the wand. These
are communicated to a PC via RF technologies. To track the direction the wand is
pointing (to determine which device we are interacting with), infrared LED’s have
been placed on the wand. Using camera tracking of the LED’s, an accurate direction
can be obtained.
A user study was conducted to determine whether the tracking technology was
sufficient, and whether audio feedback for the device was required. Ten participants
were selected to interact with a selection of 5 devices placed in a living room setting.
The participants reported to have found the device very natural to use. Information
collected showed that participants did require a large degree of tracking accuracy to
use the XWand. Results also showed that inconsistent tracking can be compensated
by clear audio feedback.
Another interesting device that is the product of research at is the Virtual Mouse
Vision Based Interface [ROB04]. This interface uses a robotic head, visual tracking
of the user’s head and hand positions and recognises user hand signs. This approach
allows control of a mouse pointer and buttons on a large display using hand signs and
movements. The focus of the research is on providing ubiquitous interaction with
intelligent kiosks. A robotic head called IGOR (Intelligent Gaze Oriented Robot)
tracks people as they approach the display. When the user makes a recognised hand
gesture (e.g. thumbs up), the system allows movement and control of the cursor on the
screen. These types of approaches have application in ubiquitous workspaces to more
naturally support team interactions. They raise some interesting possibilities if they
were to include other interaction modalities such as “look to talk” [OH02] and were
more fully integrated into the workspace infrastructure.
2
X10 is used to control electrical devices such as the coffee machine and desk lights in the UniSA
LiveSpace.
11 Michael J Vernik
two frameworks discussed in the following paragraphs are iCrafter[PON01] from
Stanford iWork and Rascal [GAJ01] from MIT Oxygen.
2.2.2.1 iCrafter
iCrafter is a software framework that supports the integration of devices into a
ubiquitous workspace. The goal of iCrafter is allow users in a workspace to naturally
interact with services (i.e. electronic devices such as lights, projectors or applications
such as PowerPoint). iCrafter does this in a number of ways. Firstly, it provides a
common aggregated repository for device services that can be easily accessible by
users. It does this ‘on the fly’ by utilizing the iROS event heap as a publish/subscribe
data mechanism.
Secondly, iCrafter manages user interface creation based on services that are
available. Rather than leaving user interface creation to individual services, it can
produce custom or generic interfaces. This allows easy utilization of legacy services,
as well as specially designed interfaces for a function required in the workspace. As
well as this, legacy services do not need to be rewritten to have a common interface
developed for them.
2.2.2.2 Rascal
Rascal is a resource manager for the MIT Intelligent Room. It takes a differing
approach to iCrafter. Rascal provides two pieces of the room infrastructure. Firstly,
the approach involves the creation of a registry of services that are available in the
environment. This differs from the iCrafter approach in that Rascal allows
applications to request a service, rather than the iCrafter approach where a service
represents a device. For example, one benefit of this approach is that it allows the
ability for multiple applications to co-exist in the environment, utilising services
without interrupting each other’s operations. Since applications are requesting
services, rather than devices, if there is a service that can serve the same functionality,
it can utilize a different device. For example, this may be something simple like
wanting to place a diagram on a display, but since that resource is being used, Rascal
12 Michael J Vernik
can place the diagram on another display that is unused. This supports a kind of
intelligence for the room
The approach taken for Workspace Navigator is to take snapshots of the workspace
from different angles at a predetermined time. For example, this may be every 20
seconds. The reason to use imagery over video was a resource requirement, as large
video can often be difficult to navigate after a session. This information is then
displayed in an application (shown figure 2-7). Navigation of the session is done
through a timeline arrangement, where a user can navigate to a particular timeframe,
and view the images developed at that instance. As well as this, other information,
such as annotations that have been created is shown for that timeframe.
One of the contributions that this work makes for this thesis project is that it is one of
the only related pieces of research that defines the notion of session. A session is
defines as: Each virtual workspace contains a series of sessions, each for a bounded
period of time. A group meeting is a typical session, or users may interpret a single
meeting as a sequence of distinct sessions if the topic changes. Each session takes place
13 Michael J Vernik
in a physical location, or setting. There can be different settings for sessions in a single
virtual workspace, e.g., when a project holds meetings in more than one place. [ION02].
Additional research is required to more formally define the notion of session to provide
the basis for session management and session support in ubiquitous workspaces.
Workspace Navigator was evaluated with two user studies to gauge the effectiveness
of the Workspace Navigator suite. The first activity involved a real entrepreneurship
meeting where a research group from a business school asked to conduct a meeting in
an iRoom and evaluate Workspace Navigator. The number of participants ranged
from 5 to 13 at any one time. One of the key problems raised was the difficulty of
searching the enormous amounts of information when sessions had not been explicitly
identified by users. The authors suggested that explicitly starting and ending sessions
helped organise the flow of meetings and allowed participants to establish a common
context for interactions.
2.2.3.2 PointRight
In ubiquitous workspaces, there is often a need to have control over displays using a
particular input device. PointRight is an application that provides input connectivity
from a range of devices (laptops, displays inputs) to many of the displays provided in
a workspace. It is part of the Stanford Interactive Workspaces project (Section 2.1.2)
Movement of a cursor across displays in an environment is done in a similar way to
which control of multiple monitors is done in popular operating systems. This gives
users a consistent and more natural feel.
14 Michael J Vernik
2.2.3.3 MeetingManager
MeetingManager is an application for the MIT iROOM to support planning,
facilitating and browsing structured meetings [OH01]. It comprises four components
that work together. Participants use the Meeting Planner to make and store an agenda.
Parameters of each agenda item include a description of the item, proposed length of
discussion, priority and state. The state allows agenda items to be continued at a later
date. The participants in a meeting session is also recorded.
The facilitator displays the agenda and records the discussions as a QuickTime video.
A multimodal interface is used which includes speech, typing, and electronic white
board support. The facilitator captures commitments and issues for each agenda item
and these can be displayed alongside the agenda. The Meeting Summariser
automatically sends a summary of the meeting to participants. The summary includes
the agenda, issues, commitments together with hyperlinks to sections of the video for
reference. The Meeting Browser allows the querying of meeting objects stored in a
database. This provides participants the ability to search for and review previous
meetings. The basic concepts have been articulated in the work by Oh et al.
However, further information has not been published and so it is not known whether
the approach has been implemented or evaluated.
There have been several applications developed to provide electronic support for
meetings. These are one type of session that needs to be supported in a ubiquitous
workspace. Much of the previous work has focused on recording of activities through
the capture of speech and video transcripts [NUN91, CHI00]. In the session
interfaces work being undertaken as part of this thesis project, broader support for
sessions is envisaged where users can interactively and more effectively initialise and
control the resources of a future workspace to support interactions between people,
technologies, media, and information. Support for the recording and browsing of
interactions is also an important requirement for session interfaces.
2.2.3.4 Multibrowse
Often in a ubiquitous work environment, there is a need to transfer information from
one display to another. For example, if one user finds a relevant URL on their
personal laptop, they may wish to share this with other members of a session.
15 Michael J Vernik
Traditionally, this would have involved retracing the steps to retrieve the document,
re-typing the URL, or in some cases, using a memory stick or similar device to
transfer a document.
2.3 LiveSpaces
LiveSpaces is a collaborative ubiquitous workspace project between University of
Australia, DSTO and DSTC. In this section, the LiveSpaces architecture is explained.
Following this, a novel use of a LiveSpace, meta-applications will be discussed. The
project has developed a number of LiveSpaces implementations. These include the
UniSA e-World Lab and DSTO Intense Collaboration Space, where further
development details will be given.
16 Michael J Vernik
research. The architecture has at its core a Workspace Infrastructure that coordinates
access to various workspace services, devices, computers and applications within a
ubiquitous workspace. The Workspace Infrastructure acts as the operating system for
the entire workspace. In LiveSpaces, the Interactive Room Operating System (iROS)
[JOH02] from Stanford University is used as the basis of the Workspace
Infrastructure. A key component of iROS is the Event Heap which acts as a
blackboard with two levels of data abstraction implemented using an extended tuple
space approach. iROS also implements a persistent context memory called the Data
Heap which allows any process to store and retrieve data that will be relevant across
applications and sessions. iROS includes several other components including iCrafter,
a system that provides support for service advertisement and invocation and the
generation of user interfaces (used for controlling lights, projectors, etc), and
PatchPanel [BAL03b], a mechanism for integrating heterogeneous software and
hardware components into a ubiquitous workspace.
Another key architectural element in LiveSpaces is the Enterprise Bus. It provides the
mechanism for communication and synchronization within and between distributed
17 Michael J Vernik
workspaces. It also provides a flexible approach for integrating various enterprise
computing elements such as ontology services, database services, workflow engines,
and other web services. In our current LiveSpaces implementation we use a publish
subscribe system called Elvin [ELV], and the associated ODSI [BON01] peer-to-peer
infrastructure as the basis of the Enterprise Bus. OSDI peers are generic software
elements that act together as application servers to discover and manage the use of
services.
A key focus of the LiveSpaces work is to investigate how various workspace support
services and knowledge services can be employed to facilitate intense group activities
such as time critical decision making, analysis, and planning. As part of the
LiveSpaces work, a novel Passive Tracking and Detection Framework [SLA03]
service, based on augmented reality technologies, has been developed to track objects
such as people and devices in a workspace. The approach uses a camera array and
fiducial marker cards [KAT99]. The cards can be used to both track objects and act as
ambient information displays. Other services that we have available in the LiveSpace
include an initial workspace orchestration service based on the use of a workflow
engine called Breeze[BRE], and speech recognition and synthesis services.
LiveSpaces researchers are currently working on advanced workspace services
including media services and automated speech transcription services.
18 Michael J Vernik
Figure 2-10. LiveSpaces Reference Architecture.
19 Michael J Vernik
a representative work environment populated with relevant tools and artifacts.
Orchestrated evaluation will be used as part of the AUSPLANS project for evaluating
the session interface approaches developed as part of this thesis project.
There are a number of implementations of LiveSpaces that have been or are being
developed. These include UniSA e-World lab, DSTO Intense Collaboration Space,
DSTC Mobile LiveSpace and C2 Developmental BattleLab at the Deployable Joint
Force Headquarters in Brisbane.
The first style of meta application was developed to be a self describing showcase of
the people and technologies present in a LiveSpace. This was of particular
importance, as many stakeholders and visitors would pass through LiveSpaces. It was
often difficult to have people available for demonstration and presentation. These
types of activities are also costly in terms of peoples’ time. e-Ghosts was the first of
meta application developed which allowed for a multimedia rich showcase of work,
initiated from a simple web front end. Figure 2-11 shows the positioning of the e-
Ghosts meta application within the LiveSpaces implementation.
