Author:
Bao Yiyang,Paszkiewicz Maria,Krimmer Jonas,Freude Wolfgang,Randel Sebastian,Rockstuhl Carsten,Koos Christian
Abstract
We present an artificial-neural-network-(ANN-)driven concept for predicting transmission losses of 3D-printed freeform waveguides within a few milliseconds and with root-mean-square errors of less than 0.5 %. Our approach enables transmission-optimized trajectory design of photonic wire bonds during fabrication.