See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228983376

A Framework for Data Collection of Collaborative Design Research

Article · January 2007

CITATIONS READS

5 455

1 author:

Peter Törlind
Luleå University of Technology
65 PUBLICATIONS 437 CITATIONS

SEE PROFILE

All content following this page was uploaded by Peter Törlind on 26 August 2016.

The user has requested enhancement of the downloaded file.

INTERNATIONAL CONFERENCE ON ENGINEERING DESIGN, ICED’07
28 - 31 AUGUST 2007, CITÉ DES SCIENCES ET DE L'INDUSTRIE, PARIS, FRANCE

A FRAMEWORK FOR DATA COLLECTION OF

COLLABORATIVE DESIGN RESEARCH
Peter Törlind
Faste Laboratory, Division of Functional Product Development
Luleå University of Technology, Sweden

ABSTRACT
In design research different methods are used to develop a descriptive understanding of a design
activity. One approach is to use video and audio recordings of complex activities such as team
meetings in engineering design. Video and audio recordings provide the researcher with a permanent
data corpus that can be used to understand events considered difficult to observe in real time. The
video approach also has several limitations, especially when observing distributed teams.
This paper describes the needs for a data collection framework of distributed meetings, and how a
prototype system has been realized. It also presents a framework on how a physical environment – the
design studio – can be used to further facilitate design research.
The presented tool is based on a conferencing system where additional video streams from the design
environment can be recorded (showing detail views of interesting areas etc.). The tool uses a layered
video approach and all video streams can be recorded and replayed for analysis. The tool supports
visual and textual bookmarks that can be used to find a particular event in the recorded data. The tool
clearly has some advantages compared to traditional data collection methods and enables researchers
to follow distributed design sessions.

Keywords: computer supported cooperative work, distributed product development, data collection,
design studies.

1 INTRODUCTION
Ethnographic techniques are often used in engineering design to collect information and data from
engineering work where manual observations, field notes and interviews are combined with video and
audio recordings of the design session, and later analyzed [1]. This approach has been used in the
Polhem Laboratory for the last ten years, when researchers have followed several product
development projects [2]. These projects involved both distributed and co-located design teams, and
have mainly been of two types, namely industry projects and university projects. In industry projects
[3,4], the researcher follows an industrial product development project over a longer period of time.
University projects have followed global student projects within the Sirius course at Luleå University
of Technology [5,6,7], i.e. the final-year course in product development for engineering
undergraduates studying towards an MSc. in Mechanical Engineering. Sirius involves teams of
students carrying out a product development project in close cooperation with industry partners. The
global teams have typically consisted of a about four local students and four students from another site
with similar courses, both nationally (Royal Institute of Technology, SE; Chalmers, SE) and
internationally (Stanford University, USA). The project also includes a large amount of
communication between the students and the industry sponsor that provides the industrial context,
requirements and funding.
The main advantage of following a university project is that the environment can be controlled more
easily, and methods, groupware and technology can easily be deployed. University projects typically
last about nine months, when several iterations of technology and tools are implemented and evaluated
[8]. Also, the researcher has the advantage of close proximity to at least one group of students.
Following teams in industry gives a more accurate view of an industry related problem, and also an
introduction to a more complex environment. Product development in industry is often done in large
teams and can be very difficult to follow, even for a large research group. It is also difficult to

ICED’07/ 529 1
instrument the team environment, and the use of video recording is often restricted due to company
regulations and is thus limited to interviews, field notes, etc.
As well, it is easier to conduct a descriptive study in industry rather than attempting to intervene and
change the work process by introducing new methods. Even more difficult is the issue of
implementing new software tools in strictly restricted environments, when the process of deploying a
new tool in the existing environment of a large company often follows strict regulations and includes a
process of conformance testing, security assessment, etc.
In Luleå, combining ethnographic observations of engineering work in industry with university
projects and controlled experiments has solved this dilemma. From descriptive studies in industry
relevant issues can be identified and then further evaluated in an experimental setting [9]. An
environment for these types of studies was realized through the design of a new collaborative design
studio.
This paper describes the advantages and limitations of video based data collection and analysis, along
with formulating the needs to overcome some of the existing limitations when following distributed
meetings. Finally, the paper describes the development of a prototype system, i.e. a framework for
data collection.