A second style of meta application has been developed to allow for evaluation of
AUSPLANS components and technologies. It is extended upon the e-ghosts
framework, and provided methods for feedback of technologies in a scenario driven
20 Michael J Vernik
environment. The use of meta applications for these Orchestrated Evaluation
Sessions, is covered in further depth in Section 5.2.
21 Michael J Vernik
Figure 2-12. The UniSA LiveSpaces e-World Lab during 2002.
The e-World lab today is a very technology rich space. Figure 2-13 shows some of
the key infrastructure. There are three projectors placed side by side, two large LCD
displays (ambient displays) and a CRT display providing display functionality.
Infrastructure such as downlights, spotlights and even the original coffee machine are
controlled through iROS via X10 mechanisms. Display interaction is provided by a
Smart Board as well as e-beam devices which can be moved to any surface in the
environment to allow handwriting tracking. A camera array in the LiveSpace is
utilized by the passive detection framework to provide marker tracking for augmented
reality technologies. A wireless network has been installed in the environment. This
gives the ability for the wireless laptops and tablet PC’s to roam freely inside the
space. Computers in the environment include two Solaris machines, a Windows
server machine, as well as numerous supporting machines for displays and cameras.
These are located in the cupboards you can see on the outside edge of the room. A
four point surround sound system has also been placed in the environment.
22 Michael J Vernik
Figure 2-13. University of South Australia LiveSpaces at e-World lab today.
Figure 2-14 highlights some of hardware technologies in the ICS. These include a
three projector array at the front of the room. These are attached to a sliding rail that
allows for microphone, camera or other display devices to be attached. Also attached
to the roof is a robotic camera array (Figure 2-16) and microphone array that is
utilized by an audio visual capturing device, MAVCap (Mobile Audio Visual
Capture) (Figure 2-18). E-beam devices are used on each whiteboard in the space to
allow for the electronic capture of handwriting and drawings by users. There are
currently four whiteboards as well as the table surface which is also made of
23 Michael J Vernik
whiteboard material (Figure 2-15). The tabletops also provide users with AV
capabilities (such as video and audio plugs), power and network capabilities (Figure
2-19).
There is one computer rack (Figure 2-17) used in the environment to house all the
computers and other devices that are utilized. Five computers are used in the
environment (1 Linux server, 1 Windows server and 3 display supporting computers).
In the rack also resides an AV Matrix switch. This switch allows any of the inputs
(such as the display computers or an individuals laptop) to be placed on any output
(such as a projectors or sound system in the environment).
Unlike the e-world lab, wireless networks are not permitted at DSTO. There are
approximately 50 wired network ports provided so that users are able to work
wherever they would like in the environment. This network has 1 Gigabit
functionality to allow for many types of media that are now and will be used in the
future.
All lights (numerous down lights and fluorescent lights) are controlled in the
environment using C-Bus technologies from Clipsal. This has provided an interface
that is controllable by a standard light switch as well as by software interfaces such as
Plexus (Section 3.4) or Ignite (Section 3.3).
An LCD touch screen has been placed in the environment to allow for interfaces such
as Ignite (Section 3.3) to be easily accessed in the environment. This touch screen is
located on the back of a ‘barn style’ door (Figure 3-5). Placing it here makes it one of
the first things that is noticed as you enter and leave a LiveSpace.
24 Michael J Vernik
Figure 2-14. DSTO Intense Collaboration Space
25 Michael J Vernik
Figure 2-17. LHS - Server rack containg hardware
Figure 2-18. RHS - MAVCap System
Figure 2-19. Input/output board at each desk in the ICS. From left to right, AC
power, network, video cable input, audio cable input
26 Michael J Vernik
Figure 2-20. AV Matrix to allow video and audio switching.
2.4 Ausplans
AUSPLANS, or Augmented Synchronised Planning Spaces, is a defence domain
project coordinated by the Collaborative Research Centre for Enterprise Distributed
Systems Technology or DSTC. The project involves participants from several
organisations including DSTO, the University of South Australia, DSTC and Boeing.
AUSPLANS focuses on the application and extension of LiveSpaces research for
distributed synchronised planning within future joint headquarters. It looks at how
physical workplaces such as meeting rooms can be rapidly augmented with and
enabled using emerging ubiquitous computing and human interaction technologies.
27 Michael J Vernik
There are a number of issues that need to be addressed to more effectively support
planning teams. These include:
Support for the pre planning, initiation and rescheduling of team activities
The platform allows for rapid integration and evaluation of session interfaces and
other innovations in areas such as advanced planning, visualisation and decision-
making tools. These are evaluated within the context of future ubiquitous team spaces
and operational situations.
These orchestrated sessions provide the context in which various capabilities can be
evaluated. LiveSpaces meta-applications coordinate a range of media, information,
simulations devices and applications around a particular story line or scenario. These
scenarios use real and represented data, together with actual applications. The high
levels of automation that can be achieved in a LiveSpace provides an effective basis
for undertaking controlled and repeatable capability evaluations.
28 Michael J Vernik
Effective synchronisation and interaction between multiple livespaces architecture are
also goals of the research. A workspace simulation system has been developed by the
project to support investigations involving several planning teams. Virtual livespaces
can be invoked through simulation to provide inputs and interactions for evaluations
involving an actual team resident in a physical LiveSpace. The use of workspace
simulation for training purposes is also being considered.
29 Michael J Vernik
3 Session Interfaces for LiveSpaces
Session Interfaces are a class of application that provide a common interface between
the workspace infrastructure, other workspace applications (including meta-
applications), devices and most importantly team members . This chapter begins by
defining the notion of session interfaces.
Session Manager. An interface that allows the set up, invocation, storing, and
retrieval of sessions.
There has been little discussion in the literature on how ubiquitous workspaces are
arranged and used for specific elements of work. The Workspace Navigator research
30 Michael J Vernik
discussed in Section 2.2.3.1 is one example of where these concepts needed to be
defined and explored since the work deals with capturing and recording workspace
activities. They define workspace activities as: “Each virtual workspace contains a
series of sessions, each for a bounded period of time. A group meeting is a typical
session, or users may interpret a single meeting as a sequence of distinct sessions if the
topic changes. Each session takes place in a physical location, or setting. There can be
different settings for sessions in a single virtual workspace [ION02].” The notion of
session used in Workspace Navigator corresponds to the dictionary definition and the
approach taken in LiveSpaces. Once again, sessions are defined as an element of work
being undertaken during a bounded period of time. The Workspace Navigator definition
also includes the concept of the structuring of sessions and the context, or setting, for
sessions. The adaptive nature of ubiquitous workspaces is also mentioned in that a
workspace can provide different settings for sessions.
In order to provide better foundations for research in this area, a more formal definition of
work in ubiquitous workspaces is required. This thesis provides an initial working
definition: A session as a definable unit of activity with specified goals and outcomes,
undertaken within and/or between ubiquitous workspaces, at a bounded and specified
point in time. The key characteristics that define a session are:
31 Michael J Vernik
More formally, a session (S) can be defined in terms of the set of characteristics stated
above.
There are also relationships between the characteristics of a session that need to be
considered. For example, there is often a relationship between a Goal, particular
Outcome(s) and underlying Tasks. This aspect can be modeled using systems
notation proposed by Kaposi and Meyers (1994). If a session (S) is considered to be a
system which is defined in terms of a set of Characteristics (C), and the relationships
between them (RC), then the following formalism can be used to represent the model:
Similarly each characteristic could be considered to be a system in its own right using
this modeling approach. For example, a characteristic such as Sg (setting) could be
described in terms of a set of elements of the setting (such as location, services
available, security aspects, etc) and the relationships between them. The setting could
then be modeled to define, for example, the relationships between the services
available and the security restrictions on accessing these services at a particular
location. This modeling approach can also extend to situations where a session is
distributed and may involve teams in a set of LiveSpaces. In these cases the setting
would be defined in terms of the set of locations and the relationships between them
where MSys(Sg) is the model of the setting which in this case includes a specific set of
locations and where RL is the set of relations across the set of locations:
The notation MSys(X) is used to indicate the system being described is a simplified
representation of the real system (in this case a model of the setting which focuses on
defining the locations and the relationships between the three locations which form
the basis of the team that is collaborating as part of a particular session).
This formalism also extends to modelling a set of related sessions. For example, there
are cases where a session might be comprised of a set of sub sessions. As described
in Section 5.2.2, the AUSPLANS Technical Exercises are defined as an Orchestrated
32 Michael J Vernik
Evaluation Sessions. Each TE, is broken down into as set of sub-sessions which
provide automated scenario briefings, capability demonstrations and evaluations, and
a scenario exercise. In this case a structural systems model for TE2.1can be used to
denote this as:
Where TE2.1 is made up of a set of seven sessions and the relations across the
sessions, Rs, is a simple sequential relationships which indicates that each session
begins after the previous one completes.
The session formalisms defined in this section are a first attempt to provide a
framework in which further discussions can ensue.
33 Michael J Vernik
those provided with Microsoft Office as well as unique applications to the
environment, such as SmartBoard software, PointRight and Multibrowse. Session
Interfaces can be used to provide a common interface where the user is not required to
know what applications are available and how to start using them. For example, a
universal display service could be used to allow a user, from a universal interface, to
start an application by dragging an icon (such as a word document) onto a display
represented on a floor map and then automatically invoke the application on the target
display. This is one of the approaches used in Plexus, a universal session interface that
has been developed for LiveSpaces and discussed in this chapter.
There are a number of attributes that should be associated with a session interface. In
summary, the requirements for session interfaces in a ubiquitous workspace include:
34 Michael J Vernik
Ubiquitously controlling the display of information on various display surfaces
using various applications
Universal – the interfaces are common in the workspace and each participant can
use them.
3.3 Ignite
Ignite is an application that provides a start and finishing point for a session in a
LiveSpace. It allows participants to easily initialize the room, having control of
common devices such as lights and displays. It also provides ambient information to
users such as displaying the room’s health.
35 Michael J Vernik
When a user enters a LiveSpace, they are greeted with a very simple interface as
shown in Figure 3-2. At the top of the display we can see 3 buttons. To start the
LiveSpace and use it in a default state, we simply press start LiveSpaces. This will
give access to applications available in the space, such as PointRight, Plexus and
Sticker as well as many others. When the user is finished, they can press stop
livespaces which will switch off devices in use so that the user can leave.
Figure 3-2. Ignite main page: User starting a LiveSpace using Ignite.
The second area of Ignite is custom or predefined session that can be invoked in
LiveSpaces. If a user requires further initialization of a LiveSpace than the start
LiveSpaces button provides, they can select “custom start up”. Figure 3-3 shows the
custom start up screen. This area allows users to choose from a number of options,
including launching meta-applications or selecting LiveSpace setups for a particular
style of session.
36 Michael J Vernik
Figure 3-3. Custom startup screen in Ignite!
