*Article* **Social Housing Life Cycle Management: Workflow for the Enhancement of Digital Management Based on Building Information Modelling (BIM)**

**Manuel Castellano-Román 1, Antonio Garcia-Martinez 2,\* and María Luisa Pérez López <sup>3</sup>**


**Abstract:** The management of the life cycle of large publicly owned social housing complexes requires a large amount of human and technological resources, the optimization of which is a desirable and shared objective. This article proposes a workflow for the enhancement of these management processes based on BIM (Building Information Modelling), a methodology capable of integrating architectural information into a three-dimensional graphic model. The proposed workflow defines the basic characteristics of the BIM model oriented toward sustainable building management and its relationship with the key moments of its life cycle. It also analyzes the architectural information associated with the models and determines which parameters are optimal for their completion from the BIM models in terms of reliability, auditability, and automation. For this purpose, a case study has been developed for a multifamily residential building in Malaga (Spain), owned by the Andalusian Housing and Rehabilitation Agency AVRA, a public agency that manages a housing stock of more than 70,000 dwellings.

**Keywords:** social housing management; building information modelling (BIM); computerised maintenance management system (CMMS)

#### **1. Introduction**

The economic, social, and environmental sustainability of public and private buildings depends, to a large extent, on efficient and effective maintenance and conservation management during their useful life cycle. To this end, numerous IT tools have been designed in recent decades, from simple databases to specialized software known as Computerized Maintenance Management Systems (CMMS) [1]. These tools provide information on the key points of buildings managed by an organization, supporting the making of informed decisions regarding them [2].

To complement these digital resources for management, some authors have proposed the convenience of incorporating Building Information Modelling (BIM), making use of its capacity to integrate architectural information into a three-dimensional graphic model [3–5]. The advantages that stand out are the three-dimensional visual presentation of the model, its ability to generate effective communication between the agents involved (e.g., technicians, users, managers, and politicians), assistance in the control of maintenance (e.g., refurbishments, repair operations, and its costs,) and its usefulness as a bridge to other technological applications [6–15]. In parallel to the development of BIM, the digital maintenance and conservation phase has experienced significant progress [16–24].

**Citation:** Castellano-Román, M.; Garcia-Martinez, A.; Pérez López, M.L. Social Housing Life Cycle Management: Workflow for the Enhancement of Digital Management Based on Building Information Modelling (BIM). *Sustainability* **2022**, *14*, 7488. https://doi.org/10.3390/ su14127488

Academic Editor: Shervin Hashemi

Received: 2 January 2022 Accepted: 16 June 2022 Published: 20 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

In the specific context of the application of BIM to the management of social housing, references are scarcer. In Europe, a case study has been carried out in Triolo, France, which proposes a social housing management strategy based on the generation of a BIM model and the direct management of it [25]. It characterises the advantages of using BIM but has the limitations of lacking external management software and not addressing the issue of managing many buildings.

Another European study, located in the United Kingdom, provides experience of the use of BIM in the refurbishment process of a social housing building. This experience focuses on the time of the intervention and not so much on the subsequent management, highlighting the advantages of the fourth dimension of BIM, i.e., the time sequence in the models [26].

In Oceania, a New Zealand experience promoted by a public institution, the Wellington City Council, is of interest. This experience focuses on the benefits that BIM methodology can offer a public manager in predicting maintenance costs and planning management to extend the useful life of buildings. BIM models act as a data collector whose further management is in the process of development [27].

In the Americas, an experience developed in Chile has been published in which BIM is introduced to manage the moment when newly constructed social housing buildings are occupied by their inhabitants. It shares with the experience presented in this article attention to the phase of the useful life of the building but focuses on the characterization of damages in the building, recorded in the model itself and not in their management from external applications [28]. Finally, an experience from Brazil provides the perspective of client requirements in social housing projects using BIM, focusing on the need to adjust the information structure of the models to the objectives of the latter [29].

#### *1.1. Problem Statement*

Despite these advances, organizations that have to manage a large number of social housing estates find it very difficult to apply these technologies. The main reason for this is that the buildings they manage are, for the most part, buildings of a certain age, built before the widespread use of digital graphics and information tools. First, that implies major problems with alphanumeric document management, generally solved with basic database tools, and, to an even greater extent, problems of graphic information management, which is stored on paper, raster images from the digitization of paper plans, or, in the best of cases, on vector CAD support. Second, this alphanumeric and graphical information must be converted into an effective real estate decision-making assistant. To do so, the data must be reliable and auditable. Furthermore, if data correction and updating is necessary, the process should be as automated as possible.

This is the case for the Andalusian Housing and Rehabilitation Agency (AVRA), one of the public organizations with the largest number of dwellings managed, with 73,989 houses that provide accommodation to more than 300,000 people, on whose experience the case study will be based [30].

