Predicting the added value of smart systems in a building

Predicting the added value of smart systems in a building: A decision model approach

Dijkstra, Niels (TU Delft Architecture and the Built Environment)

Arkesteijn, M.H. (mentor)
de Graaf, R.P. (graduation committee)
Leenders, L. (graduation committee)
Binnekamp, R. (graduation committee)

Delft University of Technology

Architecture, Urbanism and Building Sciences | Management in the Built Environment

2018-11-20

Purpose – One of the long-standing issues in the field of real estate management is the alignment of demand and supply. When building retrofit is initiated to comply with changing demands, the decision-maker has to deal with several, often competing, criteria. Smart systems have the potential to contribute to the process of optimally attuning the building with the changing demands, in other words, adding value to the building. To date, plenty of research exist on predicting the added value based on quantitative variables but fail to incorporate quantitative variables. The purpose of this paper is to provide a decision model that is capable of predicting the added value of smart systems in a building. It offers real estate managers a Preference Function Modelling (PFM) process, structure and model to approach a multi-criteria problem.

Design/methodology/approach – The PFM approach, an evaluation operational research methodology, is designed to help decision makers to choose the most preferred alternative from a set of already existing alternatives. A decision model is made in Excel that enables the construction of measurement scales to which linear algebra and calculus are applicable. A second research technique being used is the Lagrange curve. The model is tested on a single building in the portfolio of Schiphol Real Estate. Findings – The PFM model enables the decision-maker to establish a list of overall preferences scores of all relevant alternatives. The alternative with the highest overall preference score is predicted to add the most value to the building. The pilot study reveals that the procedure resulted in a result but a certain level of uncertainty exists. The cause of the uncertainty seems to be a lack of knowledge about the alternatives. Research limitations/implications – The Lagrange curve is a suitable technique for changing values of a variable. However, the curve tends to overshoot. Practical implications – The practical applicability depends on the selection of alternatives and knowledge available. If the alternatives are not known a priori, the PFM approach cannot be used and Preference Based Design (PBD) is more appropriate. If the knowledge available is low and difficult to improve the input is subject to uncertainty. The output may not project the actual situation. Originality/value – The research is based on the PAS approach from Arkesteijn et al. (2017). The PAS approach is a PBD methodology and is adapted to a PFM methodology in order to fit the intended purpose of the research. The PFM methodology takes both scaling and the Lagrange curve into account and along with an iterative procedure the decision-making process is well supported. The model provides a custom-made outcome, which is of high value for the decision-maker in the SHG case.

Smart building
Preference Function Modeling
Preference Measurement
IoT

http://resolver.tudelft.nl/uuid:4a6a1f94-b7b2-4365-b51e-2fe3c0ad7429

Student theses

master thesis

© 2018 Niels Dijkstra