The third area displays the status of the room. If a device, including a computer or
display, or a service including applications such as Pointright are not functioning, an
ambient inidication is provided to the users. Critical items that are required for a
LiveSpace to function are coloured in red. Non-critical but failed items are coloured
in orange. Fully functioning items are coloured green. In Figure 3-4 we see that there
is a problem with the right projector screen and the Workspace Simulator services are
not available.
37 Michael J Vernik
Figure 3-4. Status of services in a LiveSpace environment
The Ignite computer is comprised of a normal CRT display with mouse input. Ignite
is run as an active desktop component on a WindowsXP machine. This is particularly
useful as it is available for anybody in the space to use by minimizing all open
windows (usually the default state a LiveSpace is left in when it is shut down).
Another feature active desktop provides is the ability to continue to use that machine
during a session. This could include using it as a meta-application display.
Input to Ignite is done using a mouse. A user can walk up to the computer and
navigate the Ignite menu’s as they would use a normal computer. A description of
using Ignite is given in the previous section.
38 Michael J Vernik
interface it is controlled with. As well as this, it provides new functionality to control
basic elements of the ICS including the video/audio matrix switch.
The interface has been upgraded allowing for a touch screen monitor for control,
rather than a mouse. This has been found to be a very responsive and an easy to use
method. To allow for ease of access, Ignite is running on an LG flat panel LCD
display mounted to the rear of the ICS door. The ICS door is a two part ‘barn style’
door and Ignite positioned as such, is one of the first things seen when entering the
room. Figure 3-5 demonstrates where it is positioned in the ICS.
39 Michael J Vernik
Ignite
Figure 3-5. View of the ICS LiveSpace from front to rear of room. Ignite can be
seen on a touchscreen positioned on the door.
3.4 Plexus
Plexus is an application that provides a universal interface for people working in a
LiveSpace. We call this type of application a session interface since it allows people
to interact with each other, the environment, and required information while working
towards a particular objective.
40 Michael J Vernik
such as multiple display screens, projectors, lighting; information (for example files
containing media such as maps, text documents, presentations), and each other.
The Plexus interface has three main parts. These can be seen on Figure 3-6. The
Information Resources contains a predefined set of information that is required for a
particular session. There is no need for users to go looking around the network to find
the necessary files. If additional key information is generated during the session or
participants contribute other information this can be quickly added to the Information
Resources.
The control panels on the left allow for three types of information. Here we can see
the attendee’s category which highlights who is present in the environment. Next we
can see the devices category. This gives control of devices in the environment,
including projectors and lights. This is different from Ignite, as you can control these
devices at your desk with Plexus.
Finally we can see the ‘displays’ category, which represents each of the screens in a
LiveSpaces environment. These displays can be dragged to the 3rd part of Plexus, the
room layout. This gives a graphical representation of the room, so that users can
visually place resources on displays in the environment. For example, a resource
could be dragged from the shared repository to one of the display icons on the room
layout and it will be placed on that display.
To give an example of using plexus, consider the situation where we wanted to view a
PowerPoint file on the centre screen. We would simply drag that file from our
information resources and drop it on the icon that represents the centre screen. The
environment automatically knows what type of file it is and will display the
information with the appropriate application, in this case PowerPoint. Similarly, we
could display other information from our information resources. For example, we
could display a map of the region of interest on the right hand screen and a set of
briefing slides on the left screen by doing this drag and drop procedure.
41 Michael J Vernik
team member. This is done by simply dragging the file to the icon which represents
the device being used by the other team member. Individuals can also view their own
personal copy of information. For example, a person could open their own version of
a presentation being shown on the group display on their own display. In this way
they could look back at what was presented, refer to the presentation notes, or refer to
what might be coming up.
Plexus also allows teams the ability to control their environment in a common way.
The interface to devices such as lights and projectors are easily accessible to each
person in the space. The layout is consistent amongst each users interface, which
helps users change computer and use the functionality they require without being
overwhelmed by a different interface.
Information
Resource
Control Panel
42 Michael J Vernik
Figure 3-6. Plexus displaying the available screens where information can be
dragged.
Alternatively, when users are finished in a session, they will require to take
information created in the session away with them.
A session management prototype has been developed using Microsoft Share Point
portal.
It is comprised of a group of web parts that provide a user interface that is available
corporate wide. This allows participants to add information to the portal before a
session has been initialized.
After a session is complete, information can be dragged from the Plexus information
resource to their desktop, or alternatively saved back to the SharePoint portal.
43 Michael J Vernik
The development of future Session Management applications need to implement the
types of characteristics defined in the Definitions and Formalisms shown in Section
3.1.
44 Michael J Vernik
4 Session Interface Implementation Details
This section covers implementation details on the session interface prototypes
developed for LiveSpaces. An overview of the session interfaces can be found in
Chapter 3.
4.1 Ignite
Ignite has been developed as a web application. In the Intense Collaboration Space, it
resides on a Windows based Apache server running PHP. In the e-World lab, it
resides on a Solaris based Apache server, also with PHP.
The second stage seen in Figure 4-1 is the enterprise connectivity. After a user has
pressed a button, a PHP page is called. This page pre-loads ODSI services if they
have not already been loaded. Following this, a PHP system function call invokes a
java component. This java component contains Elvin connectivity. The reason for
the two layered approach of PHP/java is that Elvin doesn’t have a library for PHP.
The java component establishes an Elvin connection with the local LiveSpace Elvin
server. The Elvin server is federated with other Elvin servers at DSTO, UniSA and
DSTC. This gives the ability to receive the Elvin events from any of the LiveSpaces
available. This is useful when multiple LiveSpaces are used as part of an activity and
participants in one LiveSpace would like to find out which meta-application
participants in a separate LiveSpace have just invoked.
45 Michael J Vernik
are particular attributes of an ODSI peer that manages workflow connectivity. This
peer also handles other parts of enterprise connectivity, such as peer management.
Upon receiving the event, it activates the Breeze workflow and begins to initialize a
new meta-application.
At this point, if a scenario based meta-application is invoked, we may not know the
state of the environment. Lights could be at a certain brightness, or the wrong type of
lights might be switched on. The AV Matrix could be set to display the wrong
machines. This can bring unexpected behaviour to the meta-application. Ignite helps
to overcome this by running a wrapper workflow around the meta-application. This
sets the right lighting conditions and display conditions. As well as this, it can help
return the LiveSpace to its previous state when a meta-application has finished. After
this wrapper workflow has completed, the meta-application workflow can begin.
Displays are then populated and the virtual assistant Kayla will begin to present. A
description of meta-applications is given in more depth in section 2.3.2.
Another feature of Ignite is the health status indicator. This has been implemented as
a Java applet. The Java applet is embedded inside the PHP front page of Ignite below
the start/shutdown/custom buttons. Each element that displays a status for a particular
service is subscribed to an event from a particular iCrafter service. If the service no
longer functions, the iCrafter persistent event that it previously had on the EventHeap
will no longer be available. This will cause the icon to display red or orange,
depending if it is a critical service.
Initialization of the health status is done using a configuration file. This file has
information relating to:
Service to track – the iCrafter name of service so that events can be tracked.
46 Michael J Vernik
Whether the service is critical or non-critical – used for the display colour if a
service fails. If the service is critical to the space operating (such as the room
infrastructure), it will be red. If the service is non-critical (such as the
workspace simulator), it will display in orange.
Invocation
Meta-application button
selected from Ignite!
Enterprise Connectivity
Java application
called to produce
Elvin events
ODSI peer
receives Elvin
event
Meta-application
Breeze workflow
engine, residing on
ODSI peer begins
particular workflow
Coordination
Subscribed iCrafter Pre-positing data in Control of
service automatically applications. e.g Audio/Visual
controls device. Plexus infrastructure
through iCrafter
47 Michael J Vernik
4.2 Plexus
The original concepts and design of Plexus were developed by myself and Vivian
Ngyuen and Professor Rudi Vernik in early 2003 as part of a summer internship.
Vivian and Aaron Stafford subsequently implemented an initial working prototype of
Plexus as part of their software engineering honours project. I have subsequently
generated an updated version of Plexus to be used in the evaluations which have been
undertaken in this thesis project.
Plexus is implemented on the Java platform. Java was chosen as iROS (Interactive
Room Operating System) provides a stable, well supported API in Java. Also, Java
can be run on multiple operating systems. This makes it useful in a LiveSpace
because it can be run on different machines such as MacOS or Linux. This also
provides the ability to run it as a web applet, making it accessible without having to
install it on individual machines.
Plexus makes heavy use of a component of iROS called the EventHeap. Figure 4-2
shows a description of how the EventHeap can be used. The EventHeap is akin to a
blackboard where messages can be placed and accessed by different devices and all
applications. These messages are structured using XML metadata. Clients utilizing
the EventHeap (such as Plexus) are able to subscribe to any type of message with a
particular piece of metadata assigned to it. These clients are also able to publish
events so that other subscribed clients can receive them. Figure 4-2 highlights an
example of using the eventheap.
48 Michael J Vernik
Subscribed Events
LightController 1
3
AVMatrix
EventHeap
Plexus
Projector
Figure 4-3 highlights some of the communication pathways that Plexus utilizes.
When Plexus starts, it creates an event and places it on the EventHeap to beacon other
instances of Plexus that it has just started and joined the session. This event is created
with specific metadata associated with Plexus. The event is persistent and will be
available for other instances of plexus to view at anytime.
Figure 4-3 also highlights the communication methods utilized by the information
resource (as discussed in Section 3-4). When an item is dragged to the information
resource, it communicates through the EventHeap, to the DataHeap. The DataHeap is
a more persistent version of the EventHeap that utilizes a webDAV directory structure
to keep information until it is physically deleted (unlike the EventHeap which loses
data when restarted). The information that is dragged to is placed on the DataHeap
49 Michael J Vernik
with a particular type of metadata. All other subscribing plexus clients will update to
display the contents of the DataHeap.
Plexus
EventHeap WebDAV
Dataheap
iCrafter
Another function of Plexus is the Universal Display Service (UDS). This service
allows a file from the Shared Resource to be opened up on another Plexus client.
When a user drags a file to an icon representing a display, the display’s UDS will
match the type of file to the required application. The application will load with the
correct file present.
The implementation of the UDS extends an iCrafter service. This is useful as iCrafter
helps coordinate services in a LiveSpace. iCrafter is explained in further depth in
2.2.2.1. Its implementation involves an event being sent through the EventHeap to
the ‘receiving’ or ‘to be displayed’ machine. Upon receiving the event, it will fetch
the appropriate file from the DataHeap. The file is stored in a “temp” directory on the
local computer’s hard disk. Opening the application with the required file is done
using a Python script. This script matches the native operating system MIME type
and opens up the appropriate application.