AVRA has wide and varied real state under its supervision, the management of which involves a major documentation problem, including graphics, which is often present in preventive conservation projects [31]. Generally, this documentation is on paper, the result of the predigital tradition of managing records through copies and reproducible means, typical of the last century. On other occasions, especially in the 1990s, paper deliveries were accompanied by copies on CD media of planimetry developed in CAD in a digital environment. The latter generally contain dwg and dxf files, whereby the information is systematized in folders, layers, and linked files, from which a printed dump can be obtained on paper or in PDF format. Despite this, dossiers have continued to appear in print well into the current century, either on paper or in PDF format, a format that was officially launched as an open standard in 2008 and published by the International Organization for Standardization (ISO) as ISO 32000-1.

In the General Archive of AVRA, this agency has only commissioned paper projects since its foundation in 2013. However, the public housing stock is made up of social housing that comes from various entities (ministries, councils, and municipal companies), most of which is from the second half of the last century, which are graphically recorded on paper and deposited in the Provincial Offices, not in the General Archive. Project copies on CD began to be ordered at the beginning of this century but have not been inventoried. This last medium usually presents the problem of difficult long-term conservation since it is highly vulnerable and contains information that is difficult to retrieve. In the case of the rehabilitation of historical buildings, or those built prior to the middle of the twentieth century, which represent a minority in relation to the total computation of the agency's social housing, there is a definite lack of updated documentation, although on rare occasions documentation is available in provincial or municipal historical archives.

Currently, AVRA uses software for the management of its social housing, named herein as CMMS-AVRA. It organizes a database that contains essential information on each managed building. The content and structure of this database are established in the Law 8/2013, of 26 June, on urban rehabilitation, regeneration, and renovation (BOE, 27 June 2013). This law describes the information required in a Building Assessment Report, a document standardized by the Spanish Government [32] that records the basic architectural characteristics and state of conservation of buildings. Based on this information, AVRA plans the rehabilitation actions to be carried out in each campaign.

#### *1.2. Objectives*

The main objective of this article is to show how the management of the digital lifecycle of public social housing parks can be improved with the use of the BIM methodology. To achieve this main objective, the following specific objectives are designed. The first is the proposal of a strategy to be managed for the graphic documentation of the buildings, capable of providing the necessary information to set up the BIM model.

The second specific objective is the analysis of the information structure of the building managers' own applications, in this case the CMSS-AVRA software, to select the key parameters whose determination can be improved from the BIM model. The aim is to select those that the BIM model can complete with guarantees of reliability, auditability, and automation.

The third specific objective is the characterization of the BIM model oriented to the sustainable management of its life cycle. This characterization will be determined in terms of a modelling strategy and linkage with the different situations in which the building may find itself during its life cycle.

The fourth specific objective is to resolve the connection of the BIM model with the building managers' own applications, in this case CMSS-AVRA software, enabling the flow of information between the two.

#### **2. Methods**

#### *2.1. Current Workflow Analysis*

Information on the current situation of the workflows carried out by AVRA in the management of its social housing was extracted from the performance of structured faceto-face and online interviews. These interviews were carried out by the research team in the period January-February 2021. The interviewees were the technicians responsible for the maintenance of the AVRA buildings. These interviews focused on the current common process that the agency is carrying out in the management of its social housing and the processes that could be optimized using the BIM methodology.

The information required for the analysis and evaluation of the improvement possibilities was obtained from unstructured interviews with various building managers (IT experts, architects, engineers, and managers). In these interviews, an attempt was made to determine the real needs of AVRA in relation to maintenance and the real possibilities of

implementing BIM-based conservation management, considering the knowledge of these resources held by the technical personnel in charge of building conservation management.

The final evaluation of this process is presented in Section 1.1 as a problem statement. Although a management software is in place, CMSS-AVRA, the management model is highly analogue. The process of data entry and maintenance of the management system is essentially manual, and updated graphic documentation is rarely available. By not benefitting from the automation provided by methodologies such as BIM, maintenance is costly and inefficient, according to the managers themselves.

#### *2.2. Case Study Selection*

To achieve the objectives, one of the multifamily residential buildings managed by AVRA was taken as a case study (Figure 1). This building, built in Malaga, is representative of the type of property that AVRA manages, and lies both in the field of BIM management of social housing and in the generation of information models of existing buildings.

**Figure 1.** Case study, building on Lemus Street nº5 (Malaga). General Directorate of Cadaster and authors.

This building was constructed in 1987 and is located between Lemus Street and Plaza Bravo, in the Trinidad/Perchel neighborhood of Malaga. Its floor plan is rectangular, with two-storey pendent blocks connected through the interior patio and the first-floor gallery, evoking the image of an old neighborhood corral that previously existed at that location. It has two façades, Lemus Street and Plaza Bravo, from which the building is accessed. All the social housing in the promotion remains under a social rental scheme.

