[Back]


Publications in Scientific Journals:

C. Sustr, U. Pont, A. Mahdavi:
"Rule Based Building Construction Generation: An Approach Based on Formal Language Methods";
Applied Mechanics and Materials, 861 (2017), 564 - 573.



English abstract:
The composition of efficient and appropriate building constructions is a key agenda in the
building delivery process. While this process is regularly considered to be of highest importance for
the final quality of a building, many involved stakeholders regard it as a cumbersome and repetitive
routine. Therefore, approaches to facilitate this process should be investigated. Toward this end, we
address the layer-wise building component composition via formal language methods. These are
regularly used in computer science to formalize real-world processes into a language that can be
processed by a computer. Regarding building component generation, relationships and
interdependencies between different layers need to be considered. While these are easy to
understand for a skilled human planner with pertinent domain knowledge, the exact formulation of
building composition rules is far from trivial. Thus, automated building part generation requires
collection and formalization of the required knowledge regarding building component composition,
so that it can be readily transformed into a processual form. After collection, definition and
structuring of such rules, the overall process of component generation can be expressed in Pseudo-
Code. This offers three major advantages: i. Pseudo-code is vendor and platform neutral and is a
widely used concept in computer science; ii. Potential mistakes and issues can be easily identified,
iii. Flexibility, extensibility and editing ease is ensured. In this contribution we illustrate a general
approach, define certain rules and thresholds, and introduce a formalized method for building part
generation. Furthermore, we demonstrate the concept via a limited number of constructions and
discuss potential application scenarios.

German abstract:
(no english version) The composition of efficient and appropriate building constructions is a key agenda in the
building delivery process. While this process is regularly considered to be of highest importance for
the final quality of a building, many involved stakeholders regard it as a cumbersome and repetitive
routine. Therefore, approaches to facilitate this process should be investigated. Toward this end, we
address the layer-wise building component composition via formal language methods. These are
regularly used in computer science to formalize real-world processes into a language that can be
processed by a computer. Regarding building component generation, relationships and
interdependencies between different layers need to be considered. While these are easy to
understand for a skilled human planner with pertinent domain knowledge, the exact formulation of
building composition rules is far from trivial. Thus, automated building part generation requires
collection and formalization of the required knowledge regarding building component composition,
so that it can be readily transformed into a processual form. After collection, definition and
structuring of such rules, the overall process of component generation can be expressed in Pseudo-
Code. This offers three major advantages: i. Pseudo-code is vendor and platform neutral and is a
widely used concept in computer science; ii. Potential mistakes and issues can be easily identified,
iii. Flexibility, extensibility and editing ease is ensured. In this contribution we illustrate a general
approach, define certain rules and thresholds, and introduce a formalized method for building part
generation. Furthermore, we demonstrate the concept via a limited number of constructions and
discuss potential application scenarios.

Keywords:
Formal Modelling Approaches, Building Constructions, Layered Compositions, Thermal Building Performance, Automation in Construction


"Official" electronic version of the publication (accessed through its Digital Object Identifier - DOI)
http://dx.doi.org/10.4028/www.scientific.net/AMM.861.564


Created from the Publication Database of the Vienna University of Technology.