50 Michael J Vernik
During the past 12 months the Plexus prototype has been further developed so that it
could be deployed and used on a day to day basis. It is particularly important to have
a robust implementation for evaluation Purposes.
Plexus was also adapted so that it could run in various LiveSpaces. Contributions to
this included work from James Brennan. His work involved getting Plexus to
function in the Mobile LiveSpace at DSTC Brisbane.
The DataHeap component of iROS was upgraded to aid the performance of the
information resource. Previously it was quite slow to transfer files to and from the
information resource, as well as placing them on other displays. This could
sometimes take up to 5 minutes depending on the size of the file. The upgrade
increased performance significantly.
The SharePoint to Plexus bridge that is highlighted in Figure 4-4 is another key
component of the Session Manager. This component has been developed so that the
shared information resources in Plexus can be populated when a session is started.
Current implementation uses a batch file that maps a SharePoint document repository.
It does this using the SharePoint WebDAV functionality SharePoint provides. Once
the batch file has done this, it instantiates a Java application with a number of
arguments. These give it the location of the document repository. The application
then recursively accesses each document in the repository so that it can get all
documents located in the directories below it. As each document is accessed, they are
placed in the DataHeap. The metadata assigned to each file is in the same form that
Plexus utilizes in its document repository. The files can then be accessed from
Plexus.
51 Michael J Vernik
Plexus SharePoint
Portal
SharePoint to
Plexus Bridge
EventHeap WebDAV
Dataheap
iCrafter
52 Michael J Vernik
5 Evaluation
This chapter discusses the series of evaluations that have been undertaken of the
LiveSpaces session interfaces discussed in this thesis. The chapter begins by
providing an overview of various approaches that have been proposed and used to
evaluate interactive applications. These include a range of discount methods such as
cognitive walkthroughs and heuristic evaluations as well as the more recently
proposed Collaboration Usability Analysis approach [PIN02]. This is followed by a
discussion of the new orchestrated evaluation approach that is being trialled as part of
the AUSPLANS project. The session interfaces evaluations were undertaken using
this methodology. Section 5.3 discusses how the approach was applied for these
session interfaces evaluations. This is followed by a presentation and discussion of
the results of the evaluations in Section 5.3.
53 Michael J Vernik
problems in interfaces. The results of the evaluations are transformed into a cohesive
set of problem reports through a process called results synthesis.
Most discount methods suffer from limitations such as the dissociation from real work
settings and a lack of a theoretical basis. Also, these approaches are typically strongly
oriented around individual work. The contextual information they provide and
usability criteria are focused on tasks and actions that a user must carry out to achieve
a goal. These aspects are often defined as taskwork [PIN02]. Methods are needed
that focus on teamwork such as the actions needed to complete the task as a group.
Variations of traditional discount evaluation methods have been proposed for group
work such as the Collaboration Usability Analysis approach discussed in Section
5.1.2.
Since discount methods are typically used at the early stages of development they do
not generally address the social or organisational perspectives of employing new
concepts and technologies. These types of issues need to be investigated with actual
users in real work environments. The AUSPLANS Orchestrated Evaluation
Approach discussed in Section 5.2 provides an approach which combines the benefits
of discount methods while maintaining some of the benefits of field studies and
laboratory experiments such as using actual users in realistic situations and supporting
replication of evaluations.
54 Michael J Vernik
5.1.2 Collaboration Usability Analysis
Collaboration Usability Analysis (CUA) [PIN02] is a new task analysis technique
designed to represent collaboration in shared tasks in support of usability evaluations.
The approach has been developed to provide evaluators with a framework for
simulating the realistic use of a groupware system and to identify user interface
problems [PIN02]. The approach focuses on teamwork in group tasks.
CUA is based on the use of a hierarchical task model to represent the procedural
elements of a group task in a shared workspace. The task hierarchy includes
scenarios to describe the high-level context of the collaborative situation, tasks to
indicate specific goals within the scenario, and low-level task instantiations to
represent individual and collaborative ways of carrying out the task.
The AUSPLANS Project [EVD04] aims to apply and evaluate the use of enterprise-
enable ubiquitous workspaces (or LiveSpaces) for supporting multiple teams engaged
in military planning activities. As discussed in 2.4, the approach being adopted for
AUSPLANS centres on the use of an evolving experimental platform, based on
LiveSpaces infrastructure and technologies, to support prototyping, evaluation and
55 Michael J Vernik
transitioning activities through a series of research cycle increments, each culminating
in a Technical Exercise (TE) involving stakeholders and users.
The experimentation and evaluation approach that has been developed for
AUSPLANS is based on the use of Orchestrated Evaluation Sessions. The concept of
orchestrated evaluation is to provide a means of evaluating new concepts and
technology within a domain specific scenario enacted in a representative work
environment populated with relevant tools and artifacts [EVD04]. Artifacts can
include physical products such as paper maps through to digital artifacts such as
imagery, geospatial data, video footage, and planning data. The orchestrated session is
a framework which allows evaluation modules to be inserted at particular points in a
work process to assess the effectiveness of particular techniques and technologies
relative to users’ work practices.
Activity. This dimension defines the setting for the work being undertaken by
a set of one or more teams in fulfilling a particular objective. The elements of
56 Michael J Vernik
work are defined in terms of a set of activities and tasks that are performed by
team members using specified inputs to produce desired results within a
particular time. In AUSPLANS the activities relate to large-scale planning
situations and include major phases of activity such as Situation Assessment,
Mission Analysis, and Course of Action Development. Each of these
activities can be broken down into lower-level activities and a set of basic
units of work called tasks. Some tasks will be specific to the activity and
others will be standard types of tasks such as questioning, presenting, and
brainstorming.
Mechanics of Ubiquitous Group Collaboration. This dimension defines the
underlying aspects of human cognition and interaction that need to be
supported in accomplishing particular activities and tasks. The initial set of
mechanics developed by the AUSPLANS project are shown in Appendix B
and include categories such as communication, awareness, understanding,
remembering and reasoning. The work of Pinnelle et al (2003) was used as the
basis for the initial set of mechanics. Although tasks and activities might
change based on a particular work scenario, the AUSPLANS researchers
believe that the underlying primitives, collectively called mechanics, remain
constant and hence provide the stable dimension for evaluating the
effectiveness of particular capabilities in relation to the work context or
activities. It is against this dimension that qualitative and quantitative
measures are evaluated. The term ubiquitous is used to reflect the fact that
AUSPLANS research is primarily considering the mechanics of human
collaborative activities being undertaken in ubiquitous workspaces. This is an
important distinction in that as these environments become a natural extension
of human abilities, they largely bridging the cognitive gulfs between humans
and machines as defined by Norman [NOR86].
57 Michael J Vernik
the capabilities that are being evaluated. The AUSPLANS researchers believe
that technologies in themselves do not necessarily provide people with a
capability for supporting group collaborative activities in achieving desired
goals. Knowledge of organizational structures, policies, technology
adaptation and processes are important contributors as is expertise such as is
often captured in particular work practices.
Mechanics of Group
Collaboration
Evaluation
Activity
Capability
Evaluation is undertaken within the context of this framework. For example, the
evaluation of a specific capability is undertaken within the context of particular
work activities and tasks. The capability is evaluated in relation to its effect on
the underlying mechanics of group collaboration. A key focus of this work is on
the evaluation of capability rather than an individual technology. Capability
might include a set of one or more technologies, knowledge associated with
adaptation of technology for a particular use (e.g. this may be encoded in the
technology as in the case of meta-applications), and the work practices necessary
for achieving desired results.
58 Michael J Vernik
Evaluation is one particular use of the framework. Others include identifying
areas where additional research and development is needed to provide capabilities
which more effectively support collaborative work activities, such as in the
definition and validation of requirements. The framework has also been designed
to help understand the value of integrating particular off-the-shelf tools as part of
an overall tool suite, and in supporting innovation though the capturing of
innovative combinations of technologies, work practices, experience and
knowledge.
5.2.2 Methodology
The evaluation methodology in AUSPLANS is based on three cycles of evaluation
activities called Technical Exercises (TEs). The first series was conducted between
February and September 2004. The second series commenced in September 2004 and
will conclude in September 2005. The third series will run from September 2005
until April 2006. The methodology is being evaluated and evolved over these three
cycles.
59 Michael J Vernik
Figure 5-2. TechEx 1.1 conducted at UniSA LiveSpace
Actual exercise data and media were made available so to support the development of
the orchestrated evaluation sessions. So that realistic geospatial data and location
could be provided, the geography of Southern Australasia was used as the exercise
location with the island of Tindaro represented in the geography of Tasmania (see
Figure 5-3). An orchestrated evaluation meta applications was developed to help
automate the evaluation scenario and process. A realistic context for the planning
activities was established through the use of automated briefings coordinated by a
virtual workspace assistant called Kayla. Kayla uses speech synthesis to coordinate
activities and for providing awareness for the participants. The process used for the
TE1 series was to: (1) set the context for the activity through multi media presentation
and simulation, (2) automatically describe the candidate capabilities to be evaluated
within the context of the activity being performed, (3) set up the capabilities for
evaluation (4) enter into and evaluation segment by having the participants interact
with particular technologies, discuss ways in which individual or combinations of
technologies could be used to support collaborative group work within the context of
the activity being performed, and (4) have the participants complete a questionnaire
were their responses were recorded relative to the dimensions of the Orchestrated
Evaluation Framework. The TE1 series focused more heavily on the “front end”
60 Michael J Vernik
activities following the communication of a warning order from the high-level
command indicating the government intention to provide support to South Tindoro.
This phase involved pre-planning activities resulting in the establishment of the
various planning teams, information gathering, and situational awareness. Various
capabilities were evaluated including session interfaces, ticker-tape interfaces for
alerting and conveying the warning order to headquarters participants, InVision
visualisation capabilities, and ubiquitous workspaces applications such as PointRight.
