Knowledge-based building management combining human perception and building automation systems

A primary goal of building management is the efficient utilization of energy and resources in order to maximize comfort for building users. According to the current state of the art, various technical systems (installation, measurement, and automation engineering) are described that aim at providing high user satisfaction. Quantification of comfort is usually based on objective measurement and empirical models.

Although these approaches are often very complex, problems arise as the users and their continuous feedback are not directly involved in the overall process. Influences on their satisfaction are not or not adequately considered (e.g. activity, state of health, weather perception, or influences of other persons). While user feedback is an important source for energy-efficient and comfortable building operation as well as for the development of novel services for users based on their needs and actual problems, manual processing of user feedback is a high effort for building management. In order to include user feedback in an appropriate way and bring it in line with sensor measurements, an automatic integration is necessary, first. For this purpose, the design of an intuitive interface for simple and fast collection of human perception (Hum) regarding various comfort parameters is developed in the scope of the project.

Furthermore, the integration of these feedback records into existing integration solutions of building automation systems (BAS) is a central challenge that is subject to research in this project. In this context, compliance with requirements regarding privacy of user data is essential. By the combination and semantic modeling of objective sensor measurements and subjective user feedback into a common knowledge base, results of the project HumBAS will allow deriving statements on occurring problems in building management processes, such as significant deviations between measurements and user feedback. Expert knowledge and additional context information with respect to the building, the installed automation systems, the building operation strategies, and the subjectively perceived influences are integrated. Thus, a more revealing interpretation of common sensor measurements is enabled. The used methodology is based on approved approaches of knowledge engineering (e.g. ontologies and reasoning).

In summary, high need for research based on current projects and state of the art is identified in order to establish a user-centric, intelligent building management. The influence of user feedback in buildings is researched in this project. However, project findings are also portable to other application areas, where users play a central role. In particular, in passenger transportation (trains, airplanes) as well as in assisted living apart from conventional residential and commercial buildings, the developed approach will show considerable potential.

Big picture

The project "Knowledge-based building management combining human perception and building automation systems" (HumBAS) is supported by FFG (ICT of the Future, Project Number 867681)