clouds. Finally, Arif and Khan [93] used Dynamo to convert data collected through traditional site surveys to the BIM model in Revit.
云最后，Arif和Khan [93]使用Dynamo将通过传统现场调查收集的数据转换为Revit中的BIM模型。

Architects can use VP to generate forms and design options. Many recent studies demonstrated the use of VP for automated modelling in a BIM process. Each study proposed specific algorithms to generate specific forms or solve specific issues. There are still rooms for researchers to propose new algorithms to solve architectural issues related to form generation.
建筑师可以使用VP生成表单和设计选项。最近的许多研究表明，在BIM过程中使用VP进行自动建模。每项研究都提出了特定的算法来生成特定的表单或解决特定的问题。研究人员仍然有空间提出新的算法来解决与表单生成相关的建筑问题。

Building performance analysis is calculating the various building performance metrics, which should be done during the design phases to find a better design solution [124]. The performance of the building can be anything that can be measured quantitively, such as energy efficiency, ventilation performance, or lighting performance. Several software products are used to perform the performance analysis. Some are standalone products; others are plug-ins in BIM software. VP also has plug-in libraries that can be used for performance analysis. Using VP for performance analysis is more flexible because users can control all variables and use specific equations that may differ based on their standards. Furthermore, VP can perform optimization algorithms that could be used to optimize the design. As displayed in Table 4, Revit and Dynamo were used in $63\mathrm{\%}$$63\mathrm{\%}$63%63 \% and $56\mathrm{\%}$$56\mathrm{\%}$56%56 \% of the performance analysis and optimization studies. At the same time, Rhinoceros and its VP tool, Grasshopper, were used in only $56\mathrm{\%}$$56\mathrm{\%}$56%56 \% of the studies.
建筑性能分析是计算各种建筑性能指标，应在设计阶段进行，以找到更好的设计解决方案[124]。建筑物的性能可以是任何可以定量测量的东西，例如能源效率、通风性能或照明性能。几个软件产品用于执行性能分析。有些是独立的产品;有些是BIM软件中的插件。VP还具有可用于性能分析的插件库。使用VP进行性能分析更灵活，因为用户可以控制所有变量，并使用可能因其标准而异的特定方程。此外，VP可以执行可用于优化设计的优化算法。如表4所示，在 $63\mathrm{\%}$$63\mathrm{\%}$63%63 \% 和 $56\mathrm{\%}$$56\mathrm{\%}$56%56 \% 性能分析和优化研究中使用了Revit和Dynamo。 与此同时，Rhinoceros及其VP工具Grasshopper仅用于 $56\mathrm{\%}$$56\mathrm{\%}$56%56 \% 项研究。

Most studies used VP to assess the building envelope’s thermal performance. For example, Seghier et al. [53] developed an envelope thermal transfer value (ETTV) workflow using VP, and Natephra et al. [42] used Dynamo to develop an overall thermal transfer value (OTTV) evaluation tool. Salimzadeh et al. [47] used Dynamo to calculate cumulative solar radiation on building surfaces. Furthermore, some studies evaluate the thermal performance and optimise the design based on the thermal performance. For instance, Abbasi and Noorzai [63] used Revit, EnergyPlus and Grasshopper to run a parametric simulation for the building’s operational energy estimation and Hypervolume Estimation Algorithm for Multi-objective Optimization (HypE) to find the trade-off between embodied and operation energy focused on renewable energy use. Rahmani Asl et al. [9] introduced a workflow using Dynamo to optimize a house design based on the energy performance factor and daylighting performance factor. Kensek [8] demonstrated VP scripts to create models and optimize the designs based on various performance analyses. Seghier et al. [56] developed a prototype system to minimize an overall thermal transfer value (OTTV) and the building envelope’s retrofit cost using Dynamo and MATLAB. Some studies used an optimization package in Dynamo called Optimo to optimize two main insulation design aspects and the thermal performance of the building envelope [64] and optimize window design by considering heat transfer and natural lighting [65]. Also, Dynamo has a generative design function which some studies used to optimize the building envelopes [87]. The best position to install the PV panels and solar thermal system also used VP to calculate the solar radiation and optimization algorithms to find the optimal solutions $[47,52]$$[47,52]$[47,52][47,52]. LCA, embodied carbon and energy are also among the current topics that use BIM and VP to calculate and generate automated reports [51,58,60]. VP can also analyze walking paths based on site locations, destinations (Li et al., 2019), and the shortest walk for fire escape route planning [110].
大多数研究使用VP来评估建筑围护结构的热性能。例如，Seghier等人。[53]使用VP开发了一个信封热传递值（ETTV）工作流程，Natephra等人。[42]使用Dynamo开发了一个整体热传递值（OTTV）评估工具。Salimzadeh等人[47]使用Dynamo计算建筑物表面的累积太阳辐射。此外，一些研究评估热性能并基于热性能优化设计。例如，Abbasi和Noorzai [63]使用Revit、EnergyPlus和Grasshopper对建筑物的运行能耗估计和多目标优化的超体积估计算法（HypE）进行参数化模拟，以找到集中在可再生能源使用上的体现能源和运行能源之间的权衡。Rahmani Asl等人。[9]介绍了一种使用Dynamo的工作流程，该工作流程基于能源性能因子和采光性能因子来优化房屋设计。 Kensek [8]演示了基于各种性能分析创建模型和优化设计的VP脚本。Seghier等人。[56]开发了一个原型系统，使用Dynamo和MATLAB最大限度地减少总热传递值（OTTV）和建筑围护结构的改造成本。一些研究在Dynamo中使用了一个名为Optimo的优化包，以优化两个主要的隔热设计方面和建筑围护结构的热性能[64]，并通过考虑传热和自然采光来优化窗户设计[65]。此外，Dynamo具有生成设计功能，一些研究用于优化建筑物外壳[87]。光伏板和太阳能热系统的最佳安装位置也使用VP来计算太阳辐射和优化算法来找到最佳解决方案 $[47,52]$$[47,52]$[47,52][47,52] 。LCA、隐含碳和能源也是当前使用BIM和VP计算和生成自动报告的主题[51，58，60]。 VP还可以基于站点位置、目的地来分析步行路径（Li等人，2019），以及火灾逃生路线规划的最短步行距离[110]。