TE1.2 replicated the evaluation undertaken in TE1.1 but this time the participants
were the senior joint planning staff from the headquarters including the head of
planning (J5), the leader of the Joint Effects Planning (J5Jet), and a United States
exchange officer who was part of the planning branch (J53)
NEW
ZEALAND
NORTH
TINDORO
RABENNESTE Serves as ISB
SOUTH
TINDORO
Area of Operations
SCENARIO GEOGRAPHY
USCINCPAC
USCINCPAC UNCLASSIFIED JWID
JWID TASK
TASK FORCE
FORCE
4
61 Michael J Vernik
the participants could innovate and apply the capabilities at will. The approach that
has been adopted involves the following sessions: (1) An introductory session where
the participants learn about the LiveSpaces approach, the AUSPLANS project and the
objectives of the evaluations to be undertaken. (2) An activities mapping session
which validates the activity model for the activities to be undertaken for the
evaluation and a mapping of the Activities to the Mechanics. This mapping helps
define the requirements of the participants without reference to particular technologies
or capabilities. (3) Introduction to the scenario context for the evaluations. This is
provided by way of an automated briefing, similar to that which would be provided as
an initial awareness of the situation following the warning order. This helps place the
participants into the context of the problem. The participants are also introduced to the
information available for their planning activity including digital maps, databases
showing available resources, intelligence information, etc. (4) Capability evaluations
which introduce the various capabilities to be evaluated. These capabilities are
grouped into evaluation segments such as Physical Environment and Media
Translation, LiveSpaces Core Applications, Automation and Invocation, Enterprise
Enablement, and Advanced Command and Control Applications. The participants are
introduced to the capabilities, they then have a period of free play where they can use
the capabilities to explore how they might use them in innovative ways to aid their
activities, and they then complete a formal evaluation of selected capabilities. (5) The
participants undertake an actual planning activity which is called the Scenario
Exercise. In TE2.1 the participants completed the Mission Analysis activity. A series
of meta applications were used to automate the phases leading to the activity that the
team was required to complete. These included providing awareness briefs such as
the Intelligence Preparation of the BattleSpace, mission briefings, and pre-planning
aspects. This set the context from which they could undertake the Mission Analysis
activity. (6) A workshop involving participants and scientists constitutes the final
session where the results are consolidated and discussed.
The scenario for the TE2 series is a disaster relief operation called Tempest Express.
This scenario was chosen for TE2 in that it provides a richer set of artefacts than the
scenario used in the TE1 series, and it is one of the actual exercises that has been
62 Michael J Vernik
conducted by the headquarters to train new planning teams. As such, the participants
were familiar with the context of the planning activity.
Evaluate the use of a universal session interface such as Ignite against the
Mechanics for Ubiquitous Group Collaboration for initializing a
LiveSpace and for invoking pre-defined sessions.
63 Michael J Vernik
Understand how session interfaces work together with each other and other
ubiquitous workspace technologies to provide integrated capabilities.
5.3.2 Participants
The participants for TE1.2 and TE2.1 are shown in Table 5-1. We categorise the
participants into three groupings: Planning Staff, Researchers, and Observers. We
define the Planning Staff as the Focus Group in that they are the ones who will
ultimately be the recipients of the capabilities being evaluated. The Control Group is
made up of the researchers and observers. The distinction between these groups is
made because: (a) the Control group have typically been involved in the development
of the capabilities being evaluated and so may prejudice the evaluation results if their
questionnaire responses were combined with those of the Focus Group. (b) The
Control Group, as a population, is generally more focused on and familiar with
technology and hence provide an interesting contrast to the responses provided by the
Focus Group.
Participants
Planning Staff Researchers Observers
Tech Ex 1.2 Tech Ex 2.1 Tech Ex 1.2 Tech Ex 2.1 Tech Ex 1.2 Tech Ex 2.1
MP (DSTO) MP (DSTO
MV (UniSA) MV (UniSA)
JB (DSTC) DK (DSTC)
SJ (UniSA) TR (DSTC)
TB (UniSA)
EK (DSTC)
Table 5-1. Participants during TechEx 1.2 and TecEx 2.1. Note: participants in grey
were present at TechEx 1.2 and 2.1.
64 Michael J Vernik
Table 5-1 shows that some of the participants were common in both exercises as
highlighted in grey. This provides an interesting dimension to be analysed in terms of
the results.
5.3.3 Methodology
The methodology used for the session interfaces evaluations was the AUSPLANS
Orchestrated Evaluation Approach discussed in Section 6.2.2. Variations in the
methodology between TE1.2 and TE2.1 are discussed in Sections 6.2.2.1 and 6.2.2.2.
65 Michael J Vernik
5.3.5 Ethics Considerations
Ethics considerations for the evaluations being undertaken as part of the TE series are
addressed by the AUSPLANS project. Since AUSPLANS is a DSTC project, the
research team complies with the CRC participant ethics requirements as detailed by
DSTC, DSTO, and UNiSA.
“Using Ignite, is the ability to see the health status of a LiveSpace environment
useful?”
One of the reasons for the difference in response might be that the Control Group are
more appreciative of the complexity of the technology in a LiveSpace and the
potential for it to be unreliable. The control group would like to see visual indicators
66 Michael J Vernik
highlighting what they can and cannot use in the LiveSpace. The Focus Group are
probably less aware of the impact that technology malfunction might have in a
LiveSpace. They appear to have embraced the concept that the technologies should
be ubiquitous and available to serve their needs. This question highlights the
requirement for high reliability of infrastructure in a LiveSpace.
Ignite Questionnaire
12
Question 1
10 Question 2
Question 3
8 Question 4
Score (0-10)
Question 5
6 Question 6
Question 7
4 Question 8
Question 9
2 Question 10
Question 11
0 Question 12
J53 J54 J5 RANCO Focus Group Control Average Question 13
Average Question 14
Group Name
Figure 5-5 provides a graph of the results of the questions related to the coordination
mechanic. Again, there is a generally positive response, however there are two
significant deviations from the norm in Question 10 and Question 13. Question 10
states:
“Would you prefer to control and set the configuration of Livespace devices (such as
lighting, audio vidual matrix) using the Ignite touch screen, or from an interface
available to each user at their desk .”
The result for this question (an average of 4) suggests that many users would like to
have control from both session interfaces. This highlights an interesting issue that
needs to be discussed in terms of the relationship between various session interfaces.
For example, do people prefer to have a single interface which supports all their needs
67 Michael J Vernik
or do they prefer to have different applications that serve different goals? This is
discussed further in Section 6.
12
10
8 J53
J54
Score (0-10)
J5
6
RANCO
Focus Group Average
4
Control Average
0
1 2 3 4 5 6 10 13 14
Question Number
There were a number of comments made on the Ignite questionnaire that raises some
interesting questions. From the Focus Group we had:
“Very useful for setting up or configurations for planning and briefing sessions.”
The first comment shows that the Focus Group member understands the importance
of being able to rapidly set up the LiveSpace for particular activities. Recorded
discussions also highlight comments that amplify this observation. Howerver, the
second comment appears in conflict with this view. This is possibly a reflection of
the outcome-oriented nature of miliary planning staff. They are specialists and their
main aim is to produce an effective plan for a situation and hence their focus is on the
tasks that they need to undertake to do this, including reasoning, recording, and
68 Michael J Vernik
communicating. Ignite is a supporting application that enables what they do. In their
traditional environment, they would generally have other staff who would set up the
environment for them and fix any problems.
Similar results are found in TE 2.1. Figure 5-7 shows the overall responses of
participants in this technical exercise. It shows the general trend of scores of the
Focus Group to be greater than an average of 7.5. Interestingly, the average score for
the control group is considerably lower than the focus group. These are discussed
further in Section 5.3.2.
69 Michael J Vernik
TE1.1 Plexus Overall Response
12
Question 2
Question 3
Question 4
10 Question 5
Question 6
Question 7
8 Question 8
Question 9
Question 10
Score
6 Question 11
Question 12
Question 13
Question 14
4 Question 15
Question 16
Question 17
2 Question 18
Question 19
Question 20
0 Question 21
12
Qn1
10 Qn2
Qn3
8 Qn4
Score (0-10)
Qn5
Qn6
6
Qn7
Qn8
4 Qn9
Qn10
2 Qn11
Qn12
Qn13
0
Qn15
J53 J54 J5 RANCO Focus Group Control
Average Average Qn16
Group Name
70 Michael J Vernik
Looking at the overall response graph from TechEx 1.1, the group of questions at the
start and end of the questionnaire seemed to have the highest positive responses. This
is shown in the ‘J Average’ section. If we compare these questions to our mechanic
coverage matrix (Appendix B), we can see that these are questions related to the high-
level mechanics ‘communication’ and ‘coordination’ in a LiveSpace. Conversely, the
questions in the middle section appear to have a lower average response. In
particular, these are the magenta, yellow and cyan bars that can be seen. These
represent questions 11, 12 and 13; or questions related to reasoning. The Plexus
interface was not specifically designed to support reasoning, however it seems that,
since group reasoning is a complex undertaking involving other mechanics such as
communication, there was an indication that Plexus provided support for this area.
A closer view of the two mechanics, communication and coordination highlight some
interesting results. Figures 5-8 is the graph related to the questions involving the
communication mechanics in TE1.2.
Mechanic: Communication
12
10
8 J5
Score (0-10.5)
J5 JET
6 J55
Control group
4 J Average
0
Q15 Q16 Q18 Q21
Question Number
71 Michael J Vernik
The questions relating to the communication mechanic (TE 1.2 questionnaire
Questions 15, 16, 18 and 21) show a very high level of acceptance. The only
exception to this is Question 16 where 2 of the 3 respondents did not show a highly
positive response. Question 16 states:
The Coverage Matrix (Appendix B) shows that Plexus was developed with features
that might aid person to person communications. In Plexus, this can be achieved by
allowing users to drag a document to another person’s workspace (i.e. personal
laptop). In this way, information can be directly pushed to another participant in a
LiveSpace. A reason why members of the Focus Group did not see that Plexus could
be useful for this type of communication could have been a lack of understanding of
Plexus features, or the that it only provides a rudimentary way of supporting person to
person communications. They may well have thought that this approach could be
intrusive. Further investigation is required to resolve this issue.
72 Michael J Vernik
TE2.1 Plexus - Coordination
12
10
8 J53
J54
Score (0-10)
J5
6
RANCO
Focus Group Average
4 Control Average
0
2 3 5 11 15 16
Question number
Figure 5-9. Plexus Results for Coordination Mechanic questions in TechEx 2.1
The Graph of Figure 5-9 shows the results of the TE2.1 Plexus questions related to the
coordination mechanics. On this graph, there are two lines that represent the averages
of (a) the Focus Group (b) the Control Group. The averages for the control group is
consistently lower. However, in Question 11 both groups see this aspect as equally
important . Question 11 asks:
“Is the ability to control and configure devices and lights, projectors from an
individual desktop important?”
The contol group which is more technically oriented seems to have a favour those
questions that refer to control of technologies.
The following two questions, 15 and 16 also show an interesting trend. These
questions heavily relate to headquarters information and state:
73 Michael J Vernik
“15. Using Plexus, do you see viewing headquarters wide information resources
useful?”
“16. Using Plexus, do you see storing headquarters wide information resources
useful?”
One of the reasons for the observed results is that the Control Group might be looking
at Plexus in terms of its current implementation rather than the capability that is
needed (more the perspective of the Focus Group). For example, they understand
how technologies such as Microsoft SharePoint portal could support this aspect of
coordination. The Focus Group is typically looking at it from a future functionality
perspective. They see it would be useful to place information that is available across
the headquarters easily onto displays through a single interface. Alternatively, they
would be able to hand-off a document to somebody else in the headquarters using
Plexus (rather than taking another step such as emailing). This highlights the focus
groups particular need for a universal interface.
