Optimizing energy efficiency, sustainability, and cost of buildings by advanced analytics

In the construction industry the sustainability of buildings in general and the energy consumption in particular are very important issues today.  Here, optimization is a challenging task for digital planning and product selection, for example while designing ventilation, heating, air conditioning and selecting windows, doors, radiators, pipes, funnels, etc.

This research project aims at the development of an automated recommendation system mainly based on product requirements that may come from a digital twin of the building (building information modeling, BIM). Specifications of products on the market are extracted from documents on the internet. In the first phase of the project automated data collection mechanisms, including data scraping, based on adequate interfaces are required. In the second phase the large amount of documents is pre-processed to extract useful information and add semantics and relationships. This is done by text mining and natural language processing (NLP).

The extracted, classified, and interpreted data is used to develop a recommender system to support the decision making of building designers. Improving energy efficiency and sustainability based on combinations of product specifications and restrictions can be seen as a linear optimization problem. Decision trees provide a possible solution approach, looking at decision making as a tree-like procedure including potential consequences like resource costs and environmental benefits.