Evaluating the MER Display Ecology …

                              Evaluating the MER Display Ecology

Elaine M. Huang                                              Introduction
GVU Center. College of Computing                             In January of 2004, the National Aeronautics and Space
Georgia Institute of Technology                              Administration (NASA) landed two unmanned vehicles on
85 5th St NW, Atlanta, GA 30308                              the surface of Mars for the purposes of collecting scientific
elaine@cc.gatech.edu                                         information regarding the terrain, composition, and
                                                             atmosphere of the planet. The Mars Exploration Rover
                                                             (MER) mission has continued for the past 20 months, with
                                                             the two rovers, Spirit and Opportunity, continuing to
                                                             transmit data to Earth as they traverse the surface.

                                                             The actions of the rovers and the data that they collect
                                                             are guided by mission scientists and engineers, and the
                                                             mission is based at NASA Jet Propulsion Labs (JPL) in
                                                             California. To coordinate their activities, scientists and
                                                             engineers employ a variety of displays for collaboration
                                                             and information sharing. In the group workspaces
                                                             designed specifically for the MER Missions, shared
                                                             displays, including large projection screens, large
                                                             interactive plasma displays, and shared workstations with
                                                             multiple monitor setups, are ubiquitous. Together, these
                                                             surfaces form a "display ecology," in which the uses of
                                                             individual displays influence the roles of others, despite
Figure 1. MERBoards, projectors, laptops, and workstations
               in the work environment                       not having been designed as a unified, seamless system.
                                                             Of particular interest to us is the MERBoard [1, 2, 3].

                                                             In this position paper, we focus specifically on how the
Copyright is held by the author/owner(s).
                                                             display ecology supported the specific task of daily Rover
CHI 2006, April 22­27, 2006, Montreal, Canada.
                                                             activity planning. We consider how the physical aspects of
ACM 1-xxxxxxxxxxxxxxxxxx.
                                                             the displays affected their use in this task, how the
                                                             affordances of different applications such as PowerPoint
                                                                                                                           2




and SolTree supported the task on different displays at         early in the mission, images of SolTrees were often
different points in time, and how the changing nature of        exported as images or transcribed into PowerPoint for the
the collaboration and the task itself affected scientists'      purposes of displaying them on the projectors during
choice of displays and tools.                                   meetings when the plans were being presented to larger
                                                                workgroup. The size and resolution of MERBoard were
The Sol Planning Task                                           well suited for small group authoring, but were not
MERBoard provided a tree-building tool called SolTree           sufficient to make MERBoard a valuable presentation tool
(Figure 2) to support this task; it allowed the scientists to   for this type of viewing.
visualize all possible paths for the Rovers and annotate
them as necessary. Despite the availability of this tool on
the MERBoard, scientists also used other displays and
tools for planning as the task and collaboration evolved.
Used regularly during the first 70 Sols (Martian days) of
the MER missions for planning activities, SolTree was the
primary planning tool early in the mission.

The use of a structured scaffolding tool on a shared
display surface entails several assumptions; it assumes
that the task that it supports will be done by a group of
people, rather than an individual. It assumes that
collaboration will be synchronous and co-located in such a
way that a shared visual surface will be beneficial.
Additionally, the design of this tool assumes everyday or
near-everyday use during the mission, since it was
intended to support planning on a Sol by Sol basis. We
found that these assumptions did not hold throughout;              Figure 2. Scientists collaborating on a plan using SolTree
the nature and timing of the Sol planning task evolved
over the course of the mission, as did the type of              Tool structure supports early collaborative work: Most of the
collaboration used to accomplish the task. The evolution        LTP scientists appreciated SolTree's ability to keep track of
of task and practice caused Sol planning to migrate off of      all of the possible branches and options, especially in the
the MERBoard onto laptops and projection screens, as the        earlier parts of the mission. Others praised the fact that
scaffolding and the shared visual surface offered by the        SolTree imposed a structure on brainstorming options; it
large display ceased to fit the task as the mission             required planners to think down each linear path and
progressed.                                                     consider and annotate all of the possibilities. User of the tool
                                                                said that it "forced explicit logic," "offered scaffolding," and
Migration to projection screen for large meetings:              required the scientists to consider all possible ramifications.
Although MERBoard was well-suited for the planning task
                                                                                                                     3




This was important because the process that was still new to   decreased, it continued to be used for passive
the scientists and not yet routinized.                         information display.