2 DATA COLLECTION IN DESIGN RESEARCH

Data collection from design sessions normally comes from experiments or from the actual engineer’s
practice. One guiding principle of Ethnography is to study the activities of people in their natural
setting [10], without interfering with the ongoing work. In research fields such as CSCW and HCI, the
usage of video recording to observe people interacting with different types of technology (hardware,
user interfaces, etc.) has been used for many years as one component for interaction analysis [11]. In
design research, the Delft second delft workshop Research in Design Thinking II - Analysing Design
Activity [1] popularised the method of using video and audio to analyse and understand design work.
For many years, The Center for Design Research at Stanford University has used the observe-analyse-
intervene approach initially described by John Tang [12] and developed specifically designed physical
environments [13, 14, 15] to simplify data collection in design research.
Several techniques to collect video and audio exist, though data collection is traditionally based on the
use of several cameras to record the design work from several angles and microphones to collect audio
from the design session. By only recording video data much of the interaction is missed; team
members in a design session often interact with different types devices, such as laptops, stationary
computers and mobile devices.
Badke-Schaub and Frankenberger [16] differentiates between direct methods and indirect methods
where the direct method uses observation of design work (or analysis of video recording) and the
indirect method consists of analysis of, diary sheets (papers with notes from problem solving)
containing design rational etc. Badke-Schaub and Frankenberger conclude that this type of indirect
data enriches and complements the data collected from the direct methods.
The capturing of meeting information from several sources has been a topic of research in several
communities, such as the Workspace Navigator [15], Quindi Meeting Companion [17], and Ferret
Browser [18]. These tools are primarily designed to capture information in one specifically equipped
room and can store and index multiple streams of data from the work environment, e.g. video, audio,
slides from presentations, etc., all of which are time stamped and presented on a timeline.
Screen recording software is also a useful source for workplace analysis. Tang et al. [19] conclude
from a study using Camtasia screen recording software that it is “Unobtrusive but invasive”, as well
as collecting rich empirical data without interfering with their normal work practice (no physical
presence of a physical observer or video equipment). However, they also conclude that the
disadvantage is its invasiveness replicating all interactions with the computer.

2.1 A physical research environment for data collection – the Collaborative Design
Studio
The collaborative design studio is designed as a “rapid-response environment, in which the
significance of issues raised through ethnographic observations of engineering work can be evaluated
and solutions offered” [9 p1].
The main goal with the new studio was to “Create a flexible research environment where new
innovative ideas can be deployed, tested and evaluated. The environment will support local as well as

ICED’07/ 529 2
global creativity. Trough the strong connections with the Faste Laboratory and the Sirius project the
new environments will be evaluated in global product developments projects”.

