Proactive Intelligent Construction Site Management Enabled by Automated On Site Monitoring (PRISM)

**PRISM**将开发一种数字化的施工规划和管理解决方案，为现场施工提供逐步变化的生产力提高。现场施工受到多种问题的影响，这些问题可能随着时间的推移而累积，从而导致整体计划延迟并增加项目风险。短期（前瞻性）规划旨在通过在日常基础上逐步做出更好的决策来避免/最大限度地减少这些延误。要做到这一点，关于任务进度、资源可用性/利用率和场地特定因素（如要移动的材料、地面条件、通道、天气）等问题的准确信息至关重要。然而，到目前为止，这些信息只能通过人工检查的反馈或任务完成后或问题已经出现时现场管理人员的“特别”更新获得。PRISM解决了这一挑战。它建立在固定和移动相机/扫描仪等数字设备创造的机会之上，这些设备可以实现连续的现场监测。在高速、低延迟5G网络等近期创新的推动下，这些新技术在建筑中的应用正在增加，并将加速。然而，这些解决方案目前需要专业人员来监控/解释图像，这可能会导致数据过载（用户无法知道将注意力集中在哪里），或者将其全部好处限制在特定/集中的场景中。因此，PRISM将开发一种分析引擎，该引擎使用基于人工智能的算法（应用于来自现场监控设备的过滤/分类数据馈送），以近乎实时的方式生成关键施工任务的目标信息。由此产生的信息流将集成到前瞻性规划中，使现场管理人员能够迅速对不断变化的问题作出反应，通过重新安排活动或重新分配资源来避免时间/成本关键的延误。该项目将开发一种预商用的“最小可行产品”解决方案，用于快速部署，并可以随着时间的推移而增强。PRISM将由英国SME **Trackchain**（数据分析）、**Glideology**（数据整理/分类）和**Aphex**（前瞻性规划）牵头，由领先的建筑承包商**BAM**和**Costain**以及建筑研究机构（**BRE**）提供支持。

**PRISM** will develop a digitally-enabled construction planning and management solution that delivers step-change productivity improvements for on-site construction.On-site construction is impacted by multiple issues that can accrue over time to cause overall programme delays and increase project risk. Short-term (Lookahead) planning aims to avoid/minimise these delays by enabling incrementally better decisions on a daily basis. To do this, accurate information on issues such as task progress, resource availability/utilisation and site-specific factors (e.g. materials to be moved, ground conditions, access, weather) is essential. However, to date such information is obtained only via feedback from manual inspection or from 'ad-hoc' updates from site managers after task completion or when issues have already arisen.PRISM addresses this challenge. It builds on opportunities created by digital devices such as fixed and mobile cameras/scanners that enable continuous site monitoring. The use of these new technologies in construction is growing and will accelerate, facilitated by recent innovations such as high speed, low-latency 5G networks. However, these solutions currently require expert staff to monitor/interpret images, potentially leading to data overload (with users unable to know where to focus attention), or limiting their full benefit to specific/focussed scenarios.PRISM will therefore develop an analysis engine that uses AI-based algorithms (applied to filtered/triaged data feeds from on-site monitoring devices) to generate targeted information on critical construction tasks in near real-time. The resulting information streams will be integrated into lookahead planning, enabling site mangers to rapidly react to evolving issues, avoiding time/cost-critical delays by rescheduling activities or reallocating resources. The project will develop a pre-commercial 'minimum-viable-product' solution for rapid deployment and which can be enhanced over time.PRISM will be led by UK SME **Trackchain** (data analytics) with **Glideology** (data collation/triaging) and **Aphex** (lookahead planning), supported by leading construction contractors **BAM** and **Costain**, and the Building Research Establishment (**BRE**).