Dissipation rates from experimental uncertainty

Abstract: Active matter and driven systems exhibit statistical fluctuations in density and particle positions, providing an indirect indicator of dissipation across multiple length and time scales. Here, we quantitatively relate these measurable fluctuations to a thermodynamic speed limit that constrains the rates of heat and entropy production in nonequilibrium processes. By reparametrizing the speed limit, we show how to infer heat and entropy production rates from directly observable or controllable quantities. This approach can use available experimental data and avoid the need for analytically solvable microscopic models or full time-dependent probability distributions. The heat rate we predict agrees with experimental measurements for a Brownian particle and a microtubule active gel, which validates the approach and suggests potential for the design of experiments.