There were a number of comments made about Plexus in the TechEx 2.1
questionnaires. Questions from the Focus Group included:
“Potential problem with revision control with copies of actual documents being
passed between computers.”
74 Michael J Vernik
There are some interesting tradeoffs that need to be considered in this respect such as
the transportability of LiveSpace application to other implementations, and the
common “look and feel” that is so important in heterogeneous computing
environments.. The final question relates heavily to the Focus Group’s question
regarding version control. Given that both groups see this as important, it is an area
that warrants further investigation.
5.4.3 PointRight
PointRight is a technology that was only formally evaluated in TechEx 2.1. The
results discussed here are in the type that the capability provided by PointRight is
important to the effective use session interfaces. Figure 5-10, shows the overall
responses to the evaluation questions. The ‘Focus Group Average’ and ‘Control
Group Average’ columns show that the results are generally positive. There is an
unusual exception to this. The cream coloured bar shows a very low score for the
Focus Group and a very high score for the control group for Question 3. It states:
“3. Do you find the notion of moving your mouse cursor to the top of your local
display to control another display intuitive?”
It seems that the Focus Group did not find this method of interaction intuitive. As
explained in Section 5.3.2 some of the Focus Group had not previously had
experienced using the technology (J54 and RANCO). These members recorded a
strong negative to the question. It is interesting to note that those who were more
familiar with the environment (e.g. Control Group, J5 and J53) reported that the
approach was intuitive. The results suggest that the approach may not be entirely
intuitive in that it is only after learning and experience that people feel comfortable
with the approach. Unfortunately, the PointRight evaluation was not undertaken in
TE1.2 when J5 and J53 were new to the LiveSpaces applications. This would have
shown whether this observation is in fact correct.
The spread of other results shows that they found the other interaction methods of
PointRight to be very easy to use.
75 Michael J Vernik
TE2.1 PointRight Questionnaire
12
10
1
2
8
3
Score (0-10)
4
6
5
6
4
7
8
2
0
J53 J54 J5 RANCO Focus Group Control Group
Average Average
Group Name
Figure 5-10. PointRigh Results for Coordination Mechanic questions in TechEx 2.1
Figure 5-11 shows a graph with all questions from the PointRight questionnaire. It
has not been broken into individual mechanics as the previous technologies. The
reason for this is many of the questions from the questionnaire are from the
coordination mechanic category. A graph with a single piece of data would not be
very interesting to compare.
The focus group and control group average for this graph shows a very high
acceptance for each of the questions. The two exceptions are question 3 and question
6. Question 3 has been covered previously. Question 6 however, is a question
relating to coordination. It states:
“6. Do you find the ability of having multiple users control the same display useful?”
The result says that the Focus Group would not find it useful to have control of
displays simultaneously. As section 5.2.2.2 states, the Focus Group was available to
demonstrate a scenario to the research group after this evaluation took place. Upon
viewing this, it is evident that their work practices would not be considerably affected
if more than one person could use the display. Each member of the Focus Group has
their own designated task. An individual is usually controlling the different displays
at one time.
76 Michael J Vernik
TE 2.1 PointRight
12
10
8 J53
J54
Score (1-10)
J5
6
RANCO
Focus Group Average
4
Control Group Average
0
1 2 3 4 5 6 7 8
Que stion Number
“Control multi displays in a planning session; e.g. excel, word and PPT shows.”
and
“Assign priority to the computers (their users) that allows a user to override another
user(s).”
Collating these comments and information from the graph, it is indicated that the
Focus Group users would like a system that one person is able to use, however can
access many of the displays in the environment. They also believe it would be ok if
they took turns controlling machines, as long as they can over-ride other users from
using them. PointRight provides some of the functionality already, however would
require modification to add features such as override another user.
77 Michael J Vernik
6 Discussion and Future Work
This chapter provides a summary and discussion of the results of this thesis project.
The chapter begins by summarizing the aims of the work. This is followed by a
discussion of the findings. Section 6.3 then highlights the limitations of the research.
The chapter concludes by suggesting areas of future work.
78 Michael J Vernik
6.2.1 Theoretical Basis for Session Interfaces
Section 3.1.1 provides an initial formal definition of a session as the being the
underlying unit of activity being undertaken within and between ubiquitous
workspaces. A working definition was proposed as:
A systems modeling approach and notation based on the approach de veloped by Koposi
and Myer (1994) was used to provide a more formal definition of session. The intention
was to provide a method to support the discussion, capturing and reasoning about
session concepts. In addition, formal definitions may aid the development of
improved session interfaces through the better definition and validation of requirements.
79 Michael J Vernik
The approach taken in TE2.1 was much better than TE1.1 in that an approach based
on capability groupings was used focus more on the capabilities than the scenarios.
This provided users with a much better understanding of the capabilities being
evaluated. To improve results and feedback from TE’s, it would be beneficial to
provide an additional questionnaire following the scenario exercise. This would allow
the Focus Group to re-evaluate their views following their use of the capabilities in an
actual planning activity.
80 Michael J Vernik
dragging and dropping individual files from the Plexus Information Resource. There
is a requirement to have additional facility within Plexus to support the rapid access
and display of information.
The need for a walk-up interface to provide initialisation and shut down of the
environment was seen as an important requirement. Users entering a LiveSpace were
able to consistently locate Ignite and commence a session. This was particularly
evident from TE1.2 to TE2.1 where Ignite had been upgraded. Rather than
81 Michael J Vernik
participants being confused on how to continue from an evaluation discussion point
(as with TE1.2), a simple point of reference, Ignite was used.
6.3 Limitations
The evaluative work discussed in this thesis is an initial attempt at developing
approaches for evaluating capabilities within ubiquitous workspaces. A large amount
of data was captured during the TE1.2 and TE2.1 session interfaces evaluations, both
in terms of the questionnaires and the video/audio capture. Only a partial analysis of
this data was possible within the timeframe allocated for this honours project.
82 Michael J Vernik
enterprise enablement and connectivity. The following paragraphs discuss some of
these aspects in more detail.
The current LiveSpace health status indication on Ignite is an early prototype and
needs further investigation. A new visualization technique to view the status would
aid considerably in its usefulness. For example, if a technology is not functioning in
the environment, there should always be a clear indication of a problem and its
potential impact on the activities to be performed. This representation should allow
the user to quickly gain additional information, for example by selecting the ‘failed’
icon and drilling down for more information. This might include information on who
to contact to get the technology fixed, suggestions for fixing the technology, or ways
to fix the technology such as a reset button. Elvin Tickertape could be used to supply
this additional information.
83 Michael J Vernik
representation of their computer to a display represented in the environment, thereby
switching the displays. Investigations need to be undertaken to look at the adaptive
nature of such spatial interfaces.
84 Michael J Vernik
7 Conclusions
This thesis project has undertaken research into session interfaces for ubiquitous
workspaces. Ubiquitous workspaces are future media-rich environments that employ
new forms of operating systems and services to coordinate and manage interactions
between people, multiple display surfaces, information, personal devices, and
workspace applications. A class of application which will be of critical importance to
the success of these teams operating in these future work environments will be session
interfaces. These applications will provide a common interface through which users
can control and interact with the host of devices, services, media and applications, and
information available in the workspace. They are aimed at reducing the cognitive
load required by users in understanding and interacting through a vast array of
application and device interfaces. They will supports team collaboration by providing
ready access to required information and by providing access to a range of workspace
knowledge and support services such as instant messaging, speech transcription and
context awareness. Session interfaces have the potential to make future ubiquitous
workspaces accessible and usable by a broad range of users, much as WIMP
interfaces have made personal computers accessible to the general public.
The key objective of the thesis project was to investigate the impact that these new
forms of interfaces might have on the effectiveness of teams in ubiquitous
workspaces. The specific aims of research were to develop a theoretical basis for
session interfaces and session management by defining underlying concepts,
requirements, and models; define and implement a set of prototype session interfaces
to support the evaluation of novel concepts in relation to particular requirements such
as rapid set up and initialisation of sessions, workspace control, team interaction, and
ambient awareness; develop new concepts and frameworks for session management.;
and to undertake evaluation activities using the prototype implementations as part of
the DSTC AUSPLANS Project .
Several contributions have resulted from the research reported in this thesis. The area
of session interfaces has been scoped through extensive literature review. Initial
definitions and formalism that provide the foundations for future research in this area
85 Michael J Vernik
have been developed. Prototype session interfaces have been developed and
integrated into two LiveSpaces environments. Initial evaluations have been conducted
as part of the AUSPLANS project. The evaluation results discussed in this thesis
suggest that there is merit in continuing research into universal interfaces for
ubiquitous workspaces. These are preliminary results. Far more work is needed to
develop the underlying theoretical foundations for this area of research, prototype
implementations of future interfaces, approaches for setting up and managing
sessions, and methods for undertaking evaluations.
86 Michael J Vernik
8 References
[ACT] Active Spaces Project, http://www.cs.uiuc.edu/news/archives/resactsp.html,
last accessed May 2003
[BAL03a] Ballagas, R., et al. (2003) iStuff: A Physical User Interface Toolkit for
Ubiquitous Computing Environments. in CHI 2003: Human Factors in Computing
Systems. Fort Lauderdale, FL, USA: Association for Computing Machinery.
[BAL03b] Ballagas R., Szybalski A., Fox A., (2003) “Patch Panel: A General Model
for Interoperability in Incrementally Deployed Unbicomp Environments”, under
review for Unbicomp 03, basis for discussions between R. Vernik and Interactive
Workspaces team May 2003.
[CER01] Cerqueira, R., et al. (2001) Gaia: A Development Infrastructure for Active
Spaces. in Ubitools Workshop at Ubicomp 2001. Atlanta, GA.
[CHI00] Patrick Chiu, Ashutosh Kapuskar, Sarah Reitmeier, and Lynn Wilcox,
“Room with a Rear View: Meeting Capture in a Multimedia Conference Room.” In
IEEE Multimedia Magazine, vol. 7, no. 4, Oct-Dec 2000, pp. 48-54., October 1, 2000
[COE99] Coen, M., Phillips, B., Warshawsky, N., Weisman, L., Peters, S. and Finin,
P. (1999). Meeting the Computational Needs of Intelligent Environments: The
MetaGlue System. MANSE'99, Dublin, Ireland.
[COV98] Covi, L.M., et al. A room of your own: what do we learn about support of
teamwork from assessing teams in dedicated project rooms? in Cooperative Buildings
Integrating
Information, Organization, and Architecture First International Workshop CoBuild'98
Proceedings. 1998. Darmstadt, Germany: Berlin, Germany : Springer-Verlag, 1998
87 Michael J Vernik
[DSTO] Defence Science and Technology Organization,
http://www.dsto.defence.gov.au/, last accessed June 2004.
