A Generative AI-Enabled Design Tool: Analysing problematic projects for improved productivity and cost efficiency

本提案是一项可行性研究，探索一种新的生成式人工智能方法，解决由返工引起的时间和成本超支的关键施工挑战。**项目包括以下活动**:*探索开发人工智能设计工具作为解决方案的可行性，使用以前项目的数据来分析有问题的项目，确定需要额外监督和管理的问题，以避免返工和时间/成本超支。*通过需要解决方案的一方与能够开发解决方案的一方之间的合作，为建筑行业开发/部署应用人工智能解决方案提供人工智能专业知识。限制建筑行业生产力提高的关键因素是由于不可预见的挑战和返工造成的时间和成本超支。这个项目有潜力极大地提高生产力。**重点领域**我们的项目重点关注以下主题：**数据驱动决策：*** ** _更好的项目交付_**：使用过去的项目数据和行业范围的基准来开发准确的项目计划估算器；减少延迟/成本超支的可能性。** *_增强安全性_**：利用数据识别安全风险并采取主动缓解措施。** **质量控制** *：利用数据监控施工项目的工艺质量，以达到质量标准，避免返工。* *设计:* * * * _ * _ * *生成设计:使用人工智能产生新的设计选择基于设置参数/约束。** *_分析问题项目_**：使用人工智能分析以前建筑项目的数据，以识别模式并预测未来的结果。InnovationsOur项目为开发基于生成式人工智能的设计工具奠定了基础，该工具可以使用建筑公司记录的数据分析有问题的项目；提供有价值的见解，使他们能够采取积极主动的措施，有效地分配资源，做出明智的决定，以减轻风险；从而全面改善结果、生产力和成本效率。我们的解决方案为人工智能解决方案在建筑行业的应用提供了一个循序渐进的变化，从孤立的特定问题方法到基于整体系统的方法。**相关性**我们的项目探讨了使用生成式人工智能支持系统的以下功能的适用性：**数据分析**：分析大量项目数据；/ anomalies.2识别模式。**风险预测**：分析项目历史数据；识别/预测可能导致问题项目的风险/危害。**实时监控**：通过整合数据对正在进行的建设项目进行监控；发现与项目计划的偏差。**自然语言处理**：分析文本项目文档数据；提取有价值的见解。**标杆/比较分析**：将项目数据与行业标杆/最佳实践进行比较；识别执行低于预期或有较高成本/时间超支风险的项目。**Human-in-the-Loop：探索将人类纳入指导人工智能过程并影响结果。

This proposal is a **feasibility study** exploring a novel Generative AI approach addressing the critical construction challenge of time and cost overruns caused by re-work.**Project entails the following activities**:* Explore feasibility of developing an AI-enabled design tool as a solution using data from previous projects to analyse problematic projects identifying issues requiring additional oversight and management to avoid re-work and time/cost overruns.* Provide AI expertise in developing/deploying Applied AI solutions for construction industry through collaboration between a party in need of the solution and a party that can develop it.Critical factors limiting improvements in productivity in the construction industry are time and cost overruns due to unforeseen challenges and re-work. This project has the potential to unlock huge productivity improvements.**Areas of focus**Our project focuses on the following themes:**Data Driven Decision Making:*** **_Better project delivery_**: using past project data and industry wide benchmarks to develop estimators for accurate project plans; reducing likelihood of delays/cost overruns.* **_Enhanced safety_**: using data to identify safety risks and proactive mitigating measures.* _**Quality control**_: using data to monitor quality of workmanship on construction projects to meet quality standards and avoid rework.**Design:*** _**Generative design**_: using AI to generate new design options based on set parameters/constraints.* **_Analysing problematic projects_**: using AI to analyse data from previous construction projects to identify patterns and predict future outcomes.InnovationsOur project sets the foundations for developing a Generative AI-enabled design tool to analyse problematic projects using data recorded by construction companies; offering valuable insights enables them to take proactive measures, allocate resources effectively, make informed decisions to mitigate risks; resulting in overall improved outcomes, productivity, and cost efficiency.Our solution provides a step-change on how AI solutions are applied in the construction industry from a siloed problem-specific approach to a holistic system-based approach.**Relevance**Our project explores suitability of using Generative AI-enabled systems for the following capabilities:1. **Data analysis**: analyse large volumes of project data; identifying patterns/anomalies.2. **Risk prediction**: analyse historical project data; identifying/predicting risks/hazards likely leading to problematic projects.3. **Real-Time Monitoring**: monitoring ongoing construction projects by integrating data; detecting deviations from the project plan.4. **Natural Language Processing**: analyse textual project documentation data; extracting valuable insights.5. **Benchmarking/Comparative Analysis**: compare project data with industry benchmarks/best practices; identifying projects performing below expectations or at higher cost/time overrun risks.6. **Human-in-the-Loop**: explore inclusion of humans in guiding AI processes and influencing results.