Feasibility of BIM-enabled Generative AI Platform for Data-driven Incremental Productivity Enhancement of Construction Cost Management (BIM-GAI)

生成式人工智能（GAI）的应用可以通过自动估计成本计划，测量数量以及基于设计，规格，现场信息和其他非图形输入变量预测估值和现金流结果来提高建筑中小企业成本管理的生产力。目前，英国BIM框架和2011年任务在支持中小企业使用BIM提高生产力方面效果不佳，只有57%的采用率（UKBIMA，2021）。因此，证明了英国建筑中小企业生产力的不足。BIM更新在英国建筑中小企业中受到限制，主要是因为成本和易用性。最近的人工智能和机器学习发展为BIM在英国建筑中小企业中的应用提供了新的机会。GAI算法可以通过基于深度学习的生成模型（例如BIM中基于规则的生成模型），通过自动化，潜在地减少错误和返工，这与施工估算和预算过程中的成本超支同义。使用启用BIM的完整版GAI，可以显著减少制作工程量清单、成本计划、估价、最终账目和相关文本文档所需的时间。英国的建筑成本管理过程因估计和预算错误而受到批评。基础设施项目的大规模成本超支扼杀了英国中小企业建筑公司的增长。成本管理错误导致返工和索赔，阻碍了组织的生产产出。此外，在英国，69%的建筑工程都存在成本超支，这给建筑中小企业的生产力带来了困难（毕马威，2023年）。因此，GAI尚未在建筑管理过程中实施，以优化生产力并减少错误和超支事件。因此，工程造价管理需要在规划和施工阶段向提高生产率的新模式迈进。GAI在英国中小企业建筑成本管理过程中的应用将通过BIM提高生产力。该可行性研究将研究BIM和GAI的试点整合，数据集，道德考虑因素，法规，界面，GAI要求用户界面（UI）和用户体验（UX）。此外，GAI尚未应用于BlueBeam、CostX和PlainSwift等基于BIM成本管理的工具中。** 本可行性研究旨在使用来自中小企业建筑公司的现有建筑成本数据集来开发系统架构、Foundation大型语言学习模型（FLM）以及BIM和GAI的基于网络的试点集成，以自动化成本估算，从而提高生产力。

The application of Generative Artificial Intelligence (GAI) can enhance productivity in cost management in construction SMEs by automating estimating cost plans, measured quantities, and predicting valuation and cashflows outcomes based on the designs, specifications, site information, and other non-graphical input variables. Currently, the UK BIM framework and 2011 Mandate have been ineffective in supporting SMEs' productivity using BIM with only 57% uptake (UKBIMA, 2021). Thus, evidencing a shortfall in the productivity of UK construction SMEs. BIM update has been limited in UK construction SMEs mainly because of cost and ease of use. Recent AI and machine learning developments present new opportunities to make BIM easier to apply in UK construction SMEs. GAI algorithms can potentially reduce errors and rework synonymous with a cost overrun in construction estimating and budgeting processes through deep learning-based generative models such as rule-based generative models within BIM through automation. The time required in producing bills of quantities, cost plans, valuation, final accounts and associated textual documents can be reduced significantly with a full version of BIM-enabled GAI. Construction cost management processes in the UK have been critiqued for errors in estimation and budgeting. Large-scale cost overruns in infrastructure projects have stifled the growth of SME construction companies in the UK. Cost management errors have led to rework and claims, impeding organisations' productive outputs. Furthermore, 69% of construction works experience cost overruns in the UK, creating productivity difficulties in construction SMEs (KPMG, 2023).Consequently, GAI has not been implemented in the construction management process to optimise productivity and mitigate incidents of errors and overruns. Hence, construction cost management needs to advance towards the new models of productivity improvement in the planning and construction stages. The application of GAI in the construction cost management process of SME companies in the UK will empower productivity through BIM. This feasibility study will study the pilot integration of BIM and GAI, datasets, ethical considerations, regulations, interfaces, GAI requirements user interface (UI) and user experience (UX). Furthermore, GAI has not been applied in BIM cost management-based tools such as BlueBeam, CostX and PlainSwift. **This feasibility study intends to use existing construction cost datasets from an SME construction company to develop a systems architecture, a Foundation large language learning model (FLM) and a pilot web-based integration of BIM and GAI to automate the cost estimation to engender productivity.**