Controlling the Economy
In September 2005, Béla Lipták wrote an article describing how one might design a control system for the global economy, as if it were an industrial process. ("Can an Automation Engineer Control the Economy?") Has he done any more work in this area during the last five years?
Dr. David Scott III
Béla Lipták responds:
Yes, David, unfortunately the predictions of my 2005 article turned out to be fairly accurate. We have once again learned that the principles of process control are applicable to all processes, and that since the introduction of "self-training" artificial neural network (ANN) algorithms, we can control even those processes that we do not fully understand. We do that in a way similar to the tennis player who, when he is learning to serve and without knowing anything about Newton or aerodynamics, just observes the consequences of his actions and, based on that information, he learns how to control his serve.
One of these days I will write on this subject again because it seems that we are approaching an even worse upset in our economic control loop than the one caused by the "housing bubble." This new upset is coming because it is no longer sufficient to control GDP, leading economic indicators and unemployment by manipulating the flow of money into the economic process by throttling taxation, interest and exchange rates and trade barriers, etc.
Today, new disturbance variables evolved that have both long dead times and large inertia, while our control loop is nearly saturated. What are these relatively new problems? These newer upsets include a) reset windup, b) loop interactions and c) the conversion of the controlled process from batch to continuous. Each of these would deserve a full article, but because of other commitments, that has to wait.
Reset wind-up occurs when the integral mode of a loop keeps integrating the error while the controller output is blocked and the loop is in manual. This wind-up in the economy control loop results in the accumulation of debt, which, if the interest rates are kept artificially low (under manual control) and if expenditures rise without increasing taxes, causes a collapse (not just a "bubble") of control.
Interactions between control loops occur when national economies are not cascaded to a common master controller. Such interaction occurs when a mostly feed-forward totalitarian economy, such as China's, interacts with a free-market economy that is basically under feedback control.
The third and probably the most difficult task is to convert the batch control system of the present (economy that was designed with the assumption that the availability of resources is unlimited), to a continuous control strategy, serving a stable economy based on inexhaustible energy resources.
One day I will write about these in more detail.
Editor in chief Walt Boyes' survey on what kind of basic training in instrumentation is available to our readers is still open. How much on-going training in basic instrumentation have you received? What format is it in? Where to you receive it? In the classroom? Online? On the job? Through ISA or vendor conferences? Are you satisfied with it? In what ways do you think it could be better? Please go to www.controlglobal.com/industrynews/2010/050.html to complete the survey. The survey is open until April 30, and the results will be released in our June issue in conjunction with our 2010 salary survey.