WJ Clancey — Visualizing Practical Knowledge
knowledge in office settings. We are constructing a simulated 3-d interface, which will
show people and their environment, indicating the changing locations of people and
materials. Making a model accessible to non-technical people is an important design
constraint. We intend to use such visualizations in facilitating conversations between
workers from different areas of a business in work systems design projects (Greenbaum
and Kyng, 1991).
A key reason for the ethnographic study of HMP-98 was to determine whether Brahms
could be used to model scientific fieldwork on Mars. The study revealed that the system
was broadly applicable, but significant additions to the modeling language are required to
describe characteristics expected for fieldwork on Mars: a model of the terrain and
climate, how people work while they are sitting in a “rover,” interactions with robots, the
interplanetary time lag, the affect of 38% gravity, etc.
We also plan to use Brahms as a framework for carrying out future analog studies at
Haughton-Mars in a rigorous way. By describing the various dimensions of
practice—people, activities, tools, environment—we can specify what aspects are
analogs, and what is being ignored. We can then question interactions that may result and
invalidate our observations relative to what would occur on Mars. For example, is it
necessary to wear space suits with stiff gloves in order to understand work on Mars?
With the aim of developing good simulations of scientific fieldwork, to be used in robot
design, automated health and safety systems, and communication networks, we will
continue the ethnographic study of future expeditions to Haughton Crater. We plan to
carry out experiments to understand: individual and disciplinary differences in field
activities; interpretation of plans and checklists; forms of “group memory” over phases;
communication allowing “virtual presence” between the field, base camp, and scientists
in the south (“Earth”); and how scientists interact with robots and workstations.
Conclusions
The ethnographic study of practical knowledge during HMP-98 exemplifies the
distinction between the study of practice (when, how, local interactions, and subjective
viewpoints) and the traditional human factors emphasis on process (what, why, flows,
and formal descriptions). Most process models focus on “stuff” (data and work products)
that moves through space and is transformed in time. Practice models focus on
15