Camera-based window-opening estimation in a naturally ventilated office

Naturally ventilated offices enable users to control their environment through the opening of windows. Whilst this level of control is welcomed by users, it creates risk in terms of energy performance, especially during the heating season. In older office buildings, facilities managers usually obtain energy information at the building level. They are often unaware or unable to respond to non-ideal facade interaction by users often as a result of poor environmental control provision. In the summer months, this may mean poor use of free cooling opportunities, whereas in the winter space heating may be wasteful. This paper describes a low-cost, camera-based system to diagnose automatically the status of each window (open or closed) in a facade. The system is shown to achieve a window status prediction accuracy level of 90-97% across both winter and summer test periods in a case study building. A number of limitations are discussed including winter daylight hours, the impact of rain, and the use of fixed camera locations and how these may be addressed. Options to use this window-opening information to engage with office users are explored.