Ken,
playing off your post....
My senior design project when I was finishing at U of A was a self-conceived, designed and built automated light tracking system. (keep the bike in the lane and upright)
It used a rotating base with a mechanism mounted to it that allowed for pivoting the detector up and down. (Operator mechanical inputs steer the bike)
The rotation of the base combined with the pivot of the mechanism mounted to it provided a very wide range of motion. (Operator inputs control where the bike goes over a wide range of possibilities)
The relevant part of it all was I had to design and build the power supplies and the control circuits. (Operator's throttle/braking and mental processing of inputs)
Which meant I had to design, build, test and make work, a control system with adjustable sensitivity to account for light sources of varying intensity and a damping circuit to allow for smooth adjustments and tracking so as to avoid an under damped system that would result in a wildly oscillating tracking system or an over damped system that would not respond at all to lower level inputs. (Develop a reliable system for dealing with varying inputs to provide desired control effect and good motion control)
It all worked. Feedback and the proper processing and application of that feedback meant a smoothly operating system.
(Good processing of inputs and utilization of feedback can provide a stable, reliable control system - the operator)