Construction sites in Korea are the locus of many disasters and work-related illnesses, and construction workers are particularly likely to be exposed to serious disasters such as falls, collapses, and burial. At domestic construction sites, the concept of Design for Safety has been adopted from abroad, and current regulations are intended to secure the personnel safety at each site. However, current government guidelines and regulations are difficult to apply in the field, mainly because they do not clearly address work issues and safety management measures. The current safety review method depends too much on the subjective experience of site workers or managers. This study analyzes the step-by-step tasks required for more automated building information modeling (BIM)-based construction site safety management. An example BIM-based assessment of one specific construction site hazard, the risk of a worker fall, is carried out. In the risk analysis stage, all of the associated hazards are identified and the attendant risks are estimated and quantified. A broader risk rating methodology is derived based on the scenarios of each possible disaster at a construction site, and the hazards are extracted by defining the relationships between each building element based on the BIM information. The result is a risk rating methodology derived from a BIM-based risk assessment.

This paper describes an approach to extracting a predicate-argument structure (PAS) in building design rule sentences using natural language processing (NLP) and deep learning models. For the computer to reason about the compliance of building design, design rules represented by natural language must be converted into a computer-readable format. The rule interpretation and translation processes are challenging tasks because of the vagueness and ambiguity of natural language. Many studies have proposed approaches to address this problem, but most of these are dependent on manual tasks, which is the bottleneck to expanding the scope of design rule checking to design requirements from various documents. In this paper, we apply deep learning-based NLP techniques for translating design rule sentences into a computer-readable data structure. To apply deep learning-based NLP techniques to the rule interpretation process, we identified the semantic role elements of building design requirements and defined a PAS for design rule checking. Using a bidirectional long short-term memory model with a conditional random field layer, the computer can intelligently analyze constituents of building design rule sentences and automatically extract the logical elements. The proposed approach contributes to broadening the scope of building information modeling-enabled rule checking to any natural language-based design requirements.

Various laws and guidelines on designing for safety have been developed world-wide, and these were used to identify risks at construction sites in advance through qualitative and quantitative safe management. In other words, attention is being paid to safe management based on design drawings and models rather than field visits. It was predicted that the introduction of the concept of design-for-safety would minimize risk at construction sites through safe management in the design phase. This is because the efficiency and reliability of such a strategy has been demonstrated in various cases abroad. However, domestic design-for-safety suggests a very limited range of evaluation items, so there is a limit to using such strategies in construction sites. Building information modeling (BIM) technology is attracting attention in situations when safe management must also be based on design proposals. Although previous studies on the identification of falling disasters and automatic identification of hazards have been conducted, these studies only deal with a limited range of items for identifying risk items. For example, BIM-based risk extraction, which is limited to falling disasters, is only at the level of risk recognition that can be derived based on the process table, and evaluation is not made accordingly. It is difficult to assess the overall disaster risk in many of the items required for risk rating estimation. In this paper, we solve the above problems by deriving a BIM-based risk rating estimation scenario based on the disaster scenario for automating BIM-based risk rating estimation, and we developed an evaluation system using this method. The BIM-based risk rating estimation methodology was presented through BIM-based hazard extraction, evaluation of requirement information, evaluation item selection, and using the evaluation system in a design-for-safety review.

Building information modeling (BIM) technology has been utilized increasingly in quantitative ways in the architecture, engineering, and construction disciplines. However, owing to increasing requirements for improvement in qualitative factors in BIM-based design projects, it has become necessary to develop a checking and evaluation process for BIM-based architectural design. The purpose of this study is to develop and apply BIM-based quality control requirements for improving the quality of architectural design. To achieve this, the research investigated case studies for BIM data quality control and classified quality control targets according to physical/logical quality and data quality. The quality check criteria and checklists are developed through the reconfiguration of deduced quality control targets by requirement. The research developed a rule-based quality checking system using requirements for efficient quality control based on BIM.