A robotic arm operates in a controlled factory where the arm's mechanical properties are precisely characterized and no unexpected external forces act on it. An engineer proposes adding a closed-loop sensor to improve performance. What does control theory suggest?
AClosed-loop is always better, so the sensor should be added regardless of environment
BOpen-loop is preferable here — the predictable environment and accurate model eliminate the main advantage of feedback
CClosed-loop is necessary because open-loop systems cannot achieve precise positioning
DThe choice does not matter because both architectures perform identically in controlled conditions
The key insight is that the choice between open-loop and closed-loop is fundamentally about uncertainty. Closed-loop feedback earns its added complexity by correcting for disturbances and model errors. When the plant model is accurate and disturbances are negligible, those advantages disappear — and the open-loop design wins by being simpler, cheaper, and free from the stability risks that feedback introduces. Adding feedback to an already well-characterized, stable system can introduce instability without providing meaningful benefit.
Question 2 Multiple Choice
What is the fundamental mechanism by which closed-loop control handles unpredictable disturbances that open-loop control cannot?
AIt uses a more powerful actuator that overwhelms disturbances before they affect the output
BIt predicts disturbances using an internal model and preemptively cancels their effects
CIt continuously measures output error and adjusts the control input to drive that error toward zero
DIt increases system bandwidth so disturbance effects decay faster
Closed-loop control works by sensing what actually happened and correcting for it — not by predicting or overpowering disturbances. The error signal (setpoint minus actual output) is computed continuously, and the controller adjusts its input to reduce that error. This works even for disturbances the designer never anticipated, because the mechanism responds to outcomes rather than causes. Open-loop has no such mechanism: it applies its command regardless of what the plant does, so any unmodeled disturbance goes uncorrected.
Question 3 True / False
Closing the loop generally improves stability — a marginally stable open-loop plant will become more stable once feedback is added.
TTrue
FFalse
Answer: False
This is a critical misconception. Feedback does not automatically stabilize — it can destabilize. If controller gains are too high, feedback causes overcorrection: an error triggers a corrective input, which creates a larger error in the opposite direction, which triggers a still-larger correction, leading to oscillation or divergence. A plant that is stable in open-loop can be made unstable by improperly tuned closed-loop feedback. Gain and phase margins exist precisely to quantify how much margin separates stable operation from instability in a feedback system.
Question 4 True / False
A toaster with a timer is an example of an open-loop system: it cannot compensate if the bread is already partially toasted or if the heating element degrades over time.
TTrue
FFalse
Answer: True
Correct. The toaster timer applies a fixed duration regardless of the actual toast darkness — it has no sensor to measure the output. This means it cannot compensate for varying initial conditions (bread already toasted) or plant changes (weakening heating element). This brittleness is the defining liability of open-loop systems: any deviation of the actual plant from the assumed model goes uncorrected. A feedback-based toaster would sense bread color and stop when the desired darkness was achieved, but would require a sensor and introduces the complexity of a control loop.
Question 5 Short Answer
Why does adding feedback to a control system introduce the possibility of instability that was absent in the open-loop design?
Think about your answer, then reveal below.
Model answer: Feedback creates a closed loop in which the output influences the input. If controller gains are too high, the system overcorrects: an error triggers a corrective input that drives the output past the setpoint in the other direction, creating a larger error, triggering an even larger correction, and so on. This self-reinforcing oscillation is impossible in open-loop because there is no loop — the controller applies its command without reference to what the plant does. Feedback trades the brittleness of open-loop (no correction for disturbances) for the stability risk of closed-loop (possible runaway oscillation if gains are wrong).
The mathematical condition for instability is that the loop gain and phase combine to create net positive feedback at some frequency — meaning perturbations at that frequency grow rather than decay. Gain and phase margins measure how far the system is from this condition. Neither concept applies to open-loop systems, which have no loop to go unstable. This is why gain and phase margin analysis is exclusively a closed-loop concern.