[EVD04] Evdokiou, P., Thomas, B., Vernik R.J. (2004) “Augmented Sysnchronised
Planning Spaces” submitted to in 9th International Command and Control Research
and Technology Symposium, Denmark.
[GAJ01] Gajos.K. (2001) Rascal - a Resource Manager For Multi Agent Systems In
Smart Spaces. Proceedings of CEEMAS’01, Kraków, Poland, 2001.
[JOH01] Johanson, B., Ponnekanti, S., Sengupta, C., Fox, A., "Multibrowsing:
Moving Web Content Across Multiple Displays," Proceedings of Ubicomp 2001,
September 30-October 2, 2001.
[JOH02] Johanson, B., Hutchins, G., Winograd, T., Stone, M (2002) “PointRight:
Experience with Flexible Input Redirection in Interactive Workspaces“ In Proc. UIST
'02
[KAT99] Kato, H. and M. Billinghurst. Marker Tracking and HMD Calibration for a
Video-based Augmented Reality Conferencing System. in 2nd IEEE and ACM
International Workshop on Augmented Reality. 1999. San Francisco USA.
[LEW90] Lewis, C., Poison, P., Wharton, C., and Rieman, J. (1990). Testing a
walkthrough methodology for theorybased design of walk-up-and-use interfaces.
Proc. ACM CHI’ 90 Conf. (Seattle, WA, April 1-5), 235-242.
88 Michael J Vernik
[LYY02] Lyytinen K., Yoo Y., Issues and Challenges in Ubiquitous Computing,
Communications of the ACM, Vol 45 No1, December 2002.
[MAR02b] Mark, G., Collaborative Design Within and Between Warrooms. Human-
Computer Interaction Journal, 2002.
[NIE94] Nielsen, J., and Mack, R. L. (Eds.) (1994), Usability Inspection Methods,
John Wiley & Sons, New York.
[OH02] Alice Oh, Harold Fox, Max Van Kleek, Aaron Adler, Krzysztof Gajos,
Louis-Philippe Morency, and Trevor Darrell. Evaluating Look-to-Talk: A Gaze-
Aware Interface in a Collaborative Environment. In Proceedings of CHI 2002.
Minneapolis, MN, April 2002.
[PIN02] Pinelle, D., Gutwin, C. and Greenberg, S. (2002) Task Analysis for
Groupware Usability Evaluation: Modeling Shared-Workspace Tasks with the
Mechanics of Collaboration. Accepted for publications in ACM Transactions on
Human Computer Interaction.
[POL92] Polson, P., Lewis, C., Rieman, J., and Wharton, C. Cognitive Walkthroughs:
A Method for Theory-Based Evaluation of User Interfaces. International Journal of
Man-Machine Studies, 36, (1992), 741-73.
89 Michael J Vernik
[PON01] Ponnekanti, S., et al. (2001) ICrafter: A Service Framework for Ubiquitous
Computing Environments. in Ubicomp 2001. Atlanta, GA, USA.
[ROB04] Paul Robertson, Robert Laddaga, and Max Van Kleek. Virtual Mouse
Vision Based Interface. In Proceedings of the 9th International Conference on
Intelligent User Interfaces, pp.177--183. 2004.
[SLA03] Slay, H., R. Vernik and B. Thomas (2003). Using ARToolkit for Passive
Tracking and Presentation in Ubiquitous Workspaces. Second International IEEE
ARToolkit Workshop, Waseda University, Japan.
[STR98] Streitz, N.A., J. Geissler, and T. Holmer. (1998) Roomware for cooperative
buildings: integrated design of architectural spaces and information spaces. in
Cooperative Buildings, Integrating Information, Organization, and Architecture First
International Workshop CoBuild'98 Proceedings. Darmstadt, Germany Springer-
Verlag,
1998.
[STR99] Streitz, N., et al. (1999) i-LAND: An interactive Landscape for Creativity
and Innovation. In ACM Conference on Human Factors in Computing Systems
(CHI'99). Pittsburgh, PA, USA: ACM Press, New York, NY, USA.
[TAN03] Tandler, P., (2003) The BEACH Application Model and Software
Framework for Synchronous Collaboration in Ubiquitous Computing Environments.
To appear in Journal of Systems and Software, 2003 (Special Issue on Application
Models and Programming Tools for Ubiquitous Computing).
[VER03] Vernik, R., Blackburn, T. and Bright, D., (2003): Extending Interactive
Intelligent Workspace Architectures with Enterprise Services. Proc Evolve2003,
Enterprise Information Integration, Sydney, Australia, 2003.
[VER04a] Vernik M.J., Johnson S., Vernik R.J. (2004) "e-Ghosts: leaving virtual
footprints in ubiquitous workspaces", Australaisian User Interface Conference,
Dunedin NZ.
[VER04b] Vernik R.J., Johnson S., Bright D., Vernik M.J. (2004) " Using Workspace
Simulation to Support the Evaluation of LiveSpaces for Synchronised Planning
Activities", SimTecT 2004, National Convention Centre, Canberra, 24-27 May 2004.
[VER04c] Vernik, R., Johnson, S., Bright, D., and Vernik, M. (2004) 'Using
Workspace Simulation to Support the Evaluation of LiveSpaces for Synchronized
Planning Activities', in Proceedings of SimTect 2004, Canberra, Australia, 2004.
90 Michael J Vernik
recommendations. CHI'92. Proceedings of the ACM conference on human factors in
computing systems. Striking a balance, 3 - 7 May 1992, Monterey, California, pp381 -
388
[WEI91] Weiser, M. (1991) The computer of the 21st Century. Scientific American,
Sept. 1991, 94-104.
91 Michael J Vernik
Appendix A – Ethics Proposal Submission
DSTO
Command & Control Division
PO Box 1500 Edinburgh SA 5111
AUSTRALIA
Tel: 08 02596611
Fax: 08 8259 5619
February 2010
(P. Evdokiou)
Manager, Research Coordination and Collaboration
Command Decision Environments
Command and Control Division
92 Michael J Vernik
28 September 2004
Michael Vernik
1 Hobart Crs
BANKSIA PARK SA 5091
Dear Michael
I am pleased to advise that the Divisional Ethics Committee has granted approval for your project
“Session Interfaces for Ubiquitous Workspaces”.
It is essential that you conduct all research according to UniSA guidelines, which can be found at
http://www.unisa.edu.au/orc/ethics/index.htm
Yours sincerely
Dr Ian Clark
Chairperson
Divisional Ethics Committee
93 Michael J Vernik
Section 1: Coversheet
Researcher’s signature
I certify that the protocol is complete and the research will be
conducted in accordance with the protocol and in an ethical
manner.
Supervisor’s signature (if
researcher is a student)
I certify that this protocol has been completed in accordance
with the requirements of the Uni SA’s protocol guidelines. I
have considered the ethical implications of the proposed
94 Michael J Vernik
research and believe that the research will be conducted in
accordance with the relevant local, state, national and Uni SA
policies, guidelines, regulations and legislation.
Supervisor's name (if Bruce Thomas
researcher is a student)
Supervisor's contact details (mail D building, University of South Australia, Mawson
and email addresses, telephone Lakes Campus
number/s) Bruce.thomas@unisa.edu.au
8302 3464
Please note that protocols which do not provide sufficient information for HREC or DEC to make an
adequate assessment may be returned for revision.
Section 2: Checklist
Does the research involve participation of Aboriginal or Torres Strait Islander people who have
been selected YES / NO
as research participants because they are indigenous Australians?
Does the research involve any artifacts that are of cultural, spiritual or religious significance to YES /
NO
Aboriginal or Torres Strait Islander people?
Does the research involve an unusually dependent relationship between the researcher and any of
the YES / NO
research participants?
Could the research place research participants in an unusually vulnerable situation? YES / NO
Is there any potential risk (physical, emotional, social or legal) to individual participants’ well
being, YES / NO
beyond that normally encountered in everyday life, as a result of their involvement in the
research?
Does the research involve the administration or application of drugs and/or Clinical Trial Notification Scheme (CTN)
documentation? YES / NO
Is there any reasonable likelihood that the research will result in the reporting of suspected child
abuse? YES / NO
95 Michael J Vernik
Is there any potential risk to the researcher’s safety, beyond that normally encountered in everyday
life, as a result of their involvement in the research? YES / NO
Does the study have potential legal implications for the researcher or the University? YES / NO
Have you applied for funding for this research (other than Divisional funds)? YES / NO
If YES, please give the name of granting body applied to and the type of grant sought.
Staff, PhD, Professional Doctorate and Masters by Research proposals must be considered by the
Human Research Ethics Committee.
All other student research should be considered by the appropriate Divisional Ethics Committee
unless the researcher answered YES to any of the questions above.
If you are uncertain about how to answer any of the questions or whether HREC or a Divisional
Ethics Committee should consider your proposal please seek advice from the chair of the Divisional
Ethics Committee.
96 Michael J Vernik
Please answer the remaining questions
Internet questionnaires
Other questionnaires Y
Face to face interviews which do not Y Face to face interviews which request
request personal or sensitive information personal or sensitive information
Please tick the group/s from which your sample of participants will be drawn for this study
People with whom you have an ongoing Members of the armed forces Y
professional relationship (eg students or
clients)
97 Michael J Vernik
In which country is data collection to take place?
Australia____________________________________________
98 Michael J Vernik
Will the research involve access to individuals, clients or records required from any organization?YES /
NO
Will you access individually identifiable information about participants from any
government department? YES / NO
or
from another organization (for example a state based register, education department,
hospital, health department, correctional services)? YES / NO
Have you received approval to access this information from the government
department(s)/organization(s) listed above? YES / NO
Appendix 2 Research tools YES / NO/ Not required for this study
Appendix 3 Recruitment material YES / NO/ Not required for this study
99 Michael J Vernik
Thesis/dissertation Y Conference presentation
Will research participants have the opportunity to receive a copy of your final report if they wish?
YES / NO
Ethics approval will not be finalised until copies of all necessary materials have been received by the
executive officer of the ethics committee considering your protocol.
Please keep your responses as brief as possible while providing enough information for
members of the ethics committee to gain a good understanding of what your research will
involve. The Guidelines for Gaining Ethics Approval provide advice about what the committee
requires. Remember that members of the ethics committee might not have the same
background in your area of study that you have. Your responses should be written in plain
English for a non-expert audience.
The suggested length of responses is a guide only. Simple, uncontentious research might be
adequately explained more briefly. Research projects with a number of component parts or
which involve possible risks to the research participants will require more detailed
explanation.
Some questions might not be relevant to your study, for any that are not simply write N/A.
1 RESEARCH AIMS
1.2 Explain the need for, and value of, your research. (100-300
Place the aims in the context of existing research or practice. words)
Include a list of not more than 10 key references at appendix 1.
