Project Details
Projekt
Data-driven evaluation of design variants
Applicant
Professor Dr.-Ing. Frank Petzold
Subject Area
Architecture, Building and Construction History, Construction Research, Sustainable Building Technology
Term
from 2016 to 2021
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 271444440
The architectural design process is a decision-making process within which a choice is made between at least two variants (possible actions). When evaluating variants, the ACTUAL properties of individual solution-variants are compared with corresponding TARGET properties (objectives of the client, architects, legal framework, etc.). As a result, the quality of a solution can be determined as a degree of target-attainment and a decision regarding subsequent further development or rejection can be made. The assessment is carried out both qualitatively (verbally descriptive) and quantitatively (alpha-numerical data). This process is currently lacking comprehensive documentation. An initial research aspect of the present proposal focuses on the development of a methodology for the digital documentation of qualitative (descriptive) decisions. Documenting of quantitative decisions was the subject of the first funding period. The aim is to record and document variant selection decisions, including the reasoning behind decisions made, in order to ensure future traceability and to improve knowledge transferability to other projects. For the textual documentation of the design process and decision-making, established procedures such as the use of “design stories” will be investigated, technical vocabulary with regards to spatio-functional, constructive, and energy-related aspects derived, and interrelationships between building components and their parameters within semantic building models identified. In addition to a keyword-based assessment, new conceptual methods for visual qualitative assessments will be developed, prototypically implemented and evaluated in user-studies. A second facet of the project is the evaluation of variants based on references and the use of (partial) aspects for the further design development. The use of references from already built or designed buildings is an established method to evaluate variants as well as to show the potential in further variant development. To provide the overall similarity, interactive user-controlled weighting methods are developed. For the evaluation of the potential of variants in further development, including solutions in design branches of similar references, existing calculation and visualization methods are extended by imperfection (uncertainty, vagueness, inaccuracy) and evaluated exemplarily in the context of energy efficiency and environmental impacts. Prototypical implementations and a partial evaluation with the other sub-projects demonstrate the viability of the concept.
DFG Programme
Research Units
