Next generation photonic waveguide sensors enabled by machine learning

该项目旨在建立光子波导传感的下一个前沿，通过使用机器学习将复杂的传统光子波导/光纤传感器转移到智能检测算法中。多模光子波导的复杂性和不稳定性，传统上是传感的障碍，将被有利地用于训练用于传感的深度学习模型。预期的成果包括创造智能光子传感器，原则上可以使用任何光波导材料测量任何环境参数。它将创造新的迫切需要的测量能力，用于具有挑战性的恶劣环境，如极端温度和体内生化传感。

This project aims to establish the next frontier in photonic waveguide sensing, by using machine learning to shift the complexity out of conventional photonic-waveguide/optical-fibre sensors and into smart detection algorithms. The complexity and instability of multimode photonic waveguides, traditionally a hinderance to sensing, will be advantageously employed to train deep learning models for sensing. Expected outcomes include the creation of intelligent photonic sensors that can, in principle, measure any environmental parameter using any optical waveguide material. It will create new critically needed measurement capabilities for challenging harsh environments, such as extreme temperature and in-vivo biochemical sensing.