Electric lighting is responsible of about one fifth of the electricity demand of buildings in Switzerland. Therein, integrated control of sun shadings and artificial lighting can mitigate the electricity demand while maintaining the user comfort and performance. However, the drawback of the existing smart building technologies is that they do not take visual comfort effectively into account, one of the essential aspects of indoor human comfort. An ‘on-the-fly’ measurement of a visual comfort index, Daylight Glare Probability (DGP), by a novel High Dynamic Range (HDR) vision sensor was introduced into the building control system optimizing the sun shadings position and electric lighting status. Two identical office rooms (reference and advance) of the LESO solar experimental building were occupied by 30 human subjects during 15 afternoons and were used to experimentally compare the performance and users acceptance of the fuzzy logic based control system against a best practice reference system. A paper-based and a computer-based visual tests, subjective self-reported visual comfort surveys and precise monitoring of the electric lighting consumptions were carried out. The results indicates that the electricity demand of the advanced system is 32% lower than the reference one. It also shows that, while the subjects’ visual performance remain comparable in the two office rooms, the advanced control system is successful in preventing discomfort glare sensations.