#### *2.3. Digital Capture*

The generation of the BIM model is based on the 2D CAD survey of the building, supplied by AVRA. In order to verify the degree of approximation of the model with the physical reality of the property and to evaluate the margin of error that the utilization of previous standard documentation supposes, two data captures were also made using digital technology: (a) a static laser scanner, based on taking data through static scanning locations; and (b) a hand-held scanner that performs data collection dynamically, accompanying the movement of a person moving through the property [33]. The equipment used was, firstly, a LEICA BLK360 IMAGING LASER SCANNER, capable of capturing 360,000 points per second, with a range of 60 m and a precision of 4 mm; it also includes thermal, laser, and visible light images, and is able to create a 360◦ scan in just 3 min. Secondly, a ZEB-REVO hand-held scanner was used, with a maximum range of 30 m, 43,000 points per second, and a relative precision of 1 to 3 cm.

The work carried out with the LEICA BLK360 IMAGING LASER SCANNER consisted of collecting data through 21 scanning locations in the common spaces of the building, starting from the centre of the courtyard. The interior of the dwellings was avoided since it would have required a scanning location for each room and would therefore have considerably increased the time of field work.

Through the LEICA Data Manager program, the download of the 21 BLK360 files took 60 min and their importing through the LEICA Cyclone Register program took another 60 min, while the recording of the 21 positions for their unification and adjustment took 120 min, and their exporting in rcp format, occupying 18 Gb of information, took another 210 min. A total of 555 min was spent, during which a massive data collection of points was obtained, including the management of the chromatic information of each point. This allowed dynamic images of optimal resolution to be obtained that provide a virtual tour of the building.

The metric capture using the ZEB-REVO hand scanner was carried out in two successive data collections. The first collection included the interior of one of the houses, the entire gallery, stairs, and a roof terrace. In the second data collection, the exterior façades were scanned (Figure 2). The starting point for each data collection was the same, with a duration of each moving image of 15 min.

**Figure 2.** Raw point cloud obtained from the scan: sectioned axonometry of the point cloud.

The clerical work consisted of dumping the point clouds from the scanners into universal format files, with an e57 extension from the scanner rental company, and of sending them by email link, which took approximately 10 min. Our team then imported these files into Autodesk Recap, to convert them into an Autodesk rcp file, compatible with Autodesk AutoCAD and Revit programs (Figure 3), which took another 10 min.

**Figure 3.** Point cloud after importing into Autodesk Recap.

This point cloud is scaled and oriented vertically from the origin, hence it is only necessary to orient it horizontally and move it based on a known georeferenced point. In our case, this cloud was transferred to the model that had already been prepared from the existing planimetry to verify the possible readjustments to be applied from this digital capture (Figure 4); this took another 10 min. The entire data collection and information management process to make it available in CAD was 60 min.

**Figure 4.** Model in Autodesk Revit with the imported point cloud.

A structured methodological process was designed according to the following sequence: analysis of the content and structure of the information of the AVRA program, development of the BIM model, and design of the input and output algorithms between the BIM model and the AVRA program.

#### *2.4. Analysis of the Content and Structure of the Information in the Management Software*

Information from CMSS-AVRA was received as data in Excel format. The large number of information fields resulted in the output of four separate files, three of which contained general property data and a fourth specifically oriented to the construction injuries registered in the building. The structure of this information was analyzed to identify which parameters allow a link between the CMSS-AVRA and the BIM model and which of these allow its automation from the said model.

#### *2.5. BIM Model Generation*

The BIM model was developed with the Autodesk Revit program. For the foundation of the model, CAD planimetric documentation and two three-dimensional scans were employed. However, the generation of the BIM model was not based exclusively on the metric information that could be extracted from the CAD vector graphics or from the point clouds of the three-dimensional scan, but rather required a constructive interpretation of the building. This interpretation was based on information obtained from the CMSS-AVRA and from direct inspection of the building itself (Figure 5).

#### *2.6. Connection between Management Software and BIM Model*

Once the BIM model had been formed and the information fields to be linked between the AVRA application and the model defined, the algorithms that make the transfer between them possible were then programmed. To this end, the Dynamo application, integrated in Revit, was used. An input algorithm was programmed from Excel, obtained from the AVRA application, and another output algorithm from the BIM Revit model to Excel: a format from which the AVRA application could input the information.

#### **3. Results**

#### *3.1. Graphical Documentation Strategy*

The implementation of the BIM methodology for the management of large property assets of institutions such as AVRA requires suitable systems and technological resources to render processes viable and sustainable over time. One of the first problems when building these models is the lack of a proper initial graphic documentation since most of these buildings were designed in a predigital era.