Figure 1 Concept Design Sketches, courtesy of Hans Walloschke, Arkitekthuset

Monarken.
The design rational of the collaborative design studio is further described in Larsson et al. [9]. In the
studio researchers have the possibility to rig specific experiments that can be carefully monitored and
follow design in action via ‘real-world’ development projects (i.e. the university projects mentioned
above). There is also a possibility to restrict one of the design spaces to one team for an extended
period of time, thereby creating a team space within the studio.
The studio is inspired by similar approaches, such as the Design observatory [20] at Stanford
University. In our approach the design spaces (2 and 4 in Figure 1) can be easily reconfigured similar
to a scene in a theatre or a television studio. In the design spaces several HD cameras can be used to
record design activities, and the infrastructure can support real time, i.e. down conversion, recording,
and recompression of video. Control systems are highly integrated, and researchers have the
possibility to route audio and video from the small studios and view all video sources on a large
powerwall (i.e. a large high resolution display 7,2x2,2m) in a facility across the corridor (the Studio).
Here, researchers can simultaneously monitor up to four full HD streams from the design spaces.
The infrastructure provides the researchers with an excellent environment for design observations, e.g.
the design spaces can be used in experiments to simulate geographic distance, where the researchers
can observe the design session as it unfolds from the observation bridge (10 in Figure 1) or by a larger
group in the Studio. This outside view (i.e. the researcher is not inside the design space) has several
advantages, specifically if several researchers are monitoring the development in a design session
interesting issues can be found, and the discourse between the researchers as the events unfolds can
rapidly lead to an improvement of the experiment, change the experimental setup, etc.
In a true distributed project, the researchers in Luleå have limited control and overview over the
remote site. To take even greater advantage of the new research studio new tools and methods were
needed to simplify observation, documentation and storage of distributed design meetings.

3 CHALLENGES FOR DATA COLLECTION AND ANALYSIS

Rhuleder et al. [11] present several advantages of using video for interaction analysis, the main one
being that video creates permanent data corpus and can capture complex data that is impossible to note
by simply observing the event. Finally, it provides several researchers the opportunity to do their own

ICED’07/ 529 3
interpretations and a collaborative multidisciplinary analysis can create an unbiased view of the
events. The video approach also has several problems, some of which are confidentiality, complexity
and the time consuming and expensive analysis.
One challenge is to store metadata with other types of information. Distributed projects and real
product development projects further complicate data collection. Finally, the analysis of collected
data is today poorly supported in existing tools.

3.1 The challenge of storing and retrieving additional metadata

A big challenge is to store other types of metadata that enrich the data collection (e.g. storing
interactions with computer tools in the environment and quantitative data from system log files).
Screen recording can either be done by software (compressing to a video file) or hardware (full
resolution or down converted to video). Hardware solutions that save full resolution are expensive and
require large storage systems; hence, the software approach is the emerging approach used in design
research. Within the HCI (Human Computer Interaction) community the usage of video analysis of
user interaction is widely used and is often supplemented with other types of data collection, such as
screen recording and the recording of eye movements.
The main challenge is not to store information, but rather replay and analyse the data in an effortless
fashion, as well as methods to integrate and aggregate different types of information.

3.2 The challenge of distributed projects

The complexity of data capture will increase when following distributed projects, where at least two
sites should be monitored. The actual audio and video stream used for communication should also be
saved so that comparisons of what happened and what could be seen (from the remote site) can be
done in the analysis. It is important to store the video with the actual resolution and frame rate as the
remote collaborators receive.
Many existing conferencing applications, such as Marratech [21], can record a meeting (saving video,
audio, whiteboard and application sharing). Several systems are also available to store the content of a
normal video conferencing using the H.323/H.320 standard, normally storing only one video stream
and one stream for application viewing or document camera (H.239).
In a distributed design study researchers at the sites may observe the design session and use field notes
to document additional information. These field notes can save much time in the video analysis
because they can highlight a specifically interesting moment within a long session that later can be
analysed in detail using the recorded material. An onsite researcher has the possibility to change or
adjust the technical equipment to provide a rich data collection.

3.3 The challenge of analysing the data

The analysis of raw data is a tedious task and support tools are needed to synchronize the different
media stored from a design session (multiple cameras, audio, and field notes). Some systems from the
HCI area that support analysis of multiple data types have emerged, such as d.tools [22], that record
video and other metadata to be later used for comparative analysis. These tools are normally designed
to record information from only one user. When following a product development project over time,
the amount of raw data soon becomes unmanageable; therefore, it is important that the tool also
supports the retrieval and analysis of stored data. To support the researcher it is important that the
system supports indexing of the data and provides the possibility to add metadata during the design
session and in the following analysis of the collected data. Törlind and Larsson [23] describes three
types of indexing, viz. active indexing done by the users, researchers or both, automatic indexing
created automatically by the system and passive indexing indexed automatically from the usage
pattern of users who re-examine the information.

