Example of Simulation Success

["William H. Cutler" <72734.3452@]

I recently sent this to Ed Gallagher in reply to one of his postings, and he
suggested I post it to the whole group.

I recently heard a success story about the wedding of simulation with
experiment. The New Zealand Americas Cup team used simulation as a tool in the
design of their winning entry. They won by unprecedented two to five minute
margins, even though they spent far less money than the other entries.

The secret of success was having the computational fluid dynamicists and their
computers in the same building, right in the lofting room, where the naval
architects and boat building craftsmen were working. Every night the simulation
crew made their runs. Every morning they got together with the naval architects
and boat builders to decide what modifications of the boat to test that day.
The modifications were made, the boats tested, and data brought back that
afternoon as input to the next set of simulation runs. The simulators and
real-world folk got to trusting one another and working collaboratively.

In contrast, the other teams also used simulations, but they were done off-site
in some academic or industrial lab. The naval architects and experienced boat
builders dismissed the simulation results when they didnt make sense against
established doctrine or experience, so they never had any effect on the outcome.

This isnt exactly System Dynamics since the hydrodynamics doesnt involve
feedback in any major way, but it does illustrate the value of simulation
closely coupled to test.

If youd like a reference to this work, e-mail me at the address below and Ill
get it for you.

Regards,
Bill Cutler
4114 Park Blvd.
Palo Alto, CA 94306
415-493-8715
72734.3452@compuserve.com

[Jim Hines]

Concerning Bill Cutlers Posting #SD0185.

After offering a terrific example of how model-based insight moved into
practice, Bill notes that its "not exactly system dynamics".

Id just like to mention that my experience with similar situations
involving system dynamics strongly supports what Bill is saying. The
close participation of decision makers (boat designers and builders with
the simulation) increases signficantly the value of simulation to the
decision makers.

I think this is for a number of reasons:
1) The decision makers come to understand the modeling much better;

2) the communication is better because there is a convergence of
terminology (simulators talk more like the decision makers and the
decision makers borrow words);

3) the model is much better because it reflects what decision makers
know about the structure of the organization.

4) decision makers are more comfortable with "listening" to the model
because they know it contains their thinking. (note this does not mean
that they BELIEVE the model, but rather that they will think about what
the model is saying (i.e. the logical reason that the model behaves as
it does).

5) There really is not much of a separation between the simulators and
the decision makers. The discision makers contribute to the model, and
the simulators contribute to the decision.

Jim Hines
JimHines@Interserv.Com
LeapTec and MIT