The starting point for the development of a BIM model lies in knowledge of the geometry of the building [34–36]. This can be obtained from a metric capture made expressly for the purpose or through an existing previous graphic. Nevertheless, on many occasions this documentation may not exist, may not be found, or does not accurately reflect the current state of the property. Four starting points were established:


Given these situations, the development of a BIM model raises the choice between a metric capture made expressly for the purpose or the use of existing previous graphic documents. This is an important decision, since it can lead to an investment of considerable time and cost. The decisive factor is the purpose of the BIM model. If the building shows deformations and serious structural problems, or a major reformation is to be carried out thereon, then a comprehensive survey of its geometry would be necessary to develop a rigorous model in which to strictly quantify aspects that cannot be evaluated at first glance, such as overhangs. With the same starting point, if merely the daily preventive maintenance and registration of a building under normal conditions are envisaged, then not only can the model be much simpler, but the starting documentation can also be less metrically demanding. On the other hand, on many occasions, it is necessary to model the building in a simplified and sometimes repeated way for the development of computing applications for calculation, for example, for thermal efficiency.

Therefore, it is necessary to consider both questions, for which we start from Table 1, where the initial graphic information for the construction of the model is proposed according to its purpose.


**Table 1.** Relationship between geometric information support and model requirements.

(a) Analog or digital; (b) sketch, measurement with conventional techniques; (c) digital capture/laser scanner.

For this research, both options were developed. On the one hand, the existing 2D CAD survey has been used for the generation of the BIM model. On the other hand, an alternative digital capture has been performed using both a manual scanner and a static scanner.

The information obtained from the scan was converted to rcp format, thereby obtaining point clouds that were linked to the BIM model. The comparison between point clouds is one more layer of information that enables the deformations of the elaborated model to be visually verified, and, if necessary, to be modified to better approximate the actual state of the building. In this case, the deformations occurred on the floor plan, due to the adjustment in the execution of the construction to the actual site, but these minor adjustments only affected the minor width of the floor and were insufficiently important as to modify the model (Figure 6).

From this analysis it can be deduced that in those cases where there is not a 2D digital model (CAD type), the best option for the realization of the generic BIM model for the management of the property would be to have a quick data collection of the property through the handheld scanner, which we have verified to be fast and accurate enough in most cases to subsequently obtain a model that is geometrically adjusted to reality. This type of instrument also makes it possible to accompany routine records carried out by technicians to verify any incident, since the transport of the instruments is light and the information management processes are fast.

**Figure 6.** Cloud point and BIM model comparison.

Finally, the use of a ground scanner requires a greater investment in time and storage space, although its level of precision is much higher. Solutions to quickly record the data of various static captures already exist today, but typically produce a volume of information that can become unmanageable. For this reason, this type of metric capture is recommended for the analysis of structural problems, or the analysis of persistent pathologies, and should be avoided for problems that require a more generic definition of the model.

The use of the initial documentation requires, in turn, that it be located and linked to the database that contains not only all the records of the assets, but also the basic data regarding the surface area, location, etc. This provides a history to which future BIM models will be subsequently incorporated. The use of this entire graphic legacy has many other documentary functions that can be exploited in the long term, that concern the direct relationship between the forms of representation and the type of architecture developed [37].

#### *3.2. Information Structure Analysis*

As indicated previously, AVRA currently has software, CMMS-AVRA, for the management of its social housing. According to those responsible for this management, its effectiveness as a management program for the maintenance of its buildings has been sufficiently proven, at least in actions directed by the central services of the agency.

Consequently, the structure and content of the CMSS-AVRA database have been considered as the basis for the information structure to be included in the BIM model. For this, the information contained in the current CMMS-AVRA database was analyzed to determine how the use of BIM models could optimize the current action protocols. In the present investigation, the data related to the case study have been output from the CMSS-AVRA program in Excel format, since this facilitates the bidirectional transfer options with the BIM model. These data include the parameters of different formats and architectural significance, the relevance and usefulness of which in the model required their analysis and assessment. For this, two groups of data have been identified:

The first group includes the data that can be filled in indistinctly in the CMSS-AVRA or in the BIM model, that is, that in which none of the applications offers a significant advantage beyond the opportunity that some of them were already present. For example, the identification of the property will generally be completed in the AVRA program and could be output when a BIM model of one of them is generated. However, in the opposite direction, the foundation of a new BIM model on a property not yet registered in the AVRA application would allow the same identification data to be output (Table 2).


