【Control】⏱️ 05. Why “Timing-Only Control” Fails Under Friction Aging

topics: [“control engineering”, “PID control”, “reliability”, “simulation”]

Introduction

A control system is not acceptable simply because it “moves.”
In long-term operation, what breaks first is often when it moves—its timing.

In this article, we compare:

on a plant subjected to friction aging equivalent to 1000 days, and show a case where:

A control strategy that appears successful actually collapses as a reliability control.

🎯 Objective of the Experiment

The objective is not accuracy improvement.

Can response timing (Δt) be preserved under long-term degradation?

This question is evaluated explicitly from a Reliability perspective.

🧪 Experimental Conditions (Key Points)

Degradation Model

Controllers Compared

📉 Results: Timing Degradation (Δt)

The comparison result is shown below.

Timing degradation under friction aging

At first glance, AITL appears to be a “success.”

⚠️ But What Actually Happened

Closer inspection of the AITL response reveals:

In other words:

Timing was preserved at the cost of controllability.

❓ Why Did This Happen?

The reason is straightforward.

As a result:

Timing-oriented retuning alone can collapse motion authority.

This is not a bug—it is an inevitable design failure.

🧠 What Reliability Control Really Means

This result makes one thing clear:

Reliability Control ≠ optimization of a single metric

At minimum, the following must be considered jointly:

Reliability Control is therefore a constrained multi-objective design problem.

🧩 Design Lessons Learned

That’s All for This Article

In this article, we clarified:

Next, we will move on to:

Designing a Reliability FSM with amplitude and saturation constraints

🔗 References