A semantic web approach to enable the holistic environmental and energy management of buildings
MetadataShow full item record
This item's downloads: 3737 (view details)
Narrowing the performance deficit between design intent and real-time performance is a complex and involved task, impacting on all building stakeholders. Buildings are designed, built and operated with the use of increasingly complex technology and throughout their building life-cycle, produce vast quantities of data. However, many commercial buildings do not perform as originally intended. The use of data in a cross-domain manner and the concept of enterprise level data are areas in which the building industry lags considerably behind others. Traditional methods of information capture in the architecture, engineering and construction (AEC) domain do not lend themselves to the effective provision of a holistic environmental and energy management solution for building performance assessment, prediction and informed modification. Existing methods of performance assessment fail to take into account the wealth of information available throughout a building and exclude whole categories of information including social media, occupant communication, mobile communication devices, occupancy patterns, human resource information and financial data. Building codes are prescriptive and do not encourage a continuous performance assessment mindset amongst building owners and users. The performance gap is dominated by the twin concerns of interoperability and a lack of holistic environmental and energy performance information. This thesis provides a dual strand approach to the problem, describing how heterogeneous building data sources can be transformed into semantically enriched information. These data can serve as a data service for a structured performance analysis approach, at the enterprise level. In parallel, a semantically enriched performance management framework is introduced, which builds on the homogeneous data described in the first strand. The Performance Framework is an approach to performance management using scenarios to describe performance and methods for capturing this performance in a series of defined performance objectives. These techniques, when applied together, result in a more holistic interpretation of building performance. A series of demonstrations are provided which illustrate the use of cross-domain data, first in an unstructured manner, and finally, using the scenario modelling approach, where a structured path is described through cross-domain data. Although the use of cross-domain data is beneficial for many building stakeholders, the building management context is considered throughout, in the form of the building manager. This thesis demonstrates how the semantic web approach can be combined with the environmental and energy management of buildings. The work describes how multi-disciplinary data sets can be described in a homogeneous manner and leveraged to drive a performance assessment approach. The work improves on other performance management techniques as it provides a cross life-cycle, cross-domain approach to the problem, enabling a true holistic assessment of building performance.