3.4 The challenge of following real projects

When following a design team over a longer period of time the amount of recorded data soon becomes
unmanageable, and recording all design meetings is therefore impossible. One challenge of following
real projects over time is to have the possibility to record a design meeting with very low overhead and
only record data if necessary. If a researcher focuses on interaction with physical objects, it might not
be interesting to store long meetings when the team is interacting with project planning software, etc.

ICED’07/ 529 4
However, if the team chooses to skip the planning and instead do some work with mock-ups, the
session is relevant for the researcher and video and data should be recorded from the meeting.

4 DESIGN AND IMPLEMENTATION

Even though the experimental studio is designed to provide an excellent environment for design
research, there was a need to simplify data collection and analysis of design work. Hence, from the
challenges described above the objectives for the prototype were derived;
• Collect data without interfering the ongoing engineering work
• Capture of distributed meetings with several video streams (both collaboration media and local
interaction),
• Possibility to replay entire design sessions,
• Possibility to add research metadata (e.g. notes from the researcher and images),
• All data should be time stamped to allow cross referencing within the recorded data
• An awareness of remote design collaboration work.

The framework for data collection is based on the commercially available conferencing software Alkit
Confero, an integrated audio-/videoconferencing system. The Confero system has for many years
supported different types of research prototypes at Luleå University of Technology [2], such as mobile
conferencing, life size conferencing and stereo videoconferencing. By building the data collection
framework on an existing conference tool, a prototype can rapidly be designed and evaluated.
The Confero system also support a vide range of video hardware; Fire wire cameras, USB-cameras
and high quality framegrabbers.
Some basic support for data collection was implemented in earlier projects, when the user could
locally store a video stream as a QuickTime file, see Figure 2. The video could be combined statistics
from the conferencing sessions (bandwidth used, frames per second received, spatial size of video,
video compression algorithm, dropped frames, etc.), which are time-stamped to simplify quantitative
analysis of video conferencing usage.

Figure 2 Replaying of a stored session (only conferencing video).

Several new functions were added in the prototype, specifically adding support for auxiliary cameras,
the possibilities to record all video streams using a layered conference approach, and textual and visual
bookmarks.

4.1 Using a layered conference approach

A conferencing session is normally unicasted (point to point), multicasted (multipoint controlled by
the network) or transmitted via a reflector (client server approach where all users connect to one server
that transmits video to all members in the same session); to support data collection the reflector
approach was used.
To differentiate conferencing video from auxiliary cameras a layered approach was used in the
reflector where a session could inherit another session. The conferencing layer contains the conference
session used for communication between the remote teams. The above level is the data collection
layer, where researchers have the possibility to add several other video channels containing the
cameras in the room used to monitor the whole group, detailed views at whiteboards, etc. The top level

ICED’07/ 529 5
contains the research communication layer, where researchers can have their separate conversations to
comment the ongoing design session (normally only audio is recorded in this session). All video and
audio from a lower layer is inherited to the next layer so that the designers only see the video and
audio from the conference layer, and the researchers see all video and hear all audio from all layers. A
typical setup between two remote studios is visualized in Figure 3.

Figure 3 Typical setup for remote Collaboration (in this setup the researcher is at site 1).
In the setup two sites collaborates via video conferencing, CAM C1 and CAM C2. One camera at Site1
(CAM DC1) and two at Site2 (CAM DC2 and CAM DC3) is used to save video from different views.
The different layers are represented in the conferencing application as in Figure 4.

Figure 4 The video streams from the setup above and their corresponding layer.
In the setup the users only see the video from the conferencing layer while the researcher sees the Data
Collection layer which inherit the video streams from the conferencing layer (a total of five video
streams), the researcher also has the possibility to add bookmarks which is saved in the research layer.
To view all video streams simultaneously, a web interface was designed where all video streams are
presented as video thumbnails (low frame rate and resolution). Using this solution the user has an
awareness of the complete interaction and can at any time choose to view any stream with full
resolution and frame rate. The user can also decide which streams to record for further analysis.
Existing functions such as remote camera control are also very useful, allowing the researcher to
control the camera view used for data collection from a remote location. An observing researcher can
also invite a remote researcher by mailing an invitation with a session description, SDP (which
includes the address to the reflector and the session information). If the Confero software is installed
the researcher just clicks on the SDP file to follow the design session.

