Enabling scalable, human-centric daylight assessment: A validated open-source sensing system for spatial-spectral monitoring in buildings

The intensity and spectral composition of daylight influence human health and well-being, yet a methodological gap exists between scientific understanding and practical building performance monitoring. While wearable systems have matured for personal dosimetry, building science lacks affordable, open-source instruments specifically optimized for static long-term, synchronized spatial and spectral monitoring. To bridge this gap, this paper presents a validated, open-source sensing system combining a 21-channel multispectral sensor (UV–NIR) with a High Dynamic Range (HDR) fisheye camera, capturing both the spectral power distribution and the full hemispherical luminous environment. A rigorous, multi-stage calibration pipeline ensures scientific validity by addressing cosine response, thermal drift, camera response functions, vignetting, and absolute spectral and spatial responsivity. Validation results demonstrate a spectral Mean Absolute Percentage Error (MAPE) of 9.3% and a camera luminance MAPE of 4.1%. Furthermore, a radiometric cross-validation between the spectral and spatial subsystems confirmed internal consistency with a MAPE of 9.2%. By providing a fully documented open-source package, this work removes a critical financial barrier, enabling the long-term studies needed to develop and validate the next generation of human-centric building design standards and simulation tools. The system already serves as the foundational component for an ongoing larger-scale investigation utilizing a multi-node IoT network to validate simulation engines such as Radiance, thereby showcasing the system’s immediate utility in developing new evidence-based design metrics and directly correlating quantitative spectral data with qualitative visual events to push the field towards a holistic understanding of our interaction with light.