Skip to content
PaperFren

Information thermodynamics · Maxwell's demon

A feedback engine turned every bit of noisy measurement into work, in simulation

Evidence: PreliminaryOne study or a small sample; not yet replicated. What the labels mean

Study published Jan 1, 2020. PaperFren added this explanation Sep 26, 2026.

Save this development or follow its topic to track what changes.

Short answer

With reversible feedback, work equals kT times the information gained, so noisy measurements are not wasteful, just less informative per shot.

What happened

The authors modelled an overdamped particle in a harmonic trap, measured with Gaussian error, then confined reversibly according to the measurement and expanded quasistatically. The confinement cost zero work on average and the expansion returned kT times the mutual information, an efficiency of 1. Simulated averages over 200 cycles matched the prediction, though individual cycles scattered widely. Two measurements with twice the error variance yielded the same information and work as one precise measurement.

Why it matters

Maxwell-demon style engines are usually discussed with perfect measurements. Showing that measurement error lowers the information gained but not the efficiency of converting it clarifies where the thermodynamic cost of imperfect sensing actually sits.

Evidence

Study type
Theory with Langevin simulations
Sample
Simulated trajectories averaged over 200 cycles of 10 measurement steps
Journal
Entropy · peer reviewed
Replication
No experimental test in this paper
Limitations
No experiment; idealised control; part of the paper reviews known protocols rather than presenting new results.

What this connects to

Sources

The one study this explanation is built from, by the role each plays. Every source links to PaperFren’s explanation of it and to the original paper.

Primary study

  • Can sloppy measurements still power a perfect information engine?

    Dinis L, Parrondo JMR · 2020 · Entropy (Basel, Switzerland) · 3 citations

    A simulated particle engine turns every bit of information from its measurements into work, and noisier measurements can match precise ones simply by measuring more often.

    What it does not show

    This is a theoretical and simulated result; no experiment was performed, and the protocol requires very fine, instantaneous control of the trap stiffness and centre that may be hard to realise. The authors note that imperfect tuning would introduce dissipation, which they only discuss qualitatively. Part of the paper reviews known reversible-feedback protocols rather than presenting new results.

    PaperFren explanationStudy with cards and a quizOriginal paper (DOI)cc by

Before

Imprecise measurements would be expected to make an information engine inefficient, dissipating part of whatever they learn.

Now

In this idealised protocol, efficiency stays at 1 regardless of error; noise only lowers information per measurement. It is theory and simulation, requires instantaneous fine control of trap stiffness and centre, and dissipation from imperfect tuning is only discussed qualitatively.