Moreover, other building performances can be achieved by using VP. For example, Hamidavi et al. [84] used Dynamo and Python for structural optimization. Khan et al. [94] used Grasshopper to optimize the safety egress plan for excavation. Additionally, Kiamili et al. [48] created a Dynamo script to evaluate the functional flexibility of building designs. Ostrowska-Wawryniuk [86] used Dynamo to generate the wall panel pattern and calculate the material waste. Furthermore, Sheikhkhoshkar et al. [43] used Dynamo to extract data to an Excel file and then import it to MATLAB to optimize the concrete joint layout. Finally,
此外，使用VP还可以实现其他建筑性能。例如，Hamidavi等人[84]使用Dynamo和Python进行结构优化。Khan等人[94]使用Grasshopper优化挖掘的安全出口计划。此外，Kiamili等人[48]创建了一个Dynamo脚本来评估建筑设计的功能灵活性。Ostrowska-Wawryniuk [86]使用Dynamo生成墙板图案并计算材料浪费。此外，Sheikhkhoshkar等人。[43]使用Dynamo将数据提取到Excel文件中，然后将其导入MATLAB以优化混凝土接缝布局。最后，

Chang et al. [44] integrated temperature and humidity data from IoT sensors with a BIM model to calculate and visualize human comfort in a room.
Chang等人[44]将物联网传感器的温度和湿度数据与BIM模型相结合，以计算和可视化房间中的人体舒适度。

Building performance analysis and optimization using VP is also the mainstream of BIM-related research. Most of the research depends on analysis plug-ins such as Ladybug, Honeybee, and Butterfly and simulation engines such as Energy Plus and Radiance, while some researchers developed their calculation engine. It is worth noting that most of the simulation-based studies used Grasshopper as the VP platform due to the maturity level of its simulation plugins compared to Dynamo. Future research can be done based on other performance analyses, such as wind simulation, embodied carbon assessment, and artificial lighting analysis, to name a few.
利用虚拟样机进行建筑性能分析和优化也是BIM相关研究的主流。大多数研究依赖于分析插件，如Ladybug，Honeybee和Butterfly以及模拟引擎，如Energy Plus和Radiance，而一些研究人员开发了他们的计算引擎。值得注意的是，大多数基于模拟的研究都使用Grasshopper作为VP平台，因为与Dynamo相比，其模拟插件的成熟度更高。未来的研究可以基于其他性能分析进行，例如风模拟，隐含碳评估和人工照明分析，仅举几例。

VP-enabled computational BIM algorithms in building research have other roles, including connecting to hardware, data interoperability and documentation.
在建筑研究中，支持VP的计算BIM算法还有其他作用，包括连接硬件、数据互操作性和文档。

