Creation of a database of building elements which includes a description of key elements, suitable for further modelling with artificial intelligence algorithms.

• Objectives: Participants will learn to design a basic database of structural and other elements of buildings. With the help of data mining they will obtain key elements that are the carriers of the largest data. Based on the obtained results, they will be able to plan future investments costs for renovations.
• Content:
  – Theory based introduction to the fields of artificial intelligence, data mining and other necessary data preparation procedures
  – Getting to know the content of the problem and active study of the possibilities of performing data mining
  – Preparation of information for the process of data mining
  – Data Mining
  – Analysis of results and discussion
• Duration : 5 days
• Who is this training designed for: Business leaders, planners and developers, and specialists in the following companies:- Companies engaged in facility maintenance- Companies that manage their own real estate and maintain it in terms of investment

  – Companies that manage other types of databases and would like to create interdisciplinary connections and desire to aquire new knowledge

• Organizer: Faculty of Civil Engineering, Traffic Engineering and Architecture, UM
• Lecturer: dr. Daniela Dvornik Perhavec, B.Eng.
• Note: Prior-knowledge of MS Excel is required

• Energetika Maribor d.o.o.
• Eutrip d.o.o.

Project E-NIZ

In the E-NIZ project we performed data mining using a combination of data on structural elements and energy consumption for heating. We combined static (building data) and dynamic (heat consumption data) records available in the databases. The database of construction elements was created using archived plans available in the premises of Technical Faculties. In 2013, when the project began no comparable database existed. Data on the thermal energy consumption of buildings in the area of ​​the Technical Faculties for the period between 2012 and August 2015 were obtained from energy suppliers.

The consumption of thermal energy for 2014 was estimated according to the share of the total area of ​​an individual building. This procedure was used due to the lack of information on energy consumption for individual units. Depending on the consumption of thermal energy per m² of surface, energy classes were determined in accordance with the Rules on the methodology of production and issuance of energy performance certificates for buildings (Official Gazette of the Republic of Slovenia 92/2014).

The existence of an extensive database is a prerequisite for a more efficient identification of correlations between various elements with the help of artificial intelligence. Considering that the we created the first database of characteristic features of buildings the outcomes of the project were on the one hand as expected, but on the other hand surprising und unanticipated.

The most important information was the type of facade, followed by the volume of the building and the consumption of thermal energy. With the help of artificial intelligence, we gained new insight into the topic at hand. A summary of the project is presented in the article ‘Support of artificial intelligence in projects of renovation and management of buildings’ by dr. Daniele Dvornik Perhavec, Ph.D. Danijel Rebolj and dr. Milan Zorman (published in Gradbeni Vestnik).

Other current projects

• E-NIZ – E-immovable engineering database of the Faculty of Engineering

• Dvornik Perhavec D., Rebolj D. and Zorman M. (2017): Podpora umetne inteligence pri projektih prenove in upravljanja stavb.
• Dvornik Perhavec D., Rebolj D. and Šuman, N. (2015): Systematic approach for sustainable conservation.

With the knowledge of the history of construction and associated regulations we can digitize the structural elements of buildings. After obtaining the required information we utilize data mining and artificial intelligence algorithms to detect patterns of elements in different buildings. This can help us save time and the resources needed to explore an existing building, especially in the case of historic buildings where the construction plans have been long lost. The digitization of an existing building can be used as the basis for further modelling; particularly with regards to seismic engineering, energy efficiency, renovations of buildings and installation of new structural loads (e.g., renovation of attics).