Persistence and evolution of plans: Though the general         Implications for Multi-Display Environments
perception of the SolTree tool among scientists is that it     In looking at the use of the NASA MERBoard over time,
was provided for interactive planning for Rover activities,    several patterns emerge across the various applications.
their descriptions of use illustrate a broader value of the    These patterns demonstrate the evolving role of the
tool as a persistent information display for community         system in the context of a dynamic work environment,
awareness. SolTrees were often left open on the                and a complex ecology of displays:
MERBoard even after the planners had completed their
planning for the day, as a way of maintaining awareness        · Changes in the collaboration style over time ­
of the planned activities and options. The persistence of        MERBoard's value for collaboration was that it
                                                                 supported synchronous sharing of artifacts; users could
the artifact created continuity from day-to-day between
                                                                 engage in viewing and authoring material
the various planning teams.                                      simultaneously. The fact that procedures became
                                                                 familiar and routinized meant that responsibilities could
Tasks migrate to other displays as collaboration changes:        be divided up and tackled individually, thus reducing
The planning process evolved during the course of the            the need for a shared work surface for synchronous
mission, shifting gradually from unfamiliar and                  collaboration.
                                                               · Changes in the tasks of the scientists over time ­
exploratory to familiar and proceduralized. As mission
                                                                 MERBoard's value for interaction was primarily as a
goals solidified, planning became more tactical, and             ramp-up tool that allowed users to conduct exploratory
scientists generally confined their planning to the              work, especially when procedures or tasks were
consideration of a few potential options rather than a full-     unfamiliar, and scientists benefited most from doing
blown exploration of all possible next steps. The decision       them together to see and learn how the problems
making process became increasingly streamlined. The              should be addressed. Over time, workgroups found less
                                                                 need for the shared exploration afforded by MERBoard.
method of visualizing these plans evolved as well, as did
                                                               · Other displays and applications available in the
the subsequent use of the multi-display environment.             environment ­ MERBoards were one of many display
Additionally, the planning process became predictable            technologies available. Other means of displaying
enough that scientists no longer needed to create them           information that also could be used for sharing, such as
together; it was sufficient for an individual to create the      laptops and shared workstations for very small
plan on his own and get it approved by the group later.          collaborations and projection screens for large meetings
As a result of this evolution of the task, the group use of      allowed tasks to migrate off of the MERBoard as
                                                                 necessary.
the SolTree tool on MERBoard for planning eventually
gave way to the individual use of PowerPoint on laptops
                                                               Taking these factors into account in evaluating the changes
for creating "Sol Paths" that were afterwards shown
                                                               in MERBoard use during the mission and pre-mission, we
ambiently on the MERBoards for awareness. As
                                                               identified some implications for display ecologies and large
MERBoard's value for synchronous active collaboration
                                                               interactive displays for supporting group work:
                                                                                                                      4




· The transition from interactive use to ambient display ­    such displays provide when tasks call for a shared visual
  Designers of large displays should expect that the          display or interactive work surface.
  interactive use of displays may not be constant over
  time, but that users may continue to find value in the
  ambient display capabilities of the systems. Applications   In the realm of large interactive display research, a
  should not be designed only with interactive use in mind;   decrease in interactivity is often viewed as a failure of the
  attention should also be paid to how applications might     system to support workgroup practices, but the interplay
  be support passive use, what kinds of content to support    between interactive and passive use proved to be an
  while the displays are not being used interactively, and    important aspect of the support that MERBoard offered.
  how that content should be displayed.
                                                              We observed a migration from interactive use to equally
· The dynamic use of multi-display environments ­ Large
  interactive displays in multi-display environments are      valuable ambient information display. We believe that
  by nature group-owned and flexibly appropriable;            success should be assessed by looking both at
  constant, steady use need not the primary measure of        interactivity and the value of the display in passive uses.
  success. Multi-display environment designs should be
  flexible and dynamic and perhaps easily reconfigurable.     Finally, in the greater context of a display ecology, it is
  They should support the fluid migration of tasks among
                                                              misleading to evaluate the isolated use of a single
  various display surfaces.
· Support for undefined tasks and proceduralization ­         system; the existence of other displays in the
  Systems such as MERBoard support exploratory tasks          environment means that it is important to understand
  and tasks that do not have a set procedure, becoming        how the ecology functions as a whole, not just how
  less necessary when work becomes streamlined and            individual displays are used. In evaluating displays in
  routinized over time. Designing for continuity by           such multi-display environments, we believe it is better to
  making data products easily accessible and movable          examine how well and fluidly the ecology as a whole
  between displays will make transitions in work
                                                              supports the work tasks than to assume that disuse of a
  processes smoother, and help ensure that artifacts
  continue to be valuable as work progresses.                 tool is a failure of the technology to support the task.

Implications for Evaluation                                   Citations
The "success" of a large interactive display within a         [1] Huang, E.M., Mynatt, E.D., Trimble, J. Displays in
                                                              the Wild: Understanding the Dynamics and Evolution of
display ecology cannot be measured solely by whether a
                                                              a Display Ecology. Proc. of Pervasive 2006 (to appear).
steady state of use is reached. Because people
appropriate these shared displays as necessary, there         [2] Tollinger, I., McCurdy, M., Vera, A.H., Tollinger, P.
                                                              Collaborative Knowledge Management Supporting Mars
may be a natural ebb and flow of use that does not
                                                              Mission Scientists. Proc. of CSCW 2004, 29­38.
correspond to success or failure, but rather to the
                                                              [3] Trimble, J., Wales, R., Gossweiler, R. NASA's
dynamic nature of collaborative work processes. Success
                                                              MERBoard: An Interactive Collaborative Workspace
may also be found in the ease and extent of support that      Platform. In Public and Situated Displays (2003), O'Hara,
                                                              K., Perry, M., Churchill, E., Russell, D., (eds.), 18-44.