Clash Of Realities - Transims

rdg · March 7, 2008

Hi,

I've been visiting the "Clash of Realities"-Conference [1] in Cologne.

This is a games-study meeting - rather theoretical but some of the panels had interesting topics

Claus Pias's [2] speech about Simulations and Games was very inspiring.

He was showing screenshots of TRANSIMS [3] - a simulation that handles ~ 1.6 mil agents.

Their behavior is modeled after census data.

The actual purpose is to simulate traffic and - in a modified version -the spread of epidemics.

Maybe such systems can be used for CGI crowds in the future as well.

It also reminded me of this discussion at the SESI-forum about "proceduralism vs simulations".

Where some - if I understood them rightly - argued simulations to be the worst case scenario.

This is owed to the fact that currently not anything can be solved by calculations and a simulation of fluids and weather is the only way to handle such complex systems.

What might be interesting for fluids is: Some of the models that are used to simulate weather are inconsistent with current physics and thermal sciences. They know they are 'wrong' but these models actually result in reliable data.

I also enjoyed the talk of my sandbox buddy J.R. Kuecklich [4] about cheating as a scientific tool.

Together with the discussion about "games as applications - applications as games" this opens a new field.

This is a rather theoretical approach to a practical software like houdini, but isn't CGI nothing but cheats.

And if so what might be cheats in a cheating system and how can they be used?

Georg

[1] http://www.fh-koeln.de/clash-of-realities

[2] http://homepage.univie.ac.at/claus.pias/

[3] http://transims.tsasa.lanl.gov/

[4] http://playability.de/