Several studies used VP to import data from hardware, especially sensors or control hardware. For instance, Dynamo scripting was used to blend thermal room parameters extracted from the BIM model with real on-site data using sensors for performance thermal end energy monitoring [97]. VP was also employed to develop a system that automatically extracts and sends all essential information about the facility (e.g. type of failed sensors) to IoT companies as part of the facility management process [98]. Furthermore, Dynamo was used to create Revit key schedules based on sensors and geophone data collected on-site for asset management and structural health monitoring [100]. Another study used Dynamo to integrate light sensor data into a BIM model, change the parameters of the 3D model, and send the data to servo motors to change the physical model [41]. Finally, Dynamo was used to embed and visualize IoT sensor data in a BIM model [103]. However, the use of VP to import data from sensors is not real-time; most research used text files or Excel files to store data before importing them into BIM. The sensors employed in most research are light, temperature, and humidity sensors.
一些研究使用VP从硬件导入数据，特别是传感器或控制硬件。例如，使用Dynamo脚本将从BIM模型中提取的热室参数与真实的现场数据混合，使用传感器进行性能热端能量监测[97]。VP还被用来开发一个系统，该系统可以自动提取有关设施的所有重要信息（例如故障传感器的类型）并将其发送给物联网公司，作为设施管理过程的一部分[98]。此外，Dynamo还用于根据现场收集的传感器和地震检波器数据创建Revit关键时间表，用于资产管理和结构健康监测[100]。另一项研究使用Dynamo将光传感器数据集成到BIM模型中，更改3D模型的参数，并将数据发送到伺服电机以更改物理模型[41]。最后，Dynamo用于在BIM模型中嵌入和可视化物联网传感器数据[103]。 然而，使用VP从传感器导入数据并不是实时的;大多数研究在将数据导入BIM之前使用文本文件或Excel文件来存储数据。在大多数研究中使用的传感器是光、温度和湿度传感器。

Other studies used VP to create dataflows between different software to improve data interoperability. For example, a study used Grasshopper to make live-connection between CAD-based software (Rhinoceros) and BIM-based software (Revit) [70]. As a result, any changes to the model made in Rhinoceros were automatically made in Revit and vice versa. Similarly, a study used Dynamo to connect a web-based product information management (PIM) database (OpenPIM) and the corresponding IFC model developed in this research [99]. Another study used Dynamo and Revit to demonstrate a platform that integrates sensor data to visualize in a BIM model [44]. Furthermore, VP can automate or semiautomate the preparation of design documents, such as generating schedules for model element specifications [101].
其他研究使用VP在不同软件之间创建数据流，以提高数据互操作性。例如，一项研究使用Grasshopper在基于CAD的软件（Rhinoceros）和基于BIM的软件（Revit）之间进行实时连接[70]。因此，在Rhinoceros中对模型所做的任何更改都会自动在Revit中进行，反之亦然。类似地，一项研究使用Dynamo连接基于Web的产品信息管理（PIM）数据库（OpenPIM）和本研究中开发的相应IFC模型[99]。另一项研究使用Dynamo和Revit演示了一个平台，该平台集成了传感器数据以在BIM模型中可视化[44]。此外，VP可以自动化或半自动化设计文档的准备，例如为模型元素规范生成时间表[101]。

The roles of VP in BIM-based research are broad. Researchers can also use VP in other roles that were not captured in the search protocol used in the design of this review paper.
VP在BIM研究中的作用是广泛的。研究人员还可以在本综述论文设计中使用的检索方案中未捕获的其他角色中使用VP。

The adoption of digital technologies in the AECO industry is increasing rapidly, especially with the integration of Construction 4.0 technologies [125]. Therefore, the necessity for the AECO industry to maintain the same rate of progress in adopting these new technologies is now more crucial than ever [66,117]. Furthermore, it is argued that using computational BIM in the AECO industry has created more opportunities for scholars to develop and test innovative solutions for the AECO industry than ever before. This will eventually speed up the development of the AECO industry and therefore contribute to achieving Construction 4.0. This study comprehensively reviewed the application of computational BIM through bibliometric and content analysis. The
AECO行业中数字技术的采用正在迅速增加，特别是随着建筑4.0技术的集成[125]。因此，AECO行业在采用这些新技术方面保持相同进展速度的必要性现在比以往任何时候都更加重要[66，117]。此外，有人认为，在AECO行业中使用计算BIM为学者们创造了比以往任何时候都更多的机会来开发和测试AECO行业的创新解决方案。这将最终加速AECO行业的发展，从而有助于实现建筑4.0。本文通过文献计量和内容分析，对计算BIM的应用进行了全面综述。的
following sub-sections discuss the status of computational BIM, its key topics, and future research directions in the AECO industry. Then, in section 6.4, a research framework is proposed to summarise the findings of this study and future directions.
下面的小节讨论了计算BIM的现状，它的关键主题，以及AECO行业未来的研究方向。然后，在第6.4节中，提出了一个研究框架，总结了本研究的结果和未来的发展方向。

