Building Information Modelling, or BIM, was first conceptualized in the 1970s. Since then, BIM has become a widely used approach in the AEC industry. Research is being done to streamline the process even further by identifying the influential design parameters at each stage – giving architects a clear design path.
Multi-Level-of-Development Building Information Modeling
Building Information Modelling is a digital and automized version of the design process. The design is moved through a series of levels to refine the information in the design, known as the Level of Development, of LOD. Since design involves collaborating with many difference disciplines, BIM enables sharing of 3D-models between those disciplines.
In any design, the process likely will not be a linear process and designers will switch between different levels of design as more information is collected – making the final design uncertain. Multi-LOD is based on this concept of uncertainty and at each level of design the architect focuses on select parameters deemed vital to that stage of design development. When those parameters are defined, the uncertainty of the final model is reduced.
BIM and Multi-LOD fail to define what information is needed at each level of design. Creating definitions for the various levels would streamline the design process by enabling the architect to give priority to the most important design parameters at each level of design.
Sensitivity Analysis to Define Design Levels
Researchers at the Berlin University of Technology and KU Leuven used a sensitivity analysis to define the design levels of building energy models. Energy efficiency modelling was used as it is one of the most important aspects of architectural design. The study, published in Advanced Engineering Informatics released in January of 2020, defined the study goal as: “to assist the designers by identifying the most important design parameters (information), thus suggesting them to focus on some selected parameters at each level of multi-LOD.”
The model analyzed a test building model, with a rectangular shape, and five additional building shapes commonly used in office building design. All six models were assumed to be in Regensburg near Munich, Germany. Researchers identified detailed design parameters related to energy efficiency, categorized under five groups: Geometrical, Technical Specification, Window Construction, Operational Design, and System Efficiency.
Researchers selected a variance-based sensitivity analysis, versus the Morris method or a regression-based method, as it is model-independent – making it more helpful with complex models. The sensitivity indices, S and ST, were calculated using fully automated methods. From there, researchers also analyzed the uncertainty of each design parameter group to identify how each group would increase the likelihood of potential design changes at a later stage in the design.
The study found that the energy performance of a building is heavily influenced by its shape, so the ideal shape should be identified at the very beginning of the design stage. There was no difference in parameter sensitivity between the different building shapes, indicating that the following procedure can be applied to any design at an early stage. Researchers used the uncertainty to propose the following order of design preference: (1) geometrical parameters, (2) technical specifications and operational design, and (3) window construction and system efficiency. Each design parameter should be defined in the group before moving onto the next group.
Researchers recognized that these results are specific to the test project location near Munich, Germany, as weather plays a vital role in energy predictions and suggested further studies to generalize the results for another location.