**Table 2.** Informative data that can be entered in both the CMSS-AVRA and the BIM model.

The second group includes the data that can be extracted directly from the BIM model. These data provide the greatest potential for improvement among the social housing management processes. These data are related to architectural characterization (constructive, structural, functional, etc.) and its quantification. Being architectural information, these data are from the BIM model and, therefore, will have priority over those entered into the CMSS-AVRA. In general, they are quantitative data directly calculated by the BIM software from the modelled elements. For example, the constructed area of the property is an item of data that results from the BIM model and, therefore, is a graphically verifiable item of data that can be automatically updated if an error is detected. However, qualitative data derived from quantifiable data can also be defined. For example, the dominant joinery material is an item of qualitative data determined from the joinery surfaces of each material in the building (Table 3).

**Table 3.** Extract of quantitative data inherent to the BIM model, available to be exported to CMSS-AVRA.


From what is revealed in the analysis of the content and the information structure of the CMSS-AVRA, it can be inferred that the proposed BIM model must not be a mirror reflection thereof, but rather that the optimization of the information is proposed in terms of reliability, auditability, and automation: reliability in terms of the minimization of manual data transcription errors; auditability since the graphic condition of the BIM database allows both the visual recognition of the origin of the data in the three-dimensional model and the confirmation of its veracity; and automation in relation to the direct update of the data after the eventual modifications of the elements of the model. These specific advantages of employing BIM in this case study complement other advantages that are inherent in its nature as a graphical database and that are also useful in this context: the three-dimensional characterization and the availability of coherent planimetric documentation (Figure 7).

**Figure 7.** Data auditability. The parameter CV\_FP\_DA\_SUPERFICIE filtered by color, which shows the area of the exterior façades.

#### *3.3. Generation and Characterization of the BIM Model*

The generation of the model involves both the geometric 3D modelling of the building and the design of the linked information structure, described in the previous sections. Geometric 3D modelling is based on a digital capture or on existing graphic documentation. In this case, both options have been explored to build the model, although the existing CAD survey has been sufficient to produce an operational model in accordance with AVRA's management requirements. Similarly, a linked information structure has been designed. The connection between the graphic elements and their information is solved thanks to the architectural analysis, which makes it possible to determine which information corresponds to each element of the model.

The resulting model is called BIM-AVRA, suitable for attending the management of a social housing manager. Its main characteristic is to be a 'live' model, i.e., a model that is systematically updated. To ensure the genuine effectiveness of this systematic update, the BIM-AVRA model must be as simple as necessary to ensure that the manager, in our case, AVRA, can bear the costs involved in the update process. In any case, the simplification must guarantee that the model includes the basic configuration of the building, but no elements that may be altered by third parties, whether tenants or other types of users. For example, the BIM-AVRA model will not include the color of an interior wall or the form of the bathroom cladding because these elements may be altered by third parties without the knowledge of the manager, thus rendering the model out of date (Figure 8).

**Figure 8.** Horizontal and vertical section 3D views of the BIM-AVRA model of the case study.

Therefore, this model differs from others that could be created for specific new-build or refurbishment interventions. In other words, a single model cannot resolve the entire complexity of the useful life of a building, although it is possible to successfully integrate different models generated during its useful life. The BIM-AVRA model can offer basic architectural information about any of the agency's buildings and, therefore, can become the base model for any other models that may be generated for specific events during the useful life of a building, such as its extension, refurbishment, or demolition. These models would therefore serve a very different purpose, with the specific aim of documenting a design, execution, or state following an execution. Keeping these models separate from the BIM-AVRA model mitigates the problems derived from outsourcing the design and execution of a building because the models developed for these purposes will be independent from the base model and may be as detailed as required by the specific intervention addressed. Situations such as nonexecuted designs, partial studies, and building analyses, or even information regarded as an inaccurate reflection of the state of the existing building but useful to keep, will be archived as "still photographs" of the intervention without compromising the rigorous updating of the BIM-AVRA model. When these intervention models become substantial alterations of the building, they will be incorporated into the updated model in a simplified format consistent with the existing data in the model (Figure 9).

**Figure 9.** Diagram of the life cycle workflow around the updated BIM model.

#### *3.4. Data Transfer and Interoperability BIM-AVRA/CMMS-AVRA*

The algorithms that allow the transfer between the BIM model and the AVRA program are programmed with Dynamo, an application integrated in Revit that allows Excel files to be used as a format for the input and output of the model information. Both processes are structured by 'categories', that is, the set of objects that within a BIM model are associated with a certain constructive function (walls, roofs, doors, windows, etc.) or characterization of the model (information, project, built areas, etc.).

In the case of input from Excel, the Dynamo algorithm starts reading the selected data to assign them to the corresponding categories in the model. The process in the opposite direction is symmetric, although one must stop at articulating the data lists in a way that is consistent with the sequence expected by the AVRA management program (Figures 10 and 11).

