Heterogeneously integrated Multi- material Photonic Chiplets for Neuromorphic Photonic Transfer Learning AI Engines
AI technologies offer substantial benefits for automating, accelerating and maximising the efficiency of repetitive, hazardous or labour-intensive tasks, especially in sectors like manufacturing. However, these technologies are complex to develop, with machine learning in particular requiring significant time and resources. The EU-funded HAETAE project aims to revolutionise computing by developing a multi-material photonic integrated circuit, or PIC, technology platform and a photonic neural network architecture, designed to integrate with transfer learning methods. This approach will involve the creation and combination of advanced materials, optical circuit architectures and printing processes, optimising both efficiency and cost.