ICED’07/ 529 6
4.2 Simplifying analysis
To simplify analysis the use of visual and textual bookmarks was implemented. Visual bookmarks are
snapshots of the video stream or the shared application. Textual bookmarks correspond to a short note
from the researcher. These bookmarks are stored and time stamped and can be used to rapidly find a
specific part in a long session, see example in Figure 5.

Session information

Stream info
and time stamp

Textual bookmark

Visual bookmark

Figure 5 The web interface of the reflector, showing bookmarks in the recorded session.
Both textual bookmarks and visual bookmarks can be accessed from the Confero user interface, or via
keyboard shortcuts.

4.3 Reducing bandwidth

The reflector also supports the user by limiting the transmission of video to the receiver, i.e. re-
sampling (reducing the spatial resolution of each video frame) of video and changing the frame rate
(reducing temporal resolution). This enables a remote researcher (with low bandwidth) to follow the
design session with reduced video quality, whereas users with high bandwidth can receive high quality
video.
The user can also decide which compression algorithm to use; today Confero supports DV, Motion
JPEG, MPEG2 and H.264. The complexity of the encoding reduces the bitrate and increases the
process load of the computer. JPEG encoding is the most attractive in terms of computational
complexity and the H.264 codec provides the best compression performance.
In another project Motion detection was implemented in Confero. After some testing it was apparent
that the motion detection algorithm could also be used to reduce the amount of data needed to store the
video from a meeting. If a multi-camera setup is used, the amount of data stored rapidly increases (a
normal DV-camera needs about 25Mbit/s; therefore, a recording of a distributed session with 3
cameras at each site consumes about 70 GB/h). By implementing a motion detection approach that
changes, the frame rate can be changed depending on the motion in the image.

4.4 Recording of meetings

By using the layered approach the researcher can decide to record some or all communication streams
(video, audio text messaging) that can be replayed later on. Confero also has the possibility to stream a

ICED’07/ 529 7
video file from a video server; if a user chooses to stream a video file, it is also included in the
recording. Text chat within Confero is also stored and replayed.
The user (conferencing layer) or the researcher (data collection layer) also has the possibility to
change the resolution of the video (to preserve bandwidth). This event is also stored and replayed in
the replayed session.

5 CONCLUSIONS
This paper has presented several of today’s challenges in data collection and presents a tool that
clearly has some advantages compared to traditional data collection methods. It also presents a
framework on how a physical environment – the design studio – can be used to further facilitate
design research.
Recording all video streams at the reflector ensures that the recorded video corresponds to the video
sent from the conferencing application (including compression artefacts, dropped frames, video size,
delays, etc.). The tool enables remote researchers to follow a local and remote design sessions, and
have a parallel discussion when the design session unfolds. As well the researcher can easily invite
other researchers if the opportunity arises. After the meeting, the researchers can easily replay the
distributed meeting with all video and audio synchronised, regardless of where the researcher is
situated. The framework also provides an advantage when following a design team over a longer time,
when a researcher may only have a brief awareness of the ongoing design session. By eaves dropping,
the researcher can respond if an interesting topic appears, and record the session, if necessary.
One limitation is that the system is dependent on high bandwidth networks between all sites and the
reflector and that all storage of video is done at the reflector. Several ideas to limit the bandwidth have
been implemented, such as variable frame rate (triggered by motion detection) and the possibility to
use high efficiency compression algorithms.

