Cognitive Artifacts in Complex Work - Ambient Intelligence for Scientific Discovery - page 179

suggestions posted by other collaborators. A running electronic inventory of sig-
nificant papers retrieved by lab members over time would assist in providing
views of literature in constant reference or use, or to display visualizations of
research topics and areas explored within and across projects in a particular lab
or the research center.
By using electronic lab notebooks and automated data logging, streams of
experimental data and lab notes forward to all participants linking to a thread.
As ongoing findings are logged and discussed, and participants select key ideas,
an analogy agent searches online resources and presents analogous information
objects (excerpts or keywords in context) to the shared space. While few of these
“suggested analogies” may be perfect, the object is to facilitate the interpretation
process, and not an accurate analogy. The goal, using peripheral (unattended)
information display and visualization, might be to augment the social process of
discovery for collaborators on a continuous basis. A running stream of discovery-
oriented analogies, tuned to match research topics, may be attended to at any
time as a peripheral task.
Over time these information collaborations will integrate vast amounts of
data, translated to specific research objects that might be transferred and reused.
Individual interaction and feedback with the system tunes collaborative filter
algorithms and increases information specificity, serving to reduce data and noise
to minimize attentional overload for individuals and other collaborators. The
“intelligence” in such information collaboration develops as a network effect, as
each participant and lab's inputs, research data, and information resources are
stored, developed, and selected.
6 Conclusions
The chapter shows the nature of distributed cognition that can be found through
artifacts that participants develop and use to manage cognitive work. Two case
studies presented approaches to distributed cognition. Both studies described
highly-uncertain work domains with strong common collaborative objectives.
We show how cognitive artifacts reveal distributed cognition processes that sup-
port production and performance. The preference that workers show for physical
artifacts may be a factor in many distributed systems. Such artifacts reveal sig-
nificant boundaries in the ecology of practice that have yet to be reached or
surpassed by su ciently designed information technologies. Designers should
carefully approach known boundaries “marked” by artifacts to avoid disrupting
essential cognitive work, or to avoid outright failure of systems or their adoption
in complex work domains.
Advanced technologies such as ambient systems may eventually fulfill the
promise of a rich information ecology in research and medical practice. In order
to create such systems, we urge designers to involve the tangible in distributed
computing systems. The use of printed artifacts in distributing memory and
knowledge should be considered a “way in” to understand information tasks,
and not merely as printed output from searching electronic databases. The aug-
mented A-Book [21] demonstrates such an example of combining the paper ar-
tifact and electronic tools for the lab notebook.