Balancing and Reinforcing Loops

[bbens@MIT.EDU]

WRT (SD0262)

Please let me clarify the point I made earlier without distracting
people from the focus of this group and system dynamics in general.

> The feedback loop structure of
> the coffee cooling and a thermostatically controlled environment are
> identical (both negative loops). The only difference is that in one
> case the set point (goal) of the system is determined by the room
> temperature and in the other it is set by a human being.

Here is one reason I disagreed. Supposed that the hot coffee is mixed
with iced-cold water in proportionate volume, say 1:1. Is the coffee
going to cool down such that its temperature is the same as that of
the ice-cold water? No. In fact the temperature of the mixture will
be intermediate, between the original temperature of the coffee and
that of the ice-cold water. The process is a balancing process and
the loop is a "balancing loop." It is one reason that we see a
scale, in the Fifth Discipline, to describe a balancing loop.

For those who are interested, the temperature near the vicinity of
the cooling coffee actually rises while the coffee is cooling.
However, energy diffuses into air that has a much bigger thermal
capacitance than that of a cup of coffee. The final state is that
the coffee is "cold" yet the room temperature remains the same as
it was.

The thermostat example, on the other hand, is indeed a closed-loop
control system in which measured temperature is compared to the
setpoint so that the error is used to drive the furnace. It is clear
that information is fed back to the controller and an action is taken
to minimize (eliminate) error.

Similar notion also exists in daily usage of the term feedback. When
people give us feedback on something, they generally perform some
"measurement" or assessment and provide the information back to us.
After receiving such a feedback, we compare it with our own
"standards" and then act to minimize the gap between whats measured,
by various standards, and our goal.

> So the issue of what is the
> controller and what is the plant (to use control theory jargon for the
> policy levers and the physics of the system) is ambiguous, and depends
> on the model boundary.

Very true that its not easy, worse its ambiguous. But this
approach will, hopefully, in the end help identify levers and policy
implementation--only if systemic improvement is the goal of the
effort. This is way modeling, in any discipline, is both art and
science. Its an art because the model needs to be simple enough
to generate insights yet to be complex enough to produce sensible
behavior.

The issue of model boundary is another interesting topic. I think
model boundary depends on our level of structural simplification:
what we include or exclude in the model or the focus loops in our
simplified structure. Discussions on model boundary, I think, will
be very valuable in increasing our understanding and modeling
proficiency.

Thanks, I learnt a lot in the process.
-Benny Budiman-
Graduate Student
bbens@MIT.EDU