5.1 Re-Experiencing engineering meetings

The framework also supports the design team, which can use recorded data from the conferencing
layer to re-experience the meeting [23]. Team members can also use the annotation function to add
‘bookmarks’ to important moments in the meeting. By using this function a team can rapidly find and
recapture the discussions about design rationales regarding a specific concept or function instead of
replaying the complete meeting. This function is also useful to share knowledge to absent team
members.
The collected information can also be used as a knowledge bank of lessons learnt from previous
projects. Törlind and Larsson highlights that “The need to support rapid context adaptation also
includes how to quickly and effortlessly share knowledge with new, often temporary team members
who enter the project with little or no knowledge of the previous work. Enabling team members to re-
process and ultimately re-experience the dynamism and richness of meetings in which they did not
actively take part is a very appealing situation for globally distributed development teams,
particularly if they can experience a brief synopsis of the most important parts rather than repeat the
entire meeting.” [23 p. 9].

6 FUTURE WORK
The prototype system will now be evaluated in several research projects in the new collaborative
design studio. The system today only has a basic support for storing field notes, bookmarks and other
additional metadata. By combining this type of direct data with the possibility to rapidly capture and
crosslink additional indirect data (designers note, sketches etc) to a specific moment in time, analysis
can rapidly be improved. Another interesting possibility is to add support for different types of
protocol analysis methods [1]. A combined framework that combines data from both indirect and
direct methods as well as analysis support should substantially simplify the analysis of design activity.
For future work the system will need enhanced support for the analysis of the recorded data, when all
data (video, annotations, comments and other indirect data) can be presented at a common timeline.
Also the possibility to cross index and structure different sessions is both useful for both design
research and project work where users rapidly can find and re-experience past design sessions.

ICED’07/ 529 8
ACKNOWLEDGEMENTS
Support for this research is provided by VINNOVA through the Faste Laboratory at Luleå University
of Technology and the IIPD project funded by the European Union through the Interreg IIIA
Structural funds. The financial support to the Experimental studio from the Kempe Foundations and
the Knut and Alice Wallenberg Foundation is greatly appreciated. I also appreciate the feedback from
the reviewers which has substantially improved the paper.

