- 期刊
應用BIM和GPT技術於建築工程風險管理可視化之研究－以屏東○○宿舍工程為例
APPLICATION OF BIM AND GPT TECHNOLOGIES IN VISUALIZED RISK MANAGEMENT OF CONSTRUCTION ENGINEERING - A CASE STUDY OF THE PINGTUNG XX DORMITORY PROJECT
摘要
在推動國家經濟發展的同時，營建業因專案高度複雜的施工環境與條件而面臨高度風險。現有的風險管理使用傳統方法進行施工風險因素的識別和風險等級的排序，過度依賴經驗和定性分析，缺乏針對營建專案的動態和系統化風險評估工具，導致風險識別和決策過程效率不佳，無法有效應對瞬息萬變的施工環境。此外，這些方法無法提供直觀的理解給專案利害關係人，難以達成有效的溝通與決策。針對這些挑戰，本研究提出了一種整合建築資訊模型（BIM）和生成式預訓練轉換器（GPT）技術的風險管理模式。利用BIM的3D視覺化能力和GPT的深度學習算法，我們實現了一個動態且互動的風險評估工具，能夠在Power BI上即時更新風險狀態並提供基於數據的風險緩解策略建議。這一整合方案不僅提高了風險評估的準確性和時效性，其可視化展示方式增強了風險溝通的直觀性，使非技術背景的利害關係人也能有效參與決策過程。初步評估結果顯示，使用該系統在案例專案中，可以提高風險識別的準確性並縮短決策時間，有效提升專案的風險管理效率。研究成果顯示，結合BIM和GPT技術可以顯著提升營建產業的風險管理能力，值得進一步研究和投入。
並列摘要
While promoting national economic development, the construction industry faces high risks due to the highly complex construction environments and conditions of projects. Current risk management methods rely on traditional approaches for identifying and ranking the severity of construction risk factors, which are heavily dependent on experience and qualitative analysis. They lack dynamic and systematic risk assessment tools tailored to construction projects, resulting in inefficient risk identification and decision-making processes that cannot effectively respond to the rapidly changing construction environment. Additionally, these methods must provide more intuitive understanding to project stakeholders, making effective communication and decision-making difficult. To address these challenges, this study proposes an integrated risk management model that combines Building Information Modeling (BIM) and Generative Pre-trained Transformer (GPT) technology. Utilizing BIM's 3D visualization capabilities and GPT's deep learning algorithms, we developed a dynamic and interactive risk assessment tool that can update risk status in real-time on Power BI and offer data-driven risk mitigation strategy suggestions. This integrated solution not only improves the accuracy and timeliness of risk assessments but also enhances the intuitiveness of risk communication through its visual presentation, enabling stakeholders without technical backgrounds to participate effectively in the decision-making process. Preliminary evaluation results indicate that using this system in case projects can improve the accuracy of risk identification and shorten decision-making time, significantly enhancing project risk management efficiency. The findings suggest that the combination of BIM and GPT technology can substantially improve risk management capabilities in the construction industry, warranting further research and investment.