Re: talking good

[Kathy and Ray La Belle]

I am afraid I can't help but respond to this one. I am trying to convince my
management that we need to do complex modeling. I am an engineer. I design
laundry machines. And believe me they are much more complicated than you would
think. Ever tried to figure out how and why clothes move? How should they be
moved to actually clean? We have some methods to do this. We run into
difficulties in the area of sub-system intergration. It is difficult to feed the
effects of the structure system into the fluid handling system (as an example).

I keep trying to convince people that the water molecules only know certain
things. If we model all the entities and all the rules of the groups of 
entities,
it may be possible to understand these systems in a deeper way.

I am an electrical engineer. I am working with some chemical and mechanical Phd
types. I have a hubby (Ray) that just happens to program.  How do I create a
model? How do I get information from the Phd types that actually means something
in the programming world? I am trying to create a simple model of diffusion in 
2d
as an exercise.

My first attempt with the Phd types got me equations. I explained these were
observed effects. This was definately NOT what I needed. And Ray could not use
this to create a model.

I went out and got some red, blue, and white poker chips and a table cloth with 
a
"grid" on it. Of course, the Phd types laughed at me. But after I explained that
the chips were entities and what did the individual entities know about their
environment, they decided "hey this is cool". I grouped the blue chips and
assigned a chip to each of the chemists. One was in the center of the group.
Another was at the outer edge. I called the game "diffusion" and said the goal 
of
the game was to get evenly diffused. And just to make it fun, they get to make 
up
their own rules :)

We discussed if how much each molecule would "know, feel, see" about its
environment. Would they know about every molecule (chip) in existance? Or only
the ones nearby?

 Initially, I started out with one entitiy type in a "empty space". Then we
discussed using red chips as the medium and the blue as the ones to disperse.
After we master "dispersion", we can make the "game" more complicated by
assigning properties to the different entities such as hyrdophoblic &
hydrophilic, miscible & immiscible, and even combining into new compounds. The
point is to learn to think as a particular entity (or swarm) would think. We 
know
if we are right if the emerged properties match all those complicated equations
we use on the macro level.

My counterparts went away and took the chips with them. After a few good
arguments in their office, they came back with rules I could actually use! The
main purpose of the exercise was to explain agent based modelling to them. I 
need
them to learn to think "as if they were the entity". I may have to get tinker
toys to show how molecular chains move ....

I decided to tell you about this experience because this "game" could be used 
for
all types of people. Alot of people need something visual to understand ideas.
This concept can be transferred into other fields.

                                                                            
Kathy
La Belle

Note: the opinions expressed by me do not represent the company I work for
(disclaimer)



                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================