Researchers in the AECO industry have used VP-enabled computational BIM workflows in building research since 2014 (first journal article), and it has experienced a significant increase since 2019. This agrees with Collao et al. [6] findings, which focused on reviewing computational BIM applications in infrastructure projects. Compared to conventional text-based programming, VP tools have become vital for scholars to develop models and system prototypes in BIM-based research for building projects. This is because VP tools are more user-friendly to use by scholars in the built environment who are often not wellequipped with the knowledge required for text-based programming [18,126]. Additionally, developing a well-functioning system prototype with a comprehensive user interface is usually not a priority for researchers willing to investigate or solve a specific building-related problem. Instead, scholars in the built environment often focus on the problem itself and the outcome of the proposed solution. Therefore, VP tools are often a good choice for this task within BIM-based building research.
自2014年以来，AECO行业的研究人员已经在建筑研究中使用了支持VP的计算BIM工作流（第一篇期刊文章），并且自2019年以来已经经历了显着的增长。这与Collao等人[6]的研究结果一致，该研究重点是审查基础设施项目中的计算BIM应用。与传统的基于文本的编程相比，VP工具已经成为学者在基于BIM的研究中开发模型和系统原型的重要工具。这是因为VP工具对于构建环境中的学者来说更加用户友好，他们通常不具备基于文本的编程所需的知识[18，126]。此外，对于愿意调查或解决特定建筑相关问题的研究人员来说，开发具有全面用户界面的功能良好的系统原型通常不是优先事项。 相反，建筑环境领域的学者们往往关注问题本身和所提出的解决方案的结果。因此，在基于BIM的建筑研究中，VP工具通常是执行此任务的良好选择。

Additionally, the current cooperation between countries and authors is relatively weak, and there is little cooperation between countries and institutions. In this analysis timeframe, Italy, the United States, and Japan were the top three countries developing computational BIM-based solutions for the AECO industry-indicating that VP’s current status and development in developing countries are still very weak.
此外，目前国家与作者之间的合作相对薄弱，国家与机构之间的合作也很少。在此分析时间范围内，意大利、美国和日本是为AECO行业开发基于BIM的计算解决方案的前三个国家，这表明VP在发展中国家的现状和发展仍然非常薄弱。

Based on an in-depth review of the selected papers, this study identified six key research themes where computational BIM has been applied intensely in building research: building performance analysis, design optimization, heritage BIM (HBIM), project management, and material take-off and facility management. The research under these six topics investigated different workflows, methods, and systems throughout the building lifecycle, focusing more on the design and operation stages. During the design stages in a BIM environment, scholars developed different workflows to support decision-making related to building performance analysis and design optimization. Early performance analysis is crucial during the design stage [127,128], and computational BIM has made it possible to develop customized performance analysis workflows for this stage, currently unavailable in commercial tools [114]. In terms of building performance, several studies focused on implementing computational BIM to develop methods and systems for building envelope design analysis and optimization [42,55,56,65,128]. In addition, recent publications integrated optimization algorithms with computational BIM for different optimization problems related to building projects [8,9,47,52,56]. Other studies focused on developing different computational BIM methods to automate LCA [48,51,58,60,63]. For design optimization, scholars employed computational BIM to automate the design process of different components in the building with different goals, such as optimizing material usage and, therefore, reducing waste or cost [6,39,84,86-89].
基于对所选论文的深入审查，本研究确定了计算BIM在建筑研究中广泛应用的六个关键研究主题：建筑性能分析，设计优化，遗产BIM（HBIM），项目管理以及材料起飞和设施管理。这六个主题下的研究调查了整个建筑生命周期中不同的工作流程，方法和系统，更侧重于设计和运营阶段。在BIM环境中的设计阶段，学者们开发了不同的工作流程，以支持与建筑性能分析和设计优化相关的决策。早期性能分析在设计阶段至关重要[127，128]，计算BIM使得为这一阶段开发定制的性能分析工作流程成为可能，目前商业工具中无法使用[114]。 在建筑性能方面，几项研究侧重于实施计算BIM，以开发建筑围护结构设计分析和优化的方法和系统[42，55，56，65，128]。此外，最近的出版物将优化算法与计算BIM集成，用于与建筑项目相关的不同优化问题[8，9，47，52，56]。其他研究侧重于开发不同的计算BIM方法来自动化LCA [48，51，58，60，63]。对于设计优化，学者们采用计算BIM来自动化建筑中不同组件的设计过程，这些组件具有不同的目标，例如优化材料使用，从而减少浪费或成本[6，39，84，86 -89]。

