Self-updating models for sentient buildings

The objective of this project is to realize and test a prototype of sentient building technologies. To achieve this goal, already acquired scientific foundations must be translated into an industrially promising level. We believe that this needs to occur in the integration of building and control models, and autonomous generation and updating of those models. We propose building systems control as an application domain for sentient buildings as this domain has a high technical, economical, social, and environmental potential. The Fundamental Layer includes preparatory and exploratory work in the areas of building model structure and generation, model management and maintenance, and model semantics/control. The Translation Layer builds on this previous work to integrate the above components of sentient buildings. This requires further advances in each area as well as a distributed software architecture. The Application Core is a prototypical implementation of the sentient building concepts, which will be tested and evaluated in a physical test environment. It will include the deployment of an integrated building control system for lighting and heating, ventilation, and air-conditioning control in an office setting. The test-bed will demonstrate: i) how a rich model of an existing facility is created based on a combination of sensor information, spatial reasoning, and generative rules, ii) how this model is autonomously updated in view of changes in context (e.g. weather and sky conditions), occupancy (people`s presence and activities), internal configuration (e.g. actual location of furniture and equipment), and building systems/devices status), iii) how building control may be improved with such a model.