**Figure 10.** Dynamo algorithms for data output from Excel to the BIM-AVRA model (IMPORT.dyn).

**Figure 11.** Dynamo algorithms for data output from the BIM-AVRA model to Excel (EXPORT.dyn).

#### *3.5. Summary of the Proposed Workflow*

The proposed workflow is organised around two main cores: the CMMS software of the managing institution, in our case CMMS-AVRA, and the updated BIM model of the building, in our case BIM-AVRA (Figure 9). The CMMS is a single core, and the BIM model is a multiple core, so there will be as many updated BIM models as there are buildings in the building stock to be managed. This does not mean that the workflow is only operational when all updated BIM models have been generated, but it can be combined with the current workflow. CMMS management can remain operational with the current data input system, which, as described in the Introduction, is mainly manual and, as updated BIM models of the buildings are generated, the system becomes more reliable, auditable, and automated.

Once the action on a building has been decided, the process of generating the updated BIM model begins. For this, existing graphic documentation or, where appropriate, digital capture will be used, in accordance with the strategies described in the previous points. Regardless of the sources of graphic documentation for the generation of the model, it is essential to perform an architectural analysis to basically characterise the structural system, the construction system, and the MEP system. This analysis allows the level of development of the model to be adapted to the knowledge available and to the resources available to the manager to keep it updated.

During the life cycle of the building, a number of unique events will occur, depending on the different types of intervention that the building manager may decide to carry out on the building. The updated BIM model provides reliable information for the intervention project and can serve as a basis for the generation of a BIM model specifically oriented to the objectives of that intervention. In other words, a BIM model of the corresponding event will be produced based on the updated BIM model, but different from it. Once the event is completed, the team responsible for maintaining the updated BIM model will revert the changes to the BIM model, according to its own standards and level of development. The BIM model used to produce the intervention will be archived and document the event at that point in the building life cycle.

At any time, the bidirectional communication of the updated BIM model with the CMMS can be updated via the xls export and import process described in the previous sections.

#### **4. Discussion and Conclusions**

This research presents a workflow to improve social housing life cycle management using BIM. To this end, the AVRA case study has been considered, given its role as a public owner that manages more than 75,000 dwellings. Based on unstructured interviews with those responsible for AVRA's property management, the following problems were stated: alphanumeric and graphic documentation and its use as an effective decision-making assistant for real estate.

The first step for BIM-based social housing management is the definition of a strategy for the graphic documentation of the buildings to be managed. The generation of the BIM model can be approached: (a) from 2D digitized plans (dwg, dxf or similar); (b) from plans on hardcopy; and/or (c) from the building itself by means of a digital and/or photogrammetric capture. Although digital capture offers advantages such as speed and accuracy, it is also a high-cost investment when it comes to hundreds of surveys. In the case study, two options were tested: the generation of the model from 2D CAD plans and a digital capture. It was concluded that the key point in making a choice is the purpose of the model, and a table was presented with the different purposes and recommended resources. For the day-to-day management of such a number of buildings as AVRA manages, the generation of the BIM model from 2D CAD drawings was proven to meet the proposed objectives.

On the other hand, digital capture by scanner was shown to be an effective and efficient procedure for generating BIM models. Two alternatives were studied for the metric capture of buildings: I. using a manual scanner; and II. a static scanner. It has been verified that, whatever the case may be, the information taken is compatible and can be related to the final BIM model, whether it is formed based on the digital capture carried out or not. The use of a manual scanner presents the advantages of transportable equipment, speed, simplicity, and efficiency of the data capture procedure compared to the static scanner, which requires more time for preparation and data collection. However, the static scanner presents greater precision. Given these results, it can be concluded that the point cloud obtained with the handheld scanner offers suitable precision for the realization of a general building management model. The point cloud obtained by means of the static scanner, more precise than the previous scanner, is useful in buildings that require intervention from a constructive and/or structural point of view.

The key parameters of the BIM model were determined to enhance the management carried out together with the CMSS-AVRA. This means that the information structure of the model is not a mirror of the CMSS-AVRA information, but the definition of two groups of data: those that are simply informative and those that quantify the architectural characteristics of the building. Informative data can be added to both software, but the source of quantitative architectural data must be the BIM model, as this ensures reliability, auditability, and automation.

The characterization of the BIM model oriented towards the sustainable management of the building's life cycle has been determined. In the modelling strategy, the requirement for simplicity prevails, modulated by the manager's capacity to carry out up-to-date maintenance that guarantees its long-term operability. Although simplified, the model must incorporate the basic configuration of the building, leaving out everything that can be altered by third parties. Consequently, it is ruled out that a single model can assume the full complexity of the building's life cycle, although it is possible to effectively integrate different models generated throughout the building's life cycle. A workflow in which interaction with the CMSS-AVRA is always carried out from this updated BIM-AVRA model is proposed. This updated BIM-AVRA model is offered as a basis for the development of other specific models for different events in the life of the building. In turn, the BIM models for each event will be archived and only the information relevant for the day-to-day maintenance of the building will be transferred to the BIM-AVRA model.

