Dynamic Lighting Optimization for Enhanced Occupant Health and wellbeing: A Digital Twin Approach Integrating Behavioral Insights

Descrizione del progetto

Un’illuminazione che contribuisca alla salute e all’efficienza

Un’illuminazione troppo intensa non comporta solo il consumo di energia, ma incide anche sulla salute e sulla produttività. Tuttavia, anche i sistemi di illuminazione avanzati spesso ignorano l’impatto sugli occupanti. I sistemi tradizionali di gestione degli edifici utilizzano curve di controllo fisse che non tengono in considerazione le risposte umane, causando disagio e inefficienza. Con il sostegno del programma di azioni Marie Skłodowska-Curie, il progetto DyLiOpt sta cambiando la situazione, mettendo le persone al centro della progettazione illuminotecnica. Avvalendosi della scienza comportamentale, lo studio approfondisce come gli atteggiamenti, il controllo percepito e le norme sociali influenzino le preferenze di illuminazione in ambito universitario. Il fulcro di DyLiOpt è un gemello digitale intelligente che apprende dal riscontro dell’utente e adatta l’illuminazione in tempo reale usando l’inferenza bayesiana. Allineando la tecnologia al comportamento umano, DyLiOpt ridefinisce il modo in cui illuminiamo i nostri edifici, per un futuro più sano e più ecologico.

Obiettivo

The impact of energy-efficient lighting systems on occupant health and well-being is often not thoroughly studied. Further, the human behavioural response is barely acknowledged while using the predefined control curves in building management systems, resulting in inefficiency, discomfort, and health problems, along with increased economic costs due to energy wastage and reduced productivity. This study will develop a digital twin paradigm extension for dynamic light optimisation based on behavioural insights and occupant needs (DyLiOpt) by mapping user responses and continuously optimising lighting during the operational phase using a Building Management System tailored to occupant needs. To analyse individuals' interaction with varying lighting conditions, three distinct settings within university spaces will be analysed by integrating mental models and the Theory of Planned Behavior to understand how attitudes, subjective norms, and perceived control shape occupant preferences for lighting. A key innovation of this study is the integration of Bayesian inference, enhancing the stochastic models with a probabilistic framework that accounts for uncertainties and variabilities. This approach will refine the models, enabling continuous updates based on new data. The Bayesian optimisation will be implemented within a digital twin platform, utilising patented autonomous optimisation technology to facilitate autonomous and semantic optimisation and dynamically adjust lighting parameters in real time, aligning with predicted occupant behaviour and lighting requirements. This research aims to set new standards in occupant-centric lighting optimisation, contributing to the health and well-being of occupants and energy efficiency in built environments while we race towards a climate-neutral continent by 2050.