Taming dynamic complex media for imaging, communications and photonic computing

The turbulent atmosphere, flexible optical fibres, and living tissue, are all examples of opaque materials that change in time.

Light propagating through these moving complex media is subject to intricate dynamic scattering effects: transmitted optical fields are unrecognisably distorted in space and time. Such phenomena permeate real-world imaging and communications technologies.

They prevent microscopy deep into living tissue and hinder both free-space and fibre-optic data transmission.My vision is the realisation of optical technology that automatically configures itself to dynamically cloak unwanted scattering effects of any opaque medium it is coupled to making it possible to look through to an unobstructed view of scenes on the other side.

The multimodal nature of this challenge calls for a raft of new sensing and light control techniques.

The aim of this proposal is to develop these technologies.I have begun to take some of the first experimental steps to achieve these goals: I have developed a new form of adaptive optics that suppresses temporal rather than spatial light distortion.

I have also implemented prototype adaptive optical inverters, capable of all-optically unscrambling light propagation through multimode optical fibres (MMFs) supporting ~30 modes. My project builds on these concepts, and is delivered through 3 overarching objectives:1.

Real-time monitoring of spatio-temporal scattering in dynamic complex media: Creating dynamic digital twins of complex media that evolve in real-time.2.

All-optically unscrambling dynamically scattered light: Building new types of programmable optical circuit driven by our digital twins to cloak spatio-temporal scattering effects.3.

Applications: reversing the scattering effects of MMFs and atmospheric turbulence; enabling transmission of quantum entangled photons through noisy channels; and finally, investigating the broader implications of our technology to hybrid opto-electronic neural networks.