Each correction needs at least one full feedback-loop cycle

What it is

Closed-loop correction doesn't happen continuously and seamlessly at every instant — it advances in discrete rounds: only once the full chain of "sense the current state, compare it against the target, generate a new command, conduct it to the muscle and produce a change in movement" has completed does one round of correction count as done, and the next round can't begin until this one has finished. The minimum time it takes to run through that entire chain once is called the feedback loop cycle time, and it sets a hard lower bound on how fast a single round of correction can happen — there's no way to correct any faster than that.

Why it happens

This lower bound comes from the multiple mandatory steps built into the process: sensory information takes time to travel from where it arises to where it's processed centrally; processing it, comparing it against the target, and generating a correction command takes computation time; and conducting that command back out to the muscle to actually change the movement takes conduction time again. These stretches happen sequentially and can't overlap or be skipped, and adding them up gives the minimum duration of one round of correction — measured empirically at roughly on the order of a hundred milliseconds. This means that even if feedback itself is available instantly and the error signal is perfectly clear, the body has no way to compress this entire chain any shorter — the limit comes from the speed of physiological conduction itself, not from willingness to respond or level of skill.

Studying it

A common way to measure this cycle sets participants a series of target movements with different total durations, and tracks how the number of decelerate-adjust segments near the end of the movement grows as total duration increases: if each fixed increment of added time consistently allows room for one more round of correction, that fixed increment is the measured value of the loop cycle. Another approach artificially inserts a one-off feedback perturbation mid-movement and measures the time between the perturbation appearing and the first identifiable compensatory adjustment showing up in the trajectory — that time gap directly corresponds to the time needed to complete one round of sense-compare-generate-conduct.

Where it stops holding

This cycle duration isn't perfectly identical across every person or every feedback channel: visual, tactile, and proprioceptive feedback each conduct at different speeds, so the loop cycle shifts depending on which channel is being relied on; the same person's loop cycle also fluctuates somewhat with fatigue and attentional state. This cycle also describes the minimum time a single round of correction requires — it doesn't mean every round in an actual task takes exactly that long; if the error is already small enough, correction may finish early or a further round may not be needed at all.

Applying it

• When estimating how many rounds of correction a fine-precision operation still needs before settling, treat total correction time as roughly "loop cycle time times expected number of rounds," rather than as a continuous process that can be interrupted and can respond instantly at any moment; timeout durations or loading-animation lengths set on this basis will track actual elapsed time more closely.
• For controls involving continuous fine adjustment (a precision slider, a resize handle), avoid rushing to show the next state update or trigger new feedback before a full loop cycle has elapsed — pushing updates faster than that misaligns with the user's own actual correction rhythm and adds interference rather than help.
• How to check: give the same users several groups of fine-positioning tasks with different total durations, log the relationship between the number of decelerate-adjust segments in the trajectory and total duration, and use that relationship to work backward to the measured loop cycle time in that scenario — then use it to calibrate related feedback refresh rates or timeout thresholds.

Related

• Same group: A8.07.1 Closed loop relies on continuous feedback to close in on the target step by step · A8.07.3 Feedback delay beyond the loop cycle causes overshoot and oscillation · A8.07.4 Final landing accuracy is set by the closed-loop phase · A8.07.5 Closed loop degrades to open loop when the feedback channel is lost
• Nearby: A8.01 The Composition of Reaction Time · A8.08 Two-Phase Structure of Target Acquisition
• Search terms: feedback loop cycle time · visual feedback delay · correction interval