(9j) Bayesian Inference and Design for Acoustic Levitation and Propulsion

The propulsion of micro- and nanoparticles using ultrasound is an attractive strategy for the remote manipulation of colloidal matter using biocompatible energy inputs. By controlling the three-dimensional shape of the particle, it is possible to control its levitation and directed motion within simple acoustic fields. Nevertheless, there remain significant discrepancies between the experimental observations of particle motions and the predictions of hydrodynamic models. The reconciliation of these differences can benefit from the use of Bayesian inference and experimental design within a fully automated framework that iterates the processes of inference, design, and observation. Here, we demonstrate how this framework can be applied to accurately quantify the frequency dependent acoustic velocity within a resonant chamber using a minimal number of experiments. We compare the performance of such optimal experiments to alternative designs based on common simplifying assumptions (e.g., linear models). More generally, this normative framework represents a powerful strategy for automating the scientific method in the context of active colloids when strong guiding models are available.