REFERENCES
[1] Cross, N., Christiaans, H., and Dorst, K. (eds.) Analysing Design Activity, 1996 (John Wiley &
Sons, Chichester).
[2] Törlind, P., Larsson, Andreas., Löfstrand, Magnus., Karlsson, Lennart, 2005. Towards True
Collaboration in Global Design Teams? Proceedings of 15th International Conference on
Engineering Design, (ICED 05), August 15-18 2005, Melbourne, Australia.
[3] Törlind, P., Distributed Engineering via Broadband - a Case Study, Proceedings of 14th
International Conference on Engineering Design (ICED 03), Stockholm, August 19-21, 2003.
[4] Bergström, M., Törlind, P., Getting Physical – Interacting with Physical Objects in Distributed
Collaboration, Proceedings of 15th International Conference on Engineering Design (ICED
05), Melbourn, Australia, August 15-28, 2005.
[5] Törlind P. and Larsson A. Supporting Informal Communication in Distributed Engineering
Design Teams, Annals of 2002 International CIRP Design Seminar, Hong Kong, 2002.
[6] Larsson A., Törlind P., Mabogunje A. and Milne A. Distributed Design Teams: Embedded One-
on-One Conversations in One-to-Many, Common Ground: Proceedings of the Design Research
Society International Conference at Brunel University, Staffordshire University Press, 2002, pp.
604-614.
[7] Larsson, A., Making Sense of Collaboration: The Challenge of Thinking Together in Global
Design Teams, In Proceedings of GROUP 2003, Sanibel Island, FL, USA, November 9-12,
2003.
[8] Larsson, A., Törlind, P., Karlsson, L., Mabogunje, A., Leifer, L., Larsson, T and Elfström, B-O.,
Distributed Team Innovation – a Framework for Distributed Product Development, Proceedings
of 14th International Conference on Engineering Design (ICED 03), Stockholm, August 19-21,
2003.
[9] Larsson, A., Törlind, P., Bergström, M., Löfstrand, M., Karlsson, L, 2005. Design for
Versatility: The Changing Face of Workspaces for Collaborative Design, Proceedings of 15th
International Conference on Engineering Design (ICED 05), August 15-18 2005, Melbourne,
Australia.
[10] Blomberg, J., Giacomi, J., Mosher, A., and Swenton-Wall, P., Ethnographic Field Methods and
their Relation to Design, In Schuler, D., and Namioka, A., eds., Participatory Design:
Principles and Practices, Lawrence Erlbaum, Hillsdale, NJ, USA, 1993, pp. 123-154.
[11] Ruhleder K. and Jordan, B., Capturing, complex, distributed activities: video-based interaction
analysis as a component of workplace ethnography, Proceedings of the IFIP TC8 WG 8.2
international conference on Information systems and qualitative research , Philadelphia,
Pennsylvania, United States, pp. 246 – 275, 1997.
[12] J. Tang, Toward and Understanding of the Use of Shared Workspaces by Design Teams, Ph.D.
Dissertation, Stanford University, 1989.
[13] Carrizosa, K. Eris, Ö., Milne, A. and Mabogunje, A. 2002, Building the Design Observatory: a
Core Instrument for Design Research, Proceedings of International Design Conference - Design
2002, Dubrovnik, May 14 - 17, 2002.
[14] Milne, A. and Winograd, T., 2003, The iLoft project: a technologically advanced collaborative
design workspace as research instrument, International Conference on Engineering Design
(ICED03), Stockholm, August 19-21.
[15] Ju, W., Ionescu, A., Neeley, L. and Winograd, T. 2004. Where the Wild Things Work:
Capturing Shared Physical Design Workspaces, In Proceedings of Conference on Computer
Supported Cooperative Work (CSCW 2004), 6-10 Nov 2004. Chicago, Illinois, USA
[16] Badke-Schaub, P. and Frankenberger, E. 1999. Analysis of design projects. Design Studies, 20,
481-494.
[17] Rosenschein, S. J. 2004. Quindi Meeting Companion: A Personal Meeting-Capture Tool,

ICED’07/ 529 9
 CARPE’04, October 15, New York, New York, USA.
[18] M4 Multimodal meeting manager, Report on Final Demonstrator and Evaluation, IST project
report D4.3 IST-2001-34485. http://www.m4project.org/publicDelivs/D4-3.pdf
[19] Tang, J. C., Liu, S. B., Muller, M., Lin, J., and Drews, C. 2006. Unobtrusive but invasive: using
screen recording to collect field data on computer-mediated interaction. In Proceedings of the
2006 20th Anniversary Conference on Computer Supported Cooperative Work (CSCW 06),
Banff, Alberta, Canada, November 04 - 08, 2006.
[20] Carrizosa, K. Eris, Ö., Milne, A. and Mabogunje, A. 2002, Building the Design Observatory: a
Core Instrument for Design Research, Proceedings of International Design Conference (Design
2002), Dubrovnik, May 14 - 17, 2002.
[21] Marratech, URL: http://www.marratech.com, (20060126).
[22] Hartmann, B., Klemmer, S.R., Bernstein, M., Abdulla, L., Burr, B., Robinson-Mosher, A., Gee,
J. Reflective physical prototyping through integrated design, test, and analysis.
Proceedings of UIST 2006, October 2006.
[23] Törlind, P. and Larsson, A., Re-experiencing engineering meetings : knowledge reuse
challenges from virtual meetings, Proceedings of Challenges in collaborative engineering, Prag,
Czech Republic, 2006.

Contact: Peter. Törlind

Luleå University of Technology
The Faste Laboratory
Division of Functional Product Development
SE 97187, Luleå
SWEDEN
Phone: +46 920 492412
e-mail: Peter.Torlind@ltu.se
URL: http://www.ltu.se/tfm/fpd

ICED’07/ 529 10

View publication stats