An approach to measuring chemical concentrations using a slotted micro-ring resonator (sMRR) is proposed which is robust to spectral shifts caused by temperature variations. Two 1-D Convolutional Neural Network architectures, ResNet34 and VGG20, were trained for regression, achieving mean squared errors (MSEs) of 1.1251 and 1.2195 , respectively. The models predict concentrations of water, ethanol, methanol, and propanol (0–100%) from the transmission spectra of a single-ring sMRR etched in heavily doped silicon, operating in the mid-infrared range (290–310 K). Transfer learning adapted the models for datasets with different temperature ranges, analytes (e.g., butanol), and sMRR designs, achieving comparable accuracy. Variations in accuracy across these datasets are also explored.
Unidirectional invisibility (UDI) is a fascinating phenomenon that emerges in non-Hermitian systems. Traditionally, achieving UDI in optical platforms requires carefully engineered multilayer structures incorporating regions of gain and loss (or unequal loss), along with finely tuned parameters. As a result, in such systems, optimal UDI typically occurs only at isolated points or within narrow regions of the parameter space, making it difficult to optimize the effect for additional requirements such as broad bandwidth or nonlinear response. In this study, we demonstrate—both theoretically and experimentally—that unidirectional zero reflection (UDR), a precursor to UDI, can be achieved using only standard optical components such as beam splitters and mirrors. Crucially, the proposed device can be readily tuned into the UDI regime using conventional optical amplifiers and phase shifters. Our approach enables robust, on-chip realization of UDR photonic systems with high tolerance to fabrication imperfections. To underscore the versatility of our design, we validate its experimental feasibility using standard silicon photonics technology. This work paves the way for leveraging the standard photonic design toolbox to engineer complex UDR systems with tailored functionalities, including adjustable bandwidth and nonlinear responses—features with strong potential for applications in laser engineering and optical signal processing. Furthermore, the reliance on standard components makes this framework naturally extendable to the quantum regime. Beyond optics, our approach may also inspire analogous implementations in other wave-based systems, such as acoustics and cold-atom platforms.