|
2007
Summer Computer Simulation Conference (SCSC'07)
KEYNOTE SPEAKER
Team-oriented,
Design-focused Simulation Tools:
Massively Multiplayer Games in the Engineering
Workplace
Prof.
Roger A. Dougal, Thomas
Gregory Professor of Electrical Engineering. University of South Carolina.
Traditionally,
simulators have been far more useful for analysis or experimentation than
for design or synthesis. Simulations were called on to support
understanding of the dynamics of systems that were too complex or
intractable to compute via pencil and paper. They supported hardware-less
“experiments” that were carefully and fully defined apriori, and which
could be iterated many times, either to provide new understanding of the
system, or to find a better variant of a particular design. Later, HDLs helped
merge the simulation process with the design process; they provided a
description of the system that was at once executable to describe how the
system behaved, and compilable to define how it was built. HDLs were
especially successful in digital systems, where it became possible to
design, test, debug, and then construct, huge systems, containing
thousands, millions, nearly even billions of parts on a
successful-first-production-run basis.
Though large, these systems were relatively simple because the numbers
of unique components were small. Now the world needs new tools that will
permit the design of huge, and hugely-complex, systems that contain large
numbers of unique components, many of them not yet fully defined.
Simulation tools must evolve to support massively-multiplayer design teams
running simulations using multi- and variable-time steps, with models
having multi- and variable- levels of abstractions (resolutions),
and that are controlled and defined by multi- and variable- numbers of
players. In short, simulator developers can stand to learn a thing or two
from the online gaming community. Beyond that, simulators must also
increasingly interact with hardware in the loop, and with other simulators.
They must run as parts of embedded controls, and they must predict the
ranges of performances expected from not-fully-defined systems operating
under uncertain conditions. These needs present many interesting R&D
opportunities.
Bio
Roger Dougal is
the Thomas Gregory Professor of Electrical Engineering at the University of South Carolina. His early education
was in physical electronics -- an area where simulation was an
indispensable aid to understanding of device dynamics. He has spent the
past decade as leader of the Virtual Test Bed project, addressing issues
related to design-oriented simulation of multidisciplinary dynamic systems.
Return to SCSC 2007 Home Page
Return
to SummerSim'07 Home Page
Return to SCS Home Page
Copyright
2006, SCS. All Rights Reserved.
|