2 RESEARCH METHODOLOGY
The second stage involves a free-play time with the session interface. This is
intended to give the participants a hand on approach to the technology so that
they can make a more accurate evaluation. The duration of time the
participants will utilize the session interfaces will depend on what tasks they
wish to do. For example, using Plexus Plus, they may decide to move
documents to a different display many times in order to see the effectiveness
of it. Similarly, questions will be recorded, as well as suggestions towards
the technology.
After demonstrations are complete, the participants will be asked to fill out
an evaluation questionnaire. This questionnaire features questions based on a
framework that has been developed specifically for evaluation of
environments such as a LiveSpace.
2.3 Indicate whether your research is the first stage of a larger project. (50-100
If it is, briefly explain your intentions for the development of your study to words)
facilitate further ethics approval if you do extend your research project.
No, this is not the first stage of a larger project. It is the evaluation of an
honours topic and will be completed at the end of 2004.
3 RESEARCH PARTICIPANTS
3.1 Who will be approached or recruited to be research participants? How (50-100 words)
many participants will be involved in your study?
3.4 How will you provide detailed information about your study to potential (50-100
participants? words)
Include as appendix 4 the information sheet/s that you will use.
Please ensure that any documents you will provide to research participants
have been carefully proof read prior to submission to HREC or the DEC.
Secondly, defence visitors will receive a similar overview of the project from
the orchestrated evaluation session. Following this, they will be asked to fill
out the questionnaire.
3.5 Describe how you will obtain consent to participate from those (100-200
volunteering as participants for your research. words)
Include as appendix 5 the consent form or forms that you
will use.
Please note that consent is not required for anonymous questionnaires. Return
of the completed questionnaire indicates consent.
A consent request form will be asked to be filled out. This will be part of a
package given to each participant.
3.6 If your research participants will be drawn from any dependent group (100-200
(people who have an unequal power relationship with you or with an words)
organisation which is cooperating in the research) please detail how will
you ensure that participants do not feel under any obligation to assist
you with your research as participants.
All participants are either colleagues (in which case, will have anonymous
questionnaires) or visitors from DJFHQ. The visitors are not of a dependent
group. If at any time a participant (whether visitor or colleague) wishes to
stop the evaluation, they may do so. No harm will result from this.
3.7 Describe how you will preserve participants’ confidentiality as you (50-100
4.1 Describe briefly how the research data will be recorded, for example, (50-100 words)
audiotape, videotape, or written notes.
Please note that explicit consent must be obtained from participants if
material is to be audio or videotaped or photographed. Provision for this
should be included in the consent form.
People that will have access to the information includes myself, my primary
supervisor, Bruce Thomas and my secondary supervisor, Damien Bright.
5 OWNERSHIP OF THE RESEARCH
AuSPlanS
Technology Exercise 2.1
Intense Collaboration Space
DSTO Edinburgh
DSTC, SA Node
Command and Control Division, DSTC
University of South Australia
N/A
2. Using Plexus, do you see the information resource as useful for accessing a set of
key documents available to a planning team?
Not useful Extremely useful
N/A
3. Using Plexus, do you see the transferring of planning documents between group
members as useful?
Not useful Extremely useful
N/A
N/A
N/A
N/A
N/A
N/A
N/A
10. Is the graphic representation of the LiveSpace environment, together with device
and participant overlays, important for interactions using Plexus.
Not important Extremely important
N/A
11. Is the ability to control and configure devices such as lights, projectors from an
individual desktop important?
Not useful Extremely useful
N/A
12. Would the use of the Plexus interface be useful to provide a record of the session at
some later date (e.g. accessible outside of a LiveSpace)
Not useful Extremely useful
N/A
13. Would Plexus aid a remote individual who is part of the planning team?
Not useful Extremely useful
N/A
14. Would Plexus aid a remote individual who is part of the planning team?
Not useful Extremely useful
N/A
15. Using Plexus, do you see viewing headquarters wide information resources useful?
Not useful Extremely useful
N/A
N/A
15. Do you have any other comments which would help in the development of future
capabilities such as Plexus?
AuSPlanS
Technology Exercise 2.1
Intense Collaboration Space
DSTO Edinburgh
DSTC, SA Node
Command and Control Division, DSTC
University of South Australia
N/A
2. Do you think the concept of how to start up and shutdown a LiveSpace is difficult to
understand with Ignite?
Unacceptable Acceptable
N/A
3. Using Ignite, do you see the ability to invoke a predefined session through a common
interface important?
Not important Extremely important
N/A
4. Using Ignite, would you see viewing the information resources available for a
session, such (e.g. resources like documents and maps) as being useful?
Not useful Extremely useful
N/A
5. Using Ignite, would you like control and configure individual devices such as
projectors and lights etc?
Not useful Extremely useful
N/A
6. Using Ignite, would you find it useful to be able to pre-setup a LiveSpace the way
you would like it, or would you prefer to use default settings?
Not useful Extremely useful
N/A
N/A
8. Using Ignite, do you see leaving messages for a group in a LiveSpace useful (e.g. via
a ticker tape on the interface)?
No benefit High level of benefit
N/A
9. Using Ignite, do you see leaving messages for an individual in a LiveSpace useful?
Not useful Extremely useful
N/A
10. Would you prefer to control and set the configuration of Livespaces devices (such as
lighting, audio visual matrix) using the Ignite touch screen, or from an interface
available to each user at their desk (closer to centre is both)
Ignite Individual Desk
N/A
11. Using Ignite, is the ability to see the availability status of a LiveSpace from your
office useful?
No benefit High level of benefit
N/A
12. Using Ignite, is the ability to see the health status of a LiveSpace environment useful?
Not useful Extremely useful
N/A
N/A
14. Using Ignite, would you find it useful to pause a session and resume it at any stage
(e.g. resume a meeting a day after it occurs, providing the LiveSpace exactly the way
you left it)?
Not useful Extremely useful
N/A
15. Are there any other comments you would like to make about Ignite?
AuSPlanS
Technology Exercise 2.1
Intense Collaboration Space
DSTO Edinburgh
DSTC, SA Node
Command and Control Division, DSTC
University of South Australia
N/A
2. Using PointRight, do you find it is easy to control applications on the large displays?
Not easy Extremely Easy
N/A
3. Do you find the notion of moving your mouse cursor to the top of your local display
to control another display intuitive?
Not intuitive Very intuitive
N/A
4. Do you find it difficult to switch to an alternative display (e.g. moving your cursor
left or right of a display)?
Difficult Easy
N/A
5. Do you feel that the ‘highlighting cursor feature’ useful if you lose it amongst the
displays?
Not Useful Extremely Useful
N/A
6. Do you find the ability of having multiple users control the same display useful?
Not Useful Extremely Useful
N/A
N/A
8. Is the mouse movement on a display enough to indicate that the display is being
used?
Unacceptable Acceptable
N/A
9. List any other ways in which you might like to interact with information displays in
addition to Pointright.
10. Do you have any other comments that you would like to contribute?
Plexus
J
J5 J5 JET J55 Average
Q2 8.2 10.5 7.6 8.766667
Q3 10.4 10.5 10.5 10.46667
Q4 8.7 6.1 10.5 8.433333
Q5 10.5 10.5 10.5 10.5
Q6 5.25 10.5 10.5 8.75
Q7 10 10.5 10.5 10.33333
Q8 9.2 10.5 10.5 10.06667
Q9 10.3 10.5 10.5 10.43333
Q10 7.5 10.5 10.5 9.5
Q11 7.4 8.5 7.95
Q12 7.5 8.4 7.95
Q13 7.6 7.1 7.35
Q14 9.5 10.5 10
Q15 10 10.5 10.5 10.33333
Q16 6.2 10.5 7.8 8.166667
Q17 7.7 10.5 10.5 9.566667
Q18 10.2 10.5 10.5 10.4
Q19 7.3 10.5 7.8 8.533333
Q20 10.5 10.5 10.5 10.5
Q21 9.6 10.5 8.2 9.433333
J Members
Focus Group
Qn J53 J53b J54 J54b J5 J5b RANCO RANCO Average
Control
Group
Control
Qn RV DCW TR DMK JP Average
J Members
Focus Group
Qn J53 J53b J54 J54b J5 J5b RANCO RANCO Av
1 10.5 10 8 8 6 6 10 10 8.2
2 10.5 10 10.5 10 5.5 5 9.5 9 8.6
3 10.5 10 6 6 8 8 10 10 8.2
4 10.5 10 8.5 8 8 8 5.5 5 7.7
5 10.5 10 10.5 10 6 6 5.5 5 7.7
6 10.5 10 10.5 10 5.5 5 9 9 8.5
7 8 8 10.5 10 5 5 9 9 7.7
8 9 9 7 7 7 7 8.5 8 7.5
9 10.5 10 5.5 5 8.5 8 9 9 8.0
10 10.5 10 10.5 10 6 6 9 9 8.6
11 10.5 10 7 7 7 7 9 9 8.0
12 10.5 10 9 9 4.5 4 9.5 9 8.0
13 10.5 10 10.5 10 5.5 5 9.5 9 8.6
14
15 10.5 10 10.5 10 5.5 5 9 9 9.1
16 10.5 10 10.5 10 5 5 9 9 9.0
Control Group
Control
Qn RV DCW TR DMK JP PE Average
1 9.5 4 7 5 10 9 7.1
2 9.5 5.5 7.5 6 10.5 6.5 7.2
3 9.5 3 6 6.5 5.5 8.5 6.2
4 9.5 5.5 3.5 8.5 5.5 2.5 5.6
5 9.5 3.5 5.5 7 8 6.5 6.3
6 10 3 3.5 7.5 5.5 8 6.0
7 7.5 6 7 7.5 5.5 7.5 6.5
8 6 5 5.5 7.5 10 6 6.3
9 9.5 6 6.5 9 10 7.5 7.7
10 10 6.5 6 5.5 10 7 7.1
11 9.5 6 9 8.5 10 9.5 8.3
12 10 3 6 6.5 10 7 6.7
13 3.5 7.5 3 7.5 10 8.5 6.3
14 8.5
15 1 1.5 8 8 10 7 5.6
16 1.5 1 7.5 N/A 10 6.5 4.2
J Members
Focus Group
Qn J53 J53b J54 J54b J5 J5b RANCO RANCO Average
1 11 10 11 10 6.5 6 9 9 8.7
2 11 10 7.5 7 7 7 9 9 8.1
3 8 8 1 1 8.5 8 1 1 4.4
4 11 10 11 10 7.5 7 9 9 8.9
5 11 10 11 10 9 9 8.5 8 9.2
6 7.5 7 3.5 3 6 6 8.5 8 6.1
7 11 10 5.5 5 6 6 9.5 9 7.5
8 11 10 11 10 6.5 6 3.5 3 7.4
Control Group
Control Group
Qn RV DCW TR DMK JP PE BHT Average