Dynamic optical monitoring with an ESP32-CAM-based low-cost spectrophotometric prototype

Real-time optical monitoring of large area dynamic color systems remains a technical challenge, particularly when aiming for low-cost, portable, and scalable solutions. In this study, we present a spectrophotometric platform based on an Arduino-compatible ESP32-CAM, a modular 3D-printed prototype, and a custom Python interface capable of reconstructing visible-range transmittance spectra (400–700 nm) from RGB images based on a neural network protocol. The system enables real-time visualization of both full spectral evolution and single-wavelength transmittance during any process of color switching, offering valuable insight into optical kinetics without requiring expensive instrumentation. To validate its performance, thin films of five electrochromic polymers covering visible gamut color transitions (orange, magenta, cyan, green, and black) were analyzed under color-changing dynamic conditions. The transmittance evolution and reconstructed spectra obtained using the Arduino system were compared against those acquired from a reference spectrophotometer and a previously validated digital camera-based method. Results showed excellent agreement across most polymers, with deviations in transmittance values remaining below 4% even in the most challenging cases. The developed interface allows users to select specific areas of interest and control time resolution, while all data can be exported for further analysis. Additionally, all captured images are stored for subsequent high-accuracy post-processing using color correction and spectral reconstruction models. This work introduces a robust, accessible tool for dynamic optical analysis that maintains competitive performance compared to conventional systems. Its versatility, portability, and real-time functionality make it particularly suited for the in-situ optical evaluation of surfaces, smart materials and devices in both laboratory and field environments.