Integrating Internet of Things (IoT), Advanced Analytics and Immersive Computing to Enable Cognitive BIM

最近，通过将建筑信息建模（BIM）与新兴的数字技术（如物联网（IoT）和虚拟/增强现实（VR/AR））相结合来改善设施生命周期管理（FLCM）活动的兴趣激增。虽然这些新的协同增效作用显示出巨大的潜力，但重大的研究差距仍然限制了这些技术的共同效益。 主要挑战之一是这些技术学科的数据生态系统之间缺乏足够的语义互操作性。这一缺陷主要源于与封闭的数据生态系统的紧密联系，这阻碍了在设施的整个生命周期中进行有效的跨域数据交换。最近的研究使用正式本体和语义网技术，以解决互操作性的建成环境已经显示出可喜的成果。然而，这些方法的应用还有待于更广泛地研究BIM-IoT集成。 仅仅统一FLCM数据不会增加太多价值。收集到的各种数据应首先进行融合，然后转化为有用的信息/知识。此外，需要各种推理来检索相关信息/知识作为设计用于帮助目标从业者的决策支持系统（DSS）的输入。然而，由于人为干预仍将是一个关键因素，将数据分析与严肃游戏相结合，通过增强这些环境提供的可视化和交互，有望带来更多的机会。 为了提供BIM、物联网和严肃游戏环境的丰富编排，该提案研究了“认知BIM”：一个框架，用于创建以BIM为中心的、能够感知、学习和交互的设施的自我进化数字孪生模型。为了实现这一长期愿景，在7名学生的参与下制定了四个短期目标：两名博士，四个硕士学位，还有一个本科生首先，将研究本体和语义Web技术的应用，以整合BIM，IoT和记录的生命周期数据。接下来，将介绍一个数据分析框架，以应用各种先进的数据挖掘方法。然后，将开发一个基于本体的框架来检索上下文相关的信息/知识和饲料领域特定的DSS。最后，该框架将扩展到开发支持VR/AR的界面，并向各种利益相关者（例如，设计师，施工人员，设施管理人员），从而帮助他们解决基于建筑学的问题，并纳入他们的直观反馈。 拟议研究的结果将有助于建筑环境的可持续发展，为智慧城市的实现铺平道路，在智慧城市中，不同设施（如建筑物、桥梁和隧道）的单个数字孪生模型可以实时共享见解，并智能地集体行动。

Recently, there has been a surge of interest in the improvement of Facility Life-Cycle Management (FLCM) activities by integrating Building information modeling (BIM) with emerging digital technologies, such as Internet of Things (IoT) and Virtual/Augmented Reality (VR/AR). While these new synergies show great potential, major research gaps still limit the joint benefits of such technologies. One of the primary challenges is the lack of sufficient semantic interoperability between the data ecosystems of these technological disciplines. This shortcoming mainly stems from strong ties to closed data ecosystems, which impede effective cross-domain data exchange throughout the lifecycle of facilities. Recent research using formal ontologies and semantic web technologies to address interoperability for the built environment has shown promising results. However, the application of such approaches has yet to be more extensively investigated for BIM-IoT integration. Mere unification of FLCM data does not add much value. The diverse collected data should first be fused and then transformed into useful information/knowledge. Additionally, various inferences are required to retrieve relevant information/knowledge as input to Decision Support Systems (DSS) designed to assist target practitioners. Yet, since human intervention will remain as a key element, integrating data analytics with serious games promises further opportunities by enhancing the visualizations and interactions provided by such environments. To deliver rich orchestrations of BIM, IoT, and serious game environments, this proposal investigates "Cognitive BIM": a framework for creating BIM-centred and self-evolving digital twins of facilities capable of sensing, learning, and interacting. To meet this long-term vision, four short-term objectives have been established with the involvement of 7 students: two Ph.D., four M.Sc., and one undergraduate student. First, the application of ontologies and semantic web technologies will be investigated to integrate BIM, IoT, and recorded lifecycle data. Next, a data analytics framework will be presented to apply various advanced data-mining methods. Then, an ontology-based framework will be developed to retrieve contextually relevant information/knowledge and feed domain-specific DSS. Finally, the framework will be expanded to develop VR/AR-enabled interfaces and convey rich visualizations of timely insights to various stakeholders (e.g., designers, constructors, facility managers), thereby assisting them in heuristics-based problem-solving and incorporating their intuitive feedback. Results of the proposed research will contribute to the sustainable development of the built environment by paving the road for the actualization of smart cities in which individual digital twins of different facilities, such as buildings, bridges, and tunnels, can share insights in real-time and act intelligently and collectively.