This paper presents a field-based method for assessing personal light exposure using wearable spectral light sensors under real-world conditions. The method calculates melanopic irradiance and melanopic equivalent daylight illuminance from continuous spectral measurements and evaluates these measures in relation to self-reported sleep, alertness and mood data collected during daily life. The approach is demonstrated using one-week monitoring data collected from 50 university students and is evaluated against commonly used circadian lighting metrics, including WELL, Brown et al. recommendations and non-visual direct response. Results showed that participants frequently did not meet the recommended thresholds, while continuous melanopic exposure was associated with sleep and alertness. These results highlight the complementary value of continuous personal monitoring and the need for controlled studies to test the effects of meeting the recommendations.
Many people in industrialised countries would prefer to remain in their own homes for as long as possible. Remaining healthy while ageing in place will depend on maintaining suitable indoor environmental quality. This paper reports preliminary results concerning lighting conditions in the homes of older adults in Canada based on an online survey. The results suggest that older adults who spend close to 24 h daily in their homes might not receive the recommended daily light exposure of a minimum of 250 lx melanopic equivalent daylight illuminance from morning until 3 h before bed. Those who spend time outdoors might receive an adequate total exposure if shorter or intermittent very high illuminance exposures are equivalent to long periods at a lower level. Survey participants were generally satisfied with their lighting. Regression analyses revealed small but statistically significant effects of light exposures and satisfaction with lighting on the ease of falling asleep, sleep quality and quality of life. Validation of these survey results with light logger measurements and field measurements of home lighting may lead to recommendations for changes to building codes to enable those who spend much of their time indoors to achieve the target daily light exposures.
Lighting conditions can alter thermal perception, but the magnitude of these shifts on a temperature-equivalent scale is uncertain when physiological state is measured concurrently. This secondary analysis re-examined climate-chamber data from 54 young Japanese women exposed at 26°C, 28°C and 30°C to 30 combinations of illuminance (100 lx to 2900 lx) and correlated colour temperature (CCT; 2800 K to 5200 K). Conduction-corrected modified effective temperature (ETF) was used to express lighting-related differences in thermal appraisal. Mean skin temperature showed no consistent lighting-related shift. At neutral visual thermal sensation, ETF-equivalent values for six illuminance levels spanned approximately 2°C, with a maximum difference of 1.7°C between 250 lx and 1700 lx; however, the values were non-monotonic and bootstrap confidence intervals were wider than the ETF range studied. CCT showed no statistically supported association with visual thermal comfort. The ETF-equivalent values therefore describe perceptual appraisal rather than a change in physical temperature or thermophysiological heat balance. Under these short, near-neutral exposures, illuminance may influence thermal appraisal, but the present data are insufficient to support specific lighting or HVAC control recommendations.
The review provides a focused analysis of machine learning (ML) and deep learning (DL) models for managing correlated colour temperature (CCT) in adaptive road lighting systems. As urban infrastructure evolves towards smart, responsive technologies, the ability to dynamically control lighting parameters is critical for enhancing safety and energy efficiency. We systematically evaluate the motivation for moving beyond traditional control logic to data-driven ML methodologies. We conduct a detailed survey of applicable ML paradigms, including supervised, unsupervised and reinforcement learning. Specific techniques such as regression models, support vector machines and fuzzy logic systems are critically assessed for their utility in CCT selection. Furthermore, we explore the advanced capabilities of DL architectures, examining how convolutional neural networks can leverage visual data for rich contextual awareness and how recurrent neural networks can process time-series data for predictive control. A comprehensive comparative analysis evaluates these techniques against key criteria, including accuracy, computational complexity, data requirements and real-time applicability. This review synthesizes the key advantages and trade-offs of each approach, offering guidance for researchers and practitioners in selecting and implementing the most suitable computational strategies for developing next generation, intelligent road lighting systems.
