|

Collaborating to Expand Observational Data Resources
The Space Physics and Aeronomy Research Collaboratory
(SPARC), funded by KDI, brought together a team of space physicists, computer
scientists, and behavioral scientists. Among them was Dr. Thomas Finholt, a
co-principal investigator on the project and research associate professor at
the School of Information at the University of Michigan. Dr. Finholt is also
director of the Collaboratory for Research on Electronic Work.
The KDI-funded aspect of the project picked up where a
previous project, called the Upper Atmospheric Research Laboratory (UARC), had
left off. According to Dr. Finholt, "UARC was an atmospheric science project
whose main purpose originally was to obtain data from a remote observatory in
Greenland and display it on computer screens of scientists, mostly in North
America and in Western Europe. Then, over the course of that project, it became
possible to bring in other kinds of instrumentation, including two European
instruments in the Norwegian Arctic and instruments in the U.S. and South
America."
In funding SPARC,
the National Science Foundation (NSF) wanted to create a collaboratory of
revolutionary scope and power. The project team aimed to expand the field of
view of available online instrumentation by combining observational
capabilities with computational simulation capabilities. The number of
real-time observational data resources that scientists could view in this
collaboratory environment expanded dramatically. "At the end," says Dr.
Finholt, "there was something like 390 observational resources, including
global chains of magnetometers, interferometers, radars, earth orbiting
resources, the Polar spacecraft, and more."
A key difference between the UARC project and SPARC was
that, with SPARC, the team moved away from a very specialized software
environment and started to use the computer language Java, which made the
collaboratory platform more accessible to scientists all over the world.
According to Dr. Finholt, his role in SPARC was to focus "on
what kinds of collaborations and collaborative practices would emerge in a
setting where people were not co-located. There was a great deal of enthusiasm
around these collaboratory projects and the belief that they would accelerate
knowledge creation and discoveries. Although that was the hope, nobody really
understood how that was going to work. So that was one of the things that we
were looking at, along with using our findings to try to enhance the
capabilities of the system to achieve the goal of faster discovery."
Today, Dr. Finholt
is working on an NSF-funded collaboratory for earthquake engineering
simulation called the George E. Brown, Jr. Network for Earthquake Engineering
Simulation (NEES). This project is NSF's first attempt to build a national
scientific infrastructure that is intended for use in a distributed and remote
fashion. NEES plans to link together some 18 brand new earthquake engineering
laboratories.
Dr. Finholt says that his involvement in this work was a
direct follow-on from his KDI-funded project. "Our work on UARC and then SPARC
made us visible to the National Science Foundation, so at the time that they
were bringing the various elements together, they brought us in as part of what
they described as the dream team to build this system for the
earthquake engineers."
Another result of the KDI-funded project was that the
technology that was developed in the context of SPARC became the basis for a
public domain courseware environment that is currently being elaborated by the
NEES project.
A third spin-off is a startup company called Arbor Networks,
founded and run by scientists who were involved in the SPARC project. "Arbor
Networks is one of the leading purveyors of algorithms and techniques that
Internet service providers use for detecting incipient service attacks," says
Dr. Finholt, "so it would have obvious value given the nature of the Internet
these days." Arbor Networks grew out of the UARC and SPARC experience. "In the
creation of the collaboratory we were distributing many samples of the software
all over the world," says Dr. Finholt, "and that gave these guys a great test
bed for looking at network performance and monitoring network characteristics
across a broad base. From that, they developed these algorithms that are now
being used to produce this product."
To learn more about SPARC, visit the Web site at http://intel.si.umich.edu/sparc/.
For more information on the NEES Consortium, visit their Web
site at www.nees.org.
To learn more about Arbor Networks, go to their Web site:
www.arbornetworks.com.
Back to Top of
Page
|