Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2505.03354

Physics-Informed Neural Networks in Electromagnetic and Nanophotonic Design

Omar A. M. Abdelraouf·Abdulrahman M. A. Ahmed·Emadeldeen Eldele·Ahmed A. Omar
View original source ↗Read source document ↗Export metadata ↓

Abstract

The fusion of artificial intelligence (AI) with physics-guided frameworks has opened transformative avenues for advancing the design and optimization of electromagnetic and nanophotonic systems. Innovations in deep neural networks (DNNs) and physics-informed neural networks (PINNs) now provide robust tools to tackle longstanding challenges in light scattering engineering, meta-optics, and nonlinear photonics. This review outlines recent progress in leveraging these computational methodologies to enhance device performance across domains such as dynamic light modulation, antenna design, and nonlinear optical phenomena. We systematically survey advancements in AI-driven forward and inverse design strategies, which bypass conventional trial-and-error approaches by embedding physical laws directly into optimization workflows. Furthermore, the integration of AI accelerates electromagnetic simulations and enables precise modelling of complex optical effects, including topological photonic states and nonlinear interactions. A comparative evaluation of algorithmic frameworks highlights their strengths in balancing computational efficiency, multi-objective optimization, and fabrication feasibility. Challenges such as limited interpretability of AI models and data scarcity for unconventional optical modes are critically addressed. Finally, we emphasize future opportunities in scalable multi-physics modelling, adaptive architectures, and practical deployment of AI-optimized photonic devices. This work underscores the pivotal role of AI in transcending traditional design limitations, thereby propelling the development of next-generation photonic technologies with unprecedented functionality and efficiency.

Explore related subjects

physics.opticsphysics.comp-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Omar A. M. Abdelraouf, Abdulrahman M. A. Ahmed, Emadeldeen Eldele, Ahmed A. Omar. 2025-05-06. Physics-Informed Neural Networks in Electromagnetic and Nanophotonic Design. https://arxiv.org/abs/2505.03354

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Active Fourier Spectroscopy: From fixed sampling protocols to adaptive measurement

Fourier-transform spectroscopy (FTS) is based on the Nyquist-Shannon theorem, which dictates a strict sampling protocol for the measurement of continuous signals. Here we introduce Active Fourier Spectroscopy (AFS), a generalization of FTS in which measurements are still taken in the Fourier-conjugate delay domain, but where the most informative acquisition points are selected adaptively and the spectrum is reconstructed by sequential, uncertainty-aware inference rather than by fixed-grid Fourier inversion. Our approach relies on a theoretical insight that the Nyquist condition applies only under complete ignorance about the system under investigation, and shows how to systematically improve upon it when prior knowledge exists. Because AFS recovers conventional FTS whenever the prior is uninformative, its performance never falls below that of conventional sampling. We demonstrate AFS for optical spectroscopy in three settings vital to research and medicine: clinical FTIR-based blood plasma analysis, optical metrology of structured laser beams, and hyperspectral imaging with an RGB camera. In each case AFS reaches a higher target reconstruction quality from far fewer acquisitions and provides complete uncertainty quantification in real time, propagated rigorously through downstream analysis.

physics.optics↗
arXiv2026

Memory in Integrated Photonic Neural Networks: From Physical Mechanisms to Neuromorphic Architectures

The rapid scaling of artificial neural networks has exposed fundamental limitations of conventional von Neumann computing architectures. In these systems, the physical separation between memory and processing creates a bottleneck, as computational capabilities outpace the ability of memory and interconnects to supply and retrieve data. In contrast, biological neural systems inherently co-localize computation and memory through distributed, dynamical processes. Neuromorphic computing seeks to emulate this paradigm by leveraging physical substrates whose intrinsic dynamics simultaneously encode and process information. Among emerging platforms, silicon photoncis offer a compelling approach due to its high bandwidth, low-loss propagation, and inherent parallelism. This review examines the role of memory in integrated photonic neuromorphic systems, with emphasis on the physical mechanisms that provide volatile (short-term) and non-volatile (long-term) memory in silicon-on-insulator and hybrid silicon-on-insulator platforms. Drawing inspiration from digital, biological, and photonic memory architectures, we classify existing approaches based on their underlying physical principles. We cover implementations ranging from delay lines and slow-light structures to multistable dynamics and structural memory based on charge trapping and phase-change materials. We then discuss how these mechanisms support photonic neural network architectures, including feed-forward, reservoir computing, spiking and hybrid optoelectronic recurrent systems, and assess their relevance for time-dependent singal-processing tasks such as channel equalization in telecommunications. This review aims to establish a unified framework for understanding memory and learning in neuromorphic photonics and outlines key challenges and opportunities for scalable, energy-efficient neuromorphic hardware.

physics.optics↗
arXiv2026

Exoplanet Detection Using Adaptive Quantum-Optimal Measurement

Detecting terrestrial exoplanets in the habitable zones of nearby stars remains a critical challenge. Such planets can be \(10^8\) to \(10^{10}\) times fainter than their host stars and lie at diffraction-limited angular separations, where starlight strongly obscures the companion signal. Here we present an adaptive quantum measurement method for estimating the number, positions, and brightnesses of mutually incoherent point sources in the sub-Rayleigh, ultra-high-contrast regime, operating at contrasts down to \(10^{-8}\) -- five orders of magnitude beyond previous quantum imaging approaches to exoplanet detection. The method adopts a spatial-mode basis that is updated to maximize the quantum Fisher information per detected photon. Estimation is performed by maximum likelihood in log-brightness coordinates, and the source count is determined by Bayesian-information-criterion (BIC) model selection directly from photon-count statistics, without a tunable detection threshold. For point sources within sub-Rayleigh separations and with brightness ratios spanning eight orders of magnitude, the method reconstructs complete scenes with a mean success rate of \(72.5\%\). Furthermore, it is robust to misalignment, maintaining a \(71.3\%\) success rate under offsets of up to six pixels. These results demonstrate that terrestrial exoplanets can be detected below the Rayleigh limit, a regime previously inaccessible to direct imaging.

physics.optics↗