The connection between CMSS-AVRA and the BIM model has been solved, allowing the flow of information between both. A Dynamo script was programmed to automatically transfer the selected CMSS-AVRA data to the BIM model (IMPORT.dyn) via an Excel file. In turn, a Dynamo script was also programmed to automatically transfer the data contained in the BIM model to CMSS-AVRA via an Excel file (EXPORT.dyn). The latter, obtained directly from the BIM model, improves the management workflow, as they are reliable, auditable, and automated.

Finally, AVRA technicians have been shown the joint use of the CMSS-AVRA and the BIM-AVRA model of the case study, expressing their interest in the future developments of the results of this research. A first line of development would be oriented towards the extension of this experience to a sufficiently significant number of models to produce strategic decision-making assistance programmes. This would require selecting the key parameters of BIM models and combining them with others specifically designed to represent the strategic criteria of managers. A second line of development would be oriented towards the characterization of the modelling of singular buildings. In this sense, it would be particularly interesting to define specific strategies for the modelling of heritage architecture, incorporating artistic, historical, and archaeological information. Heritage applications of BIM, known as HBIM, have been widely tested in recent years, but it would be very interesting to reconcile their contributions with the management requirements of social housing use.

**Author Contributions:** Conceptualization, A.G.-M. and M.C.-R.; methodology, A.G.-M. and M.C.-R.; BIM model and Dynamo scripts, M.C.-R.; validation, M.C.-R., A.G.-M., M.L.P.L.; formal analysis, M.C.-R. and A.G.-M.; investigation, M.C.-R. and A.G.-M.; data curation, M.C.-R.; writing—original draft preparation, M.C.-R. and A.G.-M.; writing—review and editing, A.G.-M. and M.C.-R.; visualization, M.C.-R.; supervision, A.G.-M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research has been financed by European Regional Development Funds and the Junta de Andalucía, through the CEI-10-HUM799 project, within the framework of singular projects of transfer actions in the Campus of Excellence in the areas of the Research Strategy and Innovation (RIS3), within the global project of the University of Seville Innovative Ecosystem with Artificial Intelligence for Andalusia 2025.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors thank the Andalusian Housing and Rehabilitation Agency AVRA, the Ministry of Development, Infrastructures and Spatial Planning, and the Junta de Andalucía for providing all the information necessary for the development of the case study proposed in this investigation. We are especially grateful for the work carried out by Isabel Ramírez Acedo and Miguel Ángel Lobato Aguirre; the technical area members, Elena Morón, Mª María Bermejo Oroz, Jorge Ruiz García; the AVRA work team of the delegation of Málaga, Miguel Ángel Santos Amaya (Head of the Development and Quality Control Service, Ministry of Development, Infrastructure and Land Management—SGT); and Francisco Javier Ramos from the company Geoavance SL for his availability and kindness in accompanying and supporting the data collection. We also appreciate the contribution of the research project "Development of BIM-HBIM-GIS models applied to the knowledge, conservation, dissemination and management of architectural cultural heritage", especially the kind support of its coordinator Francisco Pinto Puerto and the researcher José María Guerrero Vega.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


### *Article* **BIM and Automation in Complex Building Assessment**

**Jan R ˚užiˇcka 1,2,\*, Jakub Veselka 1,2, Zden ˇek Rudovský 3,4,5, Stanislav Vitásek <sup>5</sup> and Petr Hájek <sup>2</sup>**


**Abstract:** When using Building Information Modeling (BIM) for complex building design, optimizing the building quality in a design phase becomes an important part of integrated and advanced building design. The use of data from an information model in the design phase allows efficient assessment of different design strategies and structural variants and a higher quality of the final design. This paper aims to analyze and verify possible BIM data-driven workflows for Complex Building Quality Assessment (CBQA) and a suitable BIM data structure set up for automatic assessment and evaluation. For an efficient automation process in complex quality building assessment in the design phase, it is necessary first to understand the data structure of the Industry Foundation Classes (IFC), which is widely accepted and used for buildings, and second to understand the data structure of the assessment methodology used for the assessment. This article describes possible data workflows for an automatic assessment based on the experience gained on a case study of the real pilot project of a residential building, where the complex building quality was tested using SBToolCZ, the Czech national assessment method. This article presents the experience and recommendations for setting up the data model of a building for automatic assessment.

