cyberoid@milton.u.washington.edu (Robert Jacobson) (03/08/91)
>From the SIMULATION DIGEST, courtesy of Paul Fishwick:
Date: Thu, 7 Mar 91 19:54:17 -0500
From: "Paul Fishwick" <fishwick@fish.cis.ufl.edu>
To: simulation@ufl.edu
Subject: SPECIAL: Global Simulation
[[EDITOR: This is an excerpt from a recent publication by Takeshi
Utsumi. His organization is concerned with the use of computer simulation
as a mechanism for helping to resolve world conflicts. I have included
a part of the total document that deals with computer simulation. For
more complete information, you can reach the author through e-mail:
utsumi@cunixf.cc.columbia.edu -PAF]]
Excerpt from "Glasnost in the Global Village: A GLOSAS Project":
In July of 1986 a gaming simulation session was held at the World Future
Society "Crisis Management and Conflict Resolution" conference in New York
City. The simulation included a multi-media teleconference. Nearly 1500
persons in New York, Tokyo, Honolulu, and Vancouver took part. The demon-
stration, based on the FUGI world econometric model and databank, used com-
puter conferencing, compressed video, graphic image telecommunication and
audio. With socio-economic and political data from 140 nations, FUGI has
been used by the United Nations and various governments for economic fore-
casting. In this demonstration it provided the background tool for explor-
ing questions about U.S.-Japan trade (6).
The event was the brainchild of Takeshi Utsumi and his GLObal Systems Ana-
lysis and Simulation team. The idea for a global simulation as an instruc-
tional device was born in 1972. The following year Utsumi and the politi-
cal scientist, Robert Noel, conducted the world's first political gaming
simulation over a computer network (7). Utsumi then initiated the const-
ruction of an infrastructure for a GLOSAS simulation project on Energy,
Resources and Environment (ERE) (8). He called his simulations global
peace gaming (9). Using computers as well as a combination of advanced
telecommunication channels, the gaming was intended to enable experts and
laymen in geographically distant countries to collaborate in finding new
solutions to the problems underlying conflicts and war. It was to provide
training in crisis management, conflict resolution, and negotiation techni-
ques. Time has shown that it is particularly suited to explore problems
associated with environmental degradation and the promise of sustainable
development. This task is too large for government regulation, aid agen-
cies or development banks alone. Restoration of the environment must
engage all citizens of the globe.
The purpose of a GLOSAS interactive game is to help find appropriate alter-
native policies by establishing consensus among the participants. In this
respect such a game resembles a giant Delphi experiment. However, the
GLOSAS experiment is based on factual information stored in databanks and
guided by quickly produced polling results. As in a Delphi, each player
can change his/her mind after each poll. But the speed with which the
polling can be carried out means that in some respects the system dynamics
resemble those of a neural system. As new inputs come in, consensus might
change to accommodate them in an equilibrium seeking manner. After the
simulation progresses for a time results are shown on the display units
(Figure 1). The participants interact with results which fall within their
area of concern or expertise. For example, if pollution in Japan exceeded
a certain allowable level the Japanese participant watching it on her dis-
play unit could choose to stop the entire simulation (Figure 2). The
participants must then collectively try to find consensus on a new set of
alternative pseudo-policy parameters which will be executed until a new
crisis appears. Any public debate effects similar responses. But a global
simulation permits international participation while preserving the essence
of a village debate.
Global modelling and simulation studies have been conducted by various
groups and institutions to enhance the usefulness of international model-
ling and policy making since the early 1970s. However, with the advent of
global Value Added Networks (VANs) and standard communication protocols for
interconnecting various dispersed, dissimilar host computers, the potential
exists for ensuring the coordination of international efforts. This would
permit more credible simulation study than ever before. It is now possible
to combine existing technologies to make sophisticated and more holistic
explorations of various scenarios for solving global social problems. Many
small computers in different countries can be interconnected in a globally
distributed mode. Distributed network, information and data processing in
turn create a new environment for modelling and simulation including "peace
games" on the scale of Pentagon war games (10).
The technical details of the required infrastructure for a GLOSAS "peace
game" have been discussed elsewhere (11). Only an overview can be given
here. There are several advantages to a distributed processing network.
It means that many can communicate with many on an equal basis. Dissimilar
computer models running under different operating systems and interconnect-
ed by the VANs can all participate. Thus the system celebrates diversity
and sees "incompatibility" not as a "problem to be solved" but as an ele-
ment of a solution (12). The central databank does not have to be located
at a single location. Its contents can be echoed to many distributed data-
bank servers so that participants can retrieve updated world data by simply
accessing their local server (13). This "virtual central databank" appears
as a single databank to the user, though it consists of widely distributed,
locally updated parts which store the results of each game. Such model-
saving databases would be ready for reapplication with solutions modified
to suit local environmental or other parameters.
The complementary models created by experts of various disciplines and
countries, with their preferred simulation languages, methodologies and
geographically dispersed dissimilar computers can be interfaced and execut-
ed interactively and cooperatively, as required by the simulation. A dis-
tributed computer system with databases and simulation submodels can be
implemented by connecting subsystems in individual countries to act jointly
as a single system. In this way different interest groups, rather than a
single group of experts, can directly communicate with their counterparts
in other regions of the world, thus providing credibility and accuracy for
the databases and models which will be updated and maintained autonomously
by them. Only such holistic, whole earth views will answer the human need
to see simultaneously multiprocessed data run as time-compressed, years-
per-second films. For example, seeing at the same time advances in defo-
restation coupled with topsoil erosion and species depletion (14).
A major difficulty in the further improvement of global modelling is the
gathering of data that countries consider proprietary. A possible solution
is to have each country develop a national model in which sensitive data
could be used but "hidden." Only inputs from other national models and
outputs to the global model would be needed. There would have to be a
"global shell" to establish the characteristics of the national models. To
develop it would be quite an undertaking. But it should present difficul-
ties at least an order of magnitude smaller than those encountered in
gathering sensitive data from foreign countries. Furthermore, many nations
already have models which could possibly be modified to fit a global shell
such as Akira Onishi's FUGI (15).
Another problem is the inevitable time differences amongst participants
scattered around the world. However, this can be surmounted by using asyn-
chronous scheduling for the parallel execution of distributed simulation
submodels (16). Thus all models can be executed simultaneously and concer-
tedly via satellite and terrestrial telecommunication links.
= End of Excerpt =
[Moderator's Note: Regrettably, Fishwick's excerpt did not contain the
references cited in the text. Back to the primary source, researchers!
-- Bob Jacobson]