In-sensor computing of sonar data using graphene ionic transistors

Sonar uses sound reflections for determining the presence of objects in water, but it requires complex circuitry to filter data, increasing power costs. In-senor computing addresses this by performing calculations in the sensing device. Graphene and Nafion-117 interfaces in a transistor structure have shown potential for in-sensor computing, where quasi-nonvolatile conductance states across a graphene channel are modulated by gate current pulses through the Nafion. Here, graphene/Nafion devices are designed and studied for sonar in-sensor computing applications. Both voltage pulses and continuous waveforms are applied, showing device response of frequencies 660–1,580 Hz for macroscale devices, 690–21,850 Hz for microscale devices, amplitudes down to 5 mV p , and animal sounds via a piezoelectric buzzer. Device durability is shown for temperatures up to 120°C and pressures up to 2.5 × 10 5 Torr. These results provide key insights for implementing in-sensor computing for sonar using flexible, two-dimensional materials.