During the operation stage, scholars focused on problems related to facility management and heritage building conservation. Computational BIM was mainly employed for facility management-related studies to solve problems related to object detection from point cloud data collected using 3D scanning [96], as well as developing systems for sensors’ data management for building maintenance activities [99]. Similarly, computational BIM was used for heritage buildings to convert
在运营阶段，学者们主要关注设施管理和文物建筑保护的相关问题。计算BIM主要用于与设施管理相关的研究，以解决与使用3D扫描收集的点云数据中的对象检测相关的问题[96]，以及开发用于建筑维护活动的传感器数据管理系统[99]。同样，计算BIM被用于传统建筑，
point cloud data to BIM models. In this task, the automation and accuracy of BIM model creation were one of the main objectives of the proposed methods. Additionally, several studies used computational BIM to support decision-making in historical building conservation works [61].
点云数据到BIM模型。在这项任务中，BIM模型创建的自动化和准确性是所提出的方法的主要目标之一。此外，一些研究使用计算BIM来支持历史建筑保护工作的决策[61]。

Moreover, computational BIM has been employed in building research to assist in four tasks: data management, process automation, parametric modelling, performance assessment and simulation. Whereas data management, performance analysis and simulation gained the most focus from scholars due to the current capabilities of VP tools in performing these tasks. Overall, VP tools were mainly employed to extract and manage the data of the BIM workflow and create new functionalities in the BIM tools to increase the level of automation and achieve compliance with regulations. Besides, VP tools were used as a key element for data interoperability between the different modules of the proposed systems.
此外，计算BIM已被用于建筑研究，以协助四项任务：数据管理，过程自动化，参数建模，性能评估和模拟。而数据管理，性能分析和仿真获得了学者最关注的，由于目前的VP工具在执行这些任务的能力。总体而言，VP工具主要用于提取和管理BIM工作流程的数据，并在BIM工具中创建新功能，以提高自动化水平并符合法规。此外，虚拟样机工具被用作拟议系统不同模块之间数据互操作性的关键要素。

Although some initial studies and literature have established a good foundation for applying computational BIM in building research [8,10,18,128], there are still many research gaps and issues that need to be addressed in the AECO industry.
虽然一些初步的研究和文献已经为在建筑研究中应用计算BIM奠定了良好的基础[8，10，18，128]，但AECO行业仍有许多研究空白和问题需要解决。
6.3.1. Data interoperability between different VP tools and BIM software
6.3.1.不同VP工具和BIM软件之间的数据互操作性

To improve data interoperability between building stakeholders, the AECO industry is adopting the open BIM framework, which employs IFC files as a neutral platform that can be read and edited by any BIM software. Nevertheless, the application programming interface (API) that connects VP with BIM software is still limited by software developers. This indicates that the interoperability among the studied VP tools has undergone less development than the connectivity between different BIM software packages [10]. It is believed that the BIM platform must enable an API for VP to connect with; otherwise, manual extraction of data from BIM or importing data to BIM is required. Also, if researchers need to connect data from the BIM platform to other analytic software, that software must have a compatible API with the VP tool. Based on this review, it was clear that researchers selected the VP tool based on which BIM tool was used and the available data interoperability workflows. For instance, all researchers who used the BIM authoring tool Revit used Dynamo as the VP tool because of its wellestablished connectivity.
为了提高建筑利益相关者之间的数据互操作性，AECO行业正在采用开放的BIM框架，该框架采用IFC文件作为中立平台，可以由任何BIM软件读取和编辑。然而，连接VP与BIM软件的应用程序编程接口（API）仍然受到软件开发人员的限制。这表明所研究的VP工具之间的互操作性比不同BIM软件包之间的连接性经历了更少的发展[10]。认为BIM平台必须支持VP连接的API，否则需要手动从BIM中提取数据或将数据导入BIM。此外，如果研究人员需要将来自BIM平台的数据连接到其他分析软件，则该软件必须具有与VP工具兼容的API。根据这一审查，很明显，研究人员选择了基于BIM工具和可用的数据互操作性工作流的VP工具。 例如，所有使用BIM创作工具Revit的研究人员都使用Dynamo作为VP工具，因为它具有良好的连接性。

