Researchers at Leibniz IPHT Have Developed the First Reprogrammable Fiber, Marking a Significant Advancement in Adaptive Photonic Systems

Light typically propagates through optical fibers according to fixed laws. Once manufactured, the properties of these fibers are virtually unchangeable. However, researchers at Leibniz-IPHT have demonstrated that these limitations can be overcome. They developed a fiber whose behavior can be controlled, reconfigured, and optimized specifically in real time. This creates a platform for the first time that allows for the programmable control of nonlinear optical processes.

The team, led by Prof. Markus Schmidt (head of the Fiber Photonics Department), combined a liquid-filled hollow-core fiber with a self-learning control algorithm. Inside the fiber is a liquid whose refractive index can be finely adjusted via temperature. Twenty-four microscopic heating elements along the fiber control the local temperature — and thus how light propagates within the fiber.

An algorithm monitors the resulting light spectrum and continuously adjusts the heating profiles until the output light has the desired properties. This creates a closed-loop control system between light, matter, and the computational process; the system optimizes itself as it operates.

“We wanted to demonstrate that nonlinear optical processes can be actively shaped through both fiber design and operation,” explains Markus Schmidt. “This gives us a tool that can algorithmically control light dynamics. It could form the basis for future adaptive, learning-capable photonic systems.”

Prof. Markus Schmidt developed the concept together with first author Johannes Hofmann, Dr. Ramona Scheibinger, Dr. Bennet Fischer, and Prof. Mario Chemnitz from Leibniz- IPHT. The researchers employed a method from artificial intelligence known as particle swarm optimization. Like a digital swarm, the algorithm simultaneously and independently evaluates countless temperature configurations and searches for the best solution.

The result is a precisely controlled white light spectrum that can be specifically shaped and maintained stably. Depending on the target specification, the fiber can amplify or suppress certain wavelengths, which is a way to adjust the system in real time that was previously difficult to achieve. “The fiber reacts dynamically to its own light states,” says Hofmann.

“We have created a system that actively adjusts its dispersion, and thus its behavior, to find the desired spectrum on its own.”

A New Generation of Photonic Systems

The researchers demonstrate that nonlinear interactions in light can now be controlled in a reversible and spatially resolved manner for the first time. This creates a platform for applications that were previously limited to rigid systems, such as adaptive frequency conversion, tunable soliton dynamics, spectral filtering, and nonlinear optical information processing.

“This work marks a true paradigm shift,” emphasizes Schmidt. “We combine fiber technology, nonlinear photonics, and algorithmic optimization into a new class of reprogrammable photonics. It opens up the possibility of dynamically and purposefully shaping nonlinear processes – this is a decisive step toward programmable light control in complex photonic systems.” Applications In the future, programmable liquid- core fibers could be used as adaptive light sources for supercontinuum generation, nonlinear spectroscopy, quantum and communication optics, and optical computer architectures. With algorithmic control, processes can automatically adapt to changing conditions, forming the basis for intelligent, self-regulating photonic systems.

 

Original Publication: https://doi.org/10.1038/s41467-025-63213-8