Pattern Recognition in the Box Jellyfish Rhopalial Nervous System Mimicked by an Ensemble of Pulsed Coupled Oscillators

Abstract The rapidly expanding field of biomimetics emulates biological principles, distilled through evolutionary time, to solve problems in technology, medicine and material science. Information processing in neuronal circuitry of biological models may offer guidelines for future highly efficient computing systems exhibiting, as for example, low power consumption accompanied by excellent pattern recognition capabilities. Here, we consider the visual information processing in the nervous system of the box jellyfish Tripedalia cystophora as template for pattern recognition hardware to operate self-guiding robotic vehicles or automated driving assistants. LTspice XVII simulations of coupled relaxation-type oscillators, based on programmable unijunction transistors (PUTs), enabled the emulation of basal visual functionalities of the T. cystophora central nervous system. The transfer of such simulations in real hardware circuits including possible limitations are discussed.