Even though several researchers [9,42] acknowledged the usability of VP tools in data transfer between different software, the current VP tools’ file format still depends on the software developers. For instance, the file extension used by Dynamo is not read by Grasshopper and vice versa. It is suggested to expand the concept of Open BIM not only to the BIM tools but also to the VP tools. This would make data interoperability in the AECO industry more efficient and streamlined. It is worth noting that there is already some effort towards integrating different VP tools into more than one BIM authoring tool. For instance, a plugin named Rhino.Inside®.Revit [129] was developed to allow interaction between Rhinoceros, Grasshopper and Revit. However, based on this review findings, none of the reviewed papers has implemented this plugin in their workflows. Therefore, more exploration of the practicality and capability of this workflow to overcome data interoperability issues when using BIM and VP is needed.
尽管一些研究人员[9，42]承认VP工具在不同软件之间的数据传输中的可用性，但当前VP工具的文件格式仍然取决于软件开发人员。例如，Dynamo使用的文件扩展名不会被Grasshopper读取，反之亦然。建议将开放式BIM的概念不仅扩展到BIM工具，还扩展到VP工具。这将使AECO行业的数据互操作性更加高效和简化。值得注意的是，已经有一些努力将不同的VP工具集成到多个BIM创作工具中。例如，一个名为Rhino.Inside®.Revit [129]的插件被开发出来，允许Rhinoceros、Grasshopper和Revit之间的交互。然而，根据这一审查结果，没有一篇被审查的论文在其工作流程中实现了此插件。 因此，需要更多地探索此工作流的实用性和能力，以克服使用BIM和VP时的数据互操作性问题。

Even though the development of VP tools and their application in the BIM environment has gone a long way, it still has critical limitations in terms of functionalities and performance compared to traditional textbased programming. For instance, due to the limitation of the traditional BIM-modelling approach in dealing with complex parametric geometry and high degrees of variation, many building industry
尽管VP工具的开发及其在BIM环境中的应用已经走了很长的路，但与传统的基于文本的编程相比，它在功能和性能方面仍然存在严重的局限性。例如，由于传统的BIM建模方法在处理复杂的参数化几何形状和高度变化方面的局限性，许多建筑业
practitioners working on parametric modelling seem to avoid using VP in the BIM-based approach. Wortmann and Tunçer [6] highlighted that they instead develop customized digital workflows based on the “Building Information Generation” approach to managing the design, prefabrication, and assembly using a CAD environment, where Rhinoceros and its VP tool Grasshopper are the key software. Additionally, this approach often relies on a Grasshopper plugin called Elefront that augments CAD geometry with user-defined properties. The properties turn geometry into customized BIM-like objects that contain nongeometric data [6]. This issue related to the flexibility of parametric modelling in the BIM environment is also relevant to scholars developing new BIM-based methods and models using VP in academia. It is argued that many scholars tend to develop their methods and models without adopting the BIM approach because of the challenges in the BIM ecosystem in dealing with complex parametric geometries. Consequently, it is necessary to improve the functionalities of the BIM tools regarding these limitations.
从事参数建模的从业者似乎避免在基于BIM的方法中使用VP。Wortmann和Tunçer [6]强调，他们基于“建筑信息生成”方法开发定制的数字工作流程，使用CAD环境管理设计，预制和装配，其中Rhinoceros及其VP工具Grasshopper是关键软件。此外，这种方法通常依赖于名为Elefront的Grasshopper插件，该插件使用用户定义的属性来增强CAD几何形状。这些属性将几何体转换为包含非几何数据的自定义BIM类对象[6]。这个问题与BIM环境中参数化建模的灵活性有关，也与学术界使用VP开发新的基于BIM的方法和模型的学者有关。 有人认为，许多学者倾向于开发他们的方法和模型，而不采用BIM的方法，因为在BIM生态系统中的挑战，在处理复杂的参数化几何。因此，有必要针对这些限制改进BIM工具的功能。

Additionally, the ability of VP is limited by their library and plug-ins (or packages). For example, Grasshopper tool has more plug-ins and simulation engines than Dynamo. Therefore, some operations can be carried out by a single node in Grasshopper, whereas in Dynamo, many nodes are required to develop the same script. Furthermore, most VP tools run slower than text-based programming [9]. Also, most VP tools do not support multi-core CPU and GPU processors, making the execution slow [122]. Finally, most VPs cannot compile source code. Therefore, the developed system will share all source code, but some developers or researchers might not want to share their intellectual properties publicly.
此外，VP的能力受到其库和插件（或包）的限制。例如，Grasshopper工具比Dynamo有更多的插件和模拟引擎。因此，在Grasshopper中，一些操作可以由单个节点执行，而在Dynamo中，需要许多节点来开发相同的脚本。此外，大多数VP工具运行速度比基于文本的编程慢[9]。此外，大多数VP工具不支持多核CPU和GPU处理器，使得执行缓慢[122]。最后，大多数VP不能编译源代码。因此，开发的系统将共享所有源代码，但一些开发人员或研究人员可能不希望公开共享他们的知识产权。

