Today’s building control systems are technology oriented rather than user-centric. Therefore, control actions are sometimes in conflict with the occupant’s needs for comfort and convenience. As a result, occupants often turn off automatic systems or develop strategies to circumvent them. User-centric building control systems are therefore desirable, but they require a thorough understanding of the occupant’s behaviour. Such knowledge enables control actions that ensure user comfort whilst saving energy.
The primary objective of the «SCCER – FEEBD» project is to explore methods for the recognition of relevant user activities in order to enable user-centric control strategies. To reach this goal, typical scenarios are first developed where the building control system could have a significant impact on energy savings. Then, it is analysed which user activities or activity patterns (behaviours) are associated with those scenarios. Thus, insight is gathered about the interaction between the building control system and its occupants.
In the next step, state-of-the-art sensor solutions are compared to each other in a laboratory setting. This allows one to determine the most suitable solution (single- or multisensory) for the reliable recognition of predetermined activities and behaviours. Finally, concepts based on machine learning and sensor fusion algorithms are developed in order to reliably recognize the selected activities. The development process will initially take place in a laboratory setting. The most promising solution(s) will then be validated in a real-life environment.
The project also investigates how data-driven building energy models can be automatically created and consecutively used for building characterization («fingerprint») and energy usage prediction. Data from various buildings of different types, both residential and functional, will thereby be used as input for the modelling task.