The concept of feedback is pretty simple. Take a problem, apply a solution and analyze the results. Repetition of the process with different solutions leads to what is known as a feedback loop.
The process is present in many walks of life, from finance and education to engineering.
Engineers study control theory to understand the dynamics of a system and how to adjust it to achieve a desired response.
In practice, it often involves the use of a proportional-integral-derivative control system — commonly referred to as a PID controller.
So I don’t bore you any longer with my geek speak, I’ll try to conjure an example of how it works in the real world. Monday mornings, I drive I-40 West to work in Research Triangle Park. I usually aim for a presumably safe vehicle speed of 65 mph on the stretch from the Wade Avenue merge to the I-540 spur — my input into our loop.
This is the part where the controller comes in: I could encounter a slow-moving car that requires me to change lanes. The car is a current error and is the proportional element in the equation. Along the same thread, I may have experienced sluggish traffic in the leftmost lane between the Wade Avenue merge and Airport Boulevard exit and decide to continue the remainder of the trip in a different lane. This accumulated error is the integral element. The final element is a rate of change — the derivative term — that is an effort to mollify future error. Taken together, these P, I and D terms come to define a lane-changing system I can use to achieve my goal of driving to work Monday at 65 mph.
Since you are now veterans in PID controllers, let’s try a sample problem to see if we can apply the fundamentals to a dilemma facing N.C . State.
The University aims to have 40,000 students by the fall semester of 2017. Achieving such a population will naturally require a certain population growth rate between now and then. Using a fairly simple growth rate model ( PGR = ( ln ( P2 ) – ln ( P1 ) ) / time in years ), we can approximate that growth at approximately three percent per year.
Simply, the University must maintain a three percent growth rate to reach the student population assumptions the campus infrastructure is being built to accommodate.
Here’s the problem: we have several errors. Our current error is a severe budget shortfall at the state level. The setback — not only this year, but over the last three — has left the University unable to pay for the tenured faculty positions vital to teaching this burgeoning student population. Likewise, the University has a track record of insufficient faculty hiring during the past generation — a clear pattern of an escalating student population with paltry faculty growth. Combining this accumulated error with the current error should lead a logical party to alter its actions. It’s the only possible action in successfully maintaining the growth trajectory.
Unfortunately, we’re not varying anything. We have not truly committed to a change of course that would lead to assuaging any of our errors. And thus, the University will not ever hit the desired targets. If it does, the execution will surely be in a manner unbefitting a top-flight research university.
Admittedly, PID systems are responsive in nature. Perhaps the University has a model that makes the math work. However, if the administrators have it, they sure are being secretive about how this going to work out.
I hope they’re hitting the books. It might be time for a little extra plug and chug.