proposal: civilians

From: Michael Brenner (brenner@informatik.uni-freiburg.de)
Date: Tue 09 Nov 2004 - 16:25:05 GMT


Hello everybody,

here is another proposal not yet covered by Cameron's list.

We propose a new model of civilians to capture a number of features
missing in the current simulation:
* realistic numbers of (unhurt) people (from an unrealistic few dozens
of civilians to a several hundreds or thousands).
* phenomena such as mass panics, fleeing from fires, crowds blocking roads
* a new job for police agents: directing streams of civilians to refuges

To this end we propose a network-flow-based simulation of civilians (at
least those unhurt and able to move; hurt civilians could still be
modeled as before). A number n of civilians on a road or in a
building would simply be represented by storing n in the respective node
of the simulation. Depending on parameters like proximity of fires,
smoke direction, number of agents on adjacent nodes, movement direction
in the last cycle etc. it would be easy to compute movements of (groups
of) civilians to adjacent nodes as simple updates of these numbers for
the next cycle (like in a cellular automaton.)

Police agents would be one important factor influencing the equations,
thus being able to *control* the flow of people if
positioned at appropriate places on the map. This would pose an
interesting distributed coordination problem to the one group of agents
that is currently least important.

This model does not represent every single civilian as an agent and thus
can scale up to realistic numbers without computational overhead. Using
a cellular automaton-like model will result in a complexity that depends
mainly on the number of nodes on the map,

Integration with the traffic simulator could be achieved by regarding,
say, 10 civilians as a (dynamically) blocked lane. The movement of the
civilians would not have to be computed by the traffic simulator itself,
but by the cellular automaton.

Implementing such a feature should not be too difficult and could be
easily be done until March. It would include a tool for random
generation and storage of an initial distribution of civilians over a
given map (according to size of buildings etc.)

        michael & alex
        ResQ Freiburg

-- 
Michael Brenner 		Phone: (+49) (761) 203-8226
Institute of Computer Science	Fax:   (+49) (761) 203-8222
Albert-Ludwigs-University	mailto:brenner@informatik.uni-freiburg.de
Am Flughafen 17, Geb. 52	http://www.informatik.uni-freiburg.de/~brenner
D-79110 Freiburg, Germany



This archive was generated by hypermail 2.1.3 : Tue 09 Nov 2004 - 16:25:17 GMT