BIM integration with different technologies and design paradigms caught the interest of many researchers in the last decades. Examples of BIM integration include integration to GIS [74], integration to different green rating systems [130], integration to Blockchain technology [125] and more. Based on this review’s findings, VP tools played a key role in achieving such integration. Therefore, it is believed that more development of the VP tools is required to amplify the adoption of BIM in various stages of the project lifecycle and expand its integration to more technologies and workflows. Furthermore, based on the current development in the VP tools and the finding of the reviewed papers, it is argued that there is a need to integrate more optimization algorithms in the current VP tool used in the BIM environment, such as Dynamo. Due to this limitation, many researchers [43,56,95,96] developed BIM-based prototype systems, and they used third-party software such as MATLAB for the optimization part of the system instead of using an optimization algorithm integrated into the VP tool. This might complicate data interoperability between the different modules of the developed system and result in a system that is not user-friendly for non-MATLAB users.
在过去的几十年里，BIM与不同技术和设计范式的集成引起了许多研究人员的兴趣。BIM集成的例子包括与GIS的集成[74]，与不同的绿色评级系统的集成[130]，与区块链技术的集成[125]等等。根据这次审查的结果，VP工具在实现这种整合中发挥了关键作用。因此，人们认为，需要对VP工具进行更多的开发，以在项目生命周期的各个阶段扩大BIM的采用，并将其集成扩展到更多的技术和工作流程。此外，基于当前VP工具的发展和综述论文的发现，有人认为，有必要在BIM环境中使用的当前VP工具（如Dynamo）中集成更多的优化算法。 由于这种限制，许多研究人员[43，56，95，96]开发了基于BIM的原型系统，他们使用第三方软件（如MATLAB）进行系统的优化部分，而不是使用集成到VP工具中的优化算法。这可能会使开发系统的不同模块之间的数据互操作性复杂化，并导致系统对非MATLAB用户不友好。

Moreover, the integration of simulation engines like Energy Plus and Radiance is required for the VP tools employed in the BIM environment. Currently, Dynamo, which is the most used VP tool in the BIM environment, lacks all sorts of simulation engines and workflows. Instead, all simulation engines are integrated into Grasshopper, which is less used in BIM-based building research based on this review’s findings. Another niche which should be developed in VP tools for BIM is the integration with other technologies, such as AI-based models for Machine learning (ML) and computer vision (CV), to name a few. Current VP tools for BIM (e.g., Dynamo) do not contain any workflows to implement ML comprehensively in the BIM project lifecycle. It is worth noting that the Grasshopper has several plugins for different types of simulation (e.g., Ladybug and Honeybee plugin) [131], optimization algorithms (e.g., Octopus plugin) [129], and ML models (e.g., LunchBoxML plugin) [132] developed for different applications in the building industry. However, based on this review, the application of these plugins in BIM-related research is minimal. Furthermore, Grasshopper is more integrated into the BIM authoring tool ArchiCAD, which was found to be less employed
此外，BIM环境中使用的VP工具需要集成Energy Plus和Radiance等仿真引擎。目前，Dynamo是BIM环境中最常用的VP工具，缺乏各种仿真引擎和工作流。相反，所有的模拟引擎都集成到Grasshopper中，根据本文的研究结果，Grasshopper在基于BIM的建筑研究中使用较少。在BIM的VP工具中应该开发的另一个利基是与其他技术的集成，例如用于机器学习（ML）和计算机视觉（CV）的基于AI的模型。当前用于BIM的VP工具（例如，Dynamo）不包含任何在BIM项目生命周期中全面实施ML的工作流。值得注意的是，Grasshopper有几个插件用于不同类型的模拟（例如，Ladybug和Honeybee插件）[131]，优化算法（例如，Octopus plugin）[129]，以及ML模型（例如：，LunchBoxML plugin）[132]为建筑行业的不同应用而开发。然而，根据这篇评论，这些插件在BIM相关研究中的应用是最小的。 此外，Grasshopper更多地集成到BIM创作工具ArchiCAD中，而ArchiCAD的使用率较低
in the reviewed studies than Revit. Therefore, the extension of the application of these plugins to BIM-based workflows would positively impact the development of BIM-related research and would eventually help in solving more building-related issues.
在审查的研究比Revit。因此，将这些插件的应用扩展到基于BIM的工作流程将对BIM相关研究的发展产生积极影响，并最终有助于解决更多与建筑相关的问题。

