Mechanosensing of Stimuli Changes with Magnetically Gated Adaptive Sensitivity

Inspired by biological sensors that characteristically adapt to varying stimulus ranges, efficiently detecting stimulus changes sooner than the absolute stimulus values, we propose a mechanosensing concept in which the resolution can be adapted by magnetic field (<i>H</i>) gating to detect small pressure-changes under a wide range of compressive stimuli. This is realized with resistive sensing by pillared <i>H</i>-driven assemblies of soft ferromagnetic electrically conducting particles between planar electrodes under a voltage bias. By modulation of <i>H</i>, the pillars respond with mechanically adaptable sensitivity. Higher <i>H</i> enhances current resolution, while it increases scatter among repeating measurements due to increased magnetic structural jamming between colloids in their assembly. To manage the trade-off between electrical resolution and scatter, machine learning is introduced for searching optimum <i>H</i> gatings, thus facilitating efficient pressure prediction. This approach suggests bioinspired pathways for developing adaptive stimulus-responsive mechanosensors, detecting subtle changes across varying stimuli levels with enhanced effectiveness through machine learning.