Traditional design methods (those in which a designer proposes and refines a solution) are rapidly being superseded by algorithmic alternatives. There, a computer is given only a design goal and a set of manufacturing constraints and is allowed to create any permissible solution that does the job. In nanophotonics, this is now routinely applied to photonic structures such as waveguides and demultiplexers, where the design goal could be, for example, maximised transmission at a particular wavelength. A step towards extending this to the quantum realm of single-photon interactions with atomic systems was recently taken, in which the ideas of macroscopic quantum electrodynamics are mixed with those of adjoint optimisation. In this talk, I will introduce the formalism of quantum electrodynamical inverse design, before showing a few recent proof-of-principle applications to processes such as resonant energy transfer.
|