Flexible Neuromorphic Softness Sensory System

ABSTRACT Accurate softness sensing remains a challenge for artificial tactile systems, as conventional force‐based sensing misses key deformation cues. Biological perception instead uses active, displacement‐controlled exploration. Here we report a biomimetic neuromorphic tactile system that integrates iontronic softness sensing, spike‐based signal encoding and organic synaptic processing to enable softness sensory system. Displacement‐controlled operation of a hemispherical pressure sensor produces more discriminative temporal deformation signals than simple pressure sensors. Machine learning reveals interpretable temporal features and confirms high classification accuracy. The signals are encoded into spikes and processed by an organic synaptic device, enabling in‐sensor classification without digital post‐processing. Synaptic outputs demonstrate that material softness can be inferred directly from the excitatory postsynaptic current without complex post‐processing. This work establishes a sensor‐to‐synapse neuromorphic framework for softness perception.