Research

We are pioneering a new class of photonic systems that merge optical physics with the adaptability of machine learning to unlock powerful, high-efficiency information processing. Moving beyond conventional design constraints, our research leverages rich dynamics in interferometric networks, optical gain media, and neuromorphic architectures to build systems that learn, adapt, and optimize.

By using physics-informed machine learning, we develop models and devices that push the boundaries of optical functionality while offering scalable, layout-aware solutions for complex computational tasks. These photonic learning machines are the backbone of a foundational shift toward intelligent, adaptive optical platforms.

As photonic systems evolve toward greater complexity, we are rethinking integrated device design from the ground up. We develop strategies with extreme computational efficiency through inverse design, adiabatic structures, and data-driven optimization.

These approaches enable rapid and reliable creation of photonic components with finely tuned spectral and polarization responses, making it practical to design systems that operate across broad spectral bands and withstand fabrication imperfections. By bridging physical accuracy with scalable design tools, we’re opening the door to a new era of nanophotonic innovation.

Our research addresses the critical challenge of scaling next-generation systems: the integration of active components like low-noise amplifiers and lasers within silicon-based platforms.

By harnessing rare-earth-doped gain media, we enable compact, energy-efficient devices with exceptional frequency stability. These technologies form the foundation for highly integrated systems in ultra-fast communication, precision sensing, and signal processing, reshaping what’s achievable with silicon photonics.

Funding & Support

Research at Photonic Architecture Laboratories is graciously supported by the following organizations.