Based on the discussion above, a research framework is proposed to summarise and link key research themes and areas, the methodological role of VP-enabled computational BIM algorithm application in building research and future research directions, as illustrated in Fig. 8. These future research directions can help software developers, practitioners, and scholars derive a clear, in-depth understanding of computational BIM application in building research and its prospects for the future AECO industry.
基于上述讨论，提出了一个研究框架，以总结和链接关键的研究主题和领域，VP支持的计算BIM算法应用在建筑研究中的方法学作用和未来的研究方向，如图8所示。这些未来的研究方向可以帮助软件开发人员，从业者和学者获得一个清晰，深入了解计算BIM在建筑研究中的应用及其对未来AECO行业的前景。

The move forward to increase efficiency and productivity through automation and data-driven approaches in the AECO industry has closely linked the interdisciplinary research between the AECO industry and computer science. Computer programming and particularly VP, has become a primary tool for building practitioners and researchers to automate repetitive tasks and optimize design workflows, particularly when working in a BIM environment where data management is critical. This review revealed the state-of-the-art development trends, limitations and future direction of computational BIM application in building research. Six key research themes on the application of computational BIM have been unveiled: building performance analysis, design optimization, HBIM, project management, MTO and facility management. Moreover, four methodological roles of VP-enabled computational BIM application in these topics have been revealed: data management, process automation, parametric modelling and performance analysis and optimization. The study’s main contribution is identifying prominent areas and publication outlets for the continued application of computational BIM in building research.
AECO行业通过自动化和数据驱动的方法提高效率和生产力的发展将AECO行业和计算机科学之间的跨学科研究紧密联系在一起。计算机编程，特别是VP，已成为建筑从业者和研究人员自动化重复任务和优化设计工作流程的主要工具，特别是在数据管理至关重要的BIM环境中工作时。本文对计算BIM在建筑研究中的应用现状、局限性和未来发展方向进行了综述。计算BIM应用的六个关键研究主题已经公布：建筑性能分析，设计优化，HBIM，项目管理，MTO和设施管理。 此外，VP使能的计算BIM应用程序在这些主题中的四个方法角色已经被揭示：数据管理，过程自动化，参数化建模和性能分析和优化。该研究的主要贡献是确定了在建筑研究中继续应用计算BIM的突出领域和出版渠道。

A research framework was developed to link the state-of-the-art computational BIM application in building research with the identified research gaps and future research directions. This will assist researchers in systematically acquiring and identifying recent hotspots and trends in this research area. Furthermore, the proposed research framework will promote the development and adoption of computational BIM in the AECO industry by bridging gaps among researchers in academia, software developers, and industry practitioners.
一个研究框架的开发，以连接最先进的计算BIM在建筑研究中的应用与确定的研究差距和未来的研究方向。这将有助于研究人员系统地获取和确定这一研究领域的最新热点和趋势。此外，拟议的研究框架将通过弥合学术界研究人员、软件开发人员和行业从业人员之间的差距，促进AECO行业计算BIM的开发和采用。

Despite the contributions of this paper, this study has several limitations. Firstly, the state-of-the-art on the application of computational BIM in the AECO industry is systematically reviewed; the retrieval scope was limited to research which dealt with building projects only. Secondly, other studies that used VP-enabled computational algorithms outside the BIM environment were excluded as the scope of study considered only those claiming to use the BIM approach. Therefore, broader coverage of various industries, such as infrastructure or larger scales, such as urban scale, could provide more meaningful data, especially from the perspective of software development. Finally, it is worth noting that the results obtained in this study are dynamic and will change over time. Therefore, updated publications should be included according to the cut-off date of May 2022 for this review, and it is suggested that this review be replicated over the next few years.
尽管本文的贡献，这项研究有几个局限性。首先，系统回顾了计算BIM在建筑工程领域的应用现状，检索范围仅限于建筑工程领域。其次，排除了在BIM环境之外使用VP启用的计算算法的其他研究，因为研究范围仅考虑声称使用BIM方法的研究。因此，更广泛地覆盖各种行业，如基础设施或更大的规模，如城市规模，可以提供更有意义的数据，特别是从软件开发的角度来看。最后，值得注意的是，本研究所获得的结果是动态的，会随着时间的推移而变化。因此，应根据本综述的截止日期2022年5月纳入更新的出版物，并建议在未来几年内复制本综述。

Taki Eddine Seghier: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing - original draft, Writing - review & editing. Chavanont Khosakitchalert: Data curation, Formal analysis, Investigation, Software, Validation,
Taki Eddine Seghier：概念化，数据管理，形式分析，资金获取，调查，方法论，项目管理，资源，软件，可视化，写作-原始草稿，写作-审查和编辑。Chavanont Khosakitchalert：数据管理，形式分析，调查，软件，验证，