Overcoming Disorder in rf SQUID Metamaterials through Self-Organization

Self-organization is an emergent phenomenon in nonlinear dynamics, where a global pattern forms through local interactions. It is not only the underlying principle of many natural processes, such as phase transitions in materials and neural networks in the brain, but also of great practical interest in creating adaptive systems that can respond to time-varying inputs and remain robust to noise [1]. In this work, we build an experimental platform to inform and test the development of a universal machine learning framework to shape the nonlinear dynamics of a system through self-organization. A new generation of laser scanning microscopy (LSM) [2, 3] that incorporates a spatial light modulator to produce patterned laser perturbation is developed. This technique has been applied to image the radio frequency superconducting quantum interference device (rf SQUID) metamaterial [4], a well-known nonlinear system [5]. We demonstrate both experimentally and numerically that a tailored laser perturbation pattern can improve the rf SQUID system’s coherence. A high-throughput LSM is also under construction to generate the large amount of data required by the machine learning algorithms being developed by our colleagues under MURI collaboration: AI-guided self-organization tailoring disorder to shape complex nonlinear dynamics.

[1] npj Complex 2, 10 (2025)

[2] Low Temp. Phys 32, 592–607 (2006)

[3] Appl. Phys. Lett. 114, 082601 (2019)

[4] Phys. Rev. X 3, 041029 (2013)

[5] Phys. Rev. E 105, 044202 (2022)