**Keywords:** Building Information Modeling (BIM); Life-Cycle Assessment (LCA); building optimization; Industry Founded Classes (IFC); SBToolCZ

#### **1. Introduction**

The complex quality of buildings is one of the main issues of the current building sector, which covers a wide range of criteria, including environmental quality and LCA, energy performance of buildings, durability, and resilience, quality of the internal environment, architectural and functional quality, and other social issues. Current rating systems and assessment methods and frameworks are efficient, inspiring, and motivating tools in the design process to achieve the goal of complex building quality. The complexity of the rating systems, on the one hand, and a huge number of criteria and parameters required for the assessment, on the other hand, are a challenge when searching for new ways for automation in this field. Connecting Building Information Modeling (BIM) as a structured database of building parameters with sufficient data workflows linked to the assessment scheme can provide an efficient tool for complex building quality assessment.

BIM is still a new idea in construction, representing significant technological progress in information technology that has been visible in recent years. It combines a digital environment with a real one and has great potential to change traditional ways of working in the design, construction, and management of construction projects. The BIM method has several definitions from leading construction experts and building associations. The most

**Citation:** R ˚užiˇcka, J.; Veselka, J.; Rudovský, Z.; Vitásek, S.; Hájek, P. BIM and Automation in Complex Building Assessment. *Sustainability* **2022**, *14*, 2237. https://doi.org/ 10.3390/su14042237

Academic Editor: Ali Bahadori-Jahromi

Received: 31 December 2021 Accepted: 11 February 2022 Published: 16 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

frequently inflected interpretation gives the authors of the article a definition: "BIM technology represents a digitally created one or more precise virtual building models. Supporting object design across all phases of the object life cycle, enabling better analysis and control than manual (traditional) processes. When these computer-generated models are ready, they contain the exact geometry and data needed for the construction, production, and procurement process according to which the building is carried out, including subsequent management" [1]. In general, all interpretations associated with the definition of the BIM method have common features based on the following:


However, according to the perspective of the global consulting company McKinsey & Company and its Industry Digitization Index, the Architecture, Engineering, and Construction (AEC) field is still one of the least digitalized fields of human activities [2,3]. Other recognized publications express in a similar vein the views that the construction industry is among the least digitized industries, and the lack of innovation in construction project management practices has led to decreased productivity [4]. This includes the point of view of public contracting authorities (public sector), who have modern technologies required for more efficient spending of public funds, where selected articles recommend starting to "push" BIM in this way and thus start market development [5–7]. It also includes the point of view of private investors, where the "correct" grasp of these modern processes carries the private companies themselves into the new digital era, and therefore, there are desired savings of time and, especially, costs [8–10]. In the overview of professional publications and studies, the authors in the following subchapters focused on topics that are closely related to the subject of the article, where their conclusions and messages were considered in addressing the issue (in addition to obvious benefits and pitfalls of the digital implementation process).

#### *1.1. Sustainability of Construction*

The sustainability of construction, together with its digitization, is one of the most current topics, both in the field of research and in construction practice. Recent studies based on more than 650 buildings produced by several research groups [4,11–13] show that in the following decades, the importance of embodied greenhouse gas (GHG) emissions is greater than operational emissions. BIM can be involved in many particular tasks within design, construction, and operation within this issue. Relevant Asian projects linking BIM and sustainability were in Malaysia, China, and Kazakhstan. In all cases, it was a strategic decision to place a BIM method to determine the degree of sustainability in the solution of a project identified as sustainable, including the subsequent use of data for its management. The conclusions of the articles are related to balancing the advantages and disadvantages of digital technologies, where the benefits of these modern techniques clearly prevail [14–16]. Another important article on this issue is research on the Korean environment, which focuses on the impact of the BIM method and sustainability on highrise buildings. The detection process took place through Exploratory Factor Analysis (EFA) and Structural Equation Modeling (SEM). The result of the research was the identification of Critical Success Factors (CSFs) for the implementation of BIM on sustainable and high-rise construction, where 205 construction projects participated in the questionnaire survey. Key criteria identified by this research were included in a modified form in the solution of their article [17].

The authors of practical studies have also developed them in the energy sector. Specifically, these are two articles focused on applying BIM in the field of energy with a direct link to sustainable topics. Both articles deal with identifying data needs and subsequent analyses for the energy and sustainable model of building in the European environment. The primary goal was to automatically generate an accurate and flexible model called Building Energy Modeling (BEM). BEM made it possible to automate the calculations of primary energy savings and CO2 emissions for the solved projects and thus met the pilot objectives of the project [18,19]. However, from the point of view of the authors of their own article, some visions were too progressive and may have a practical impact on routine operation in the next few years.

In general, the topic of building sustainability has seen significant progress in recent years, both in research and in practice, where several digital tools have been developed to allow calculations based on input parameters. It is these input parameters that BIM can provide if entered correctly. In creating their own article, the authors used the applied methods in selected publications, primarily in connection with the method of defining data needs for similarly large projects intended for sustainable construction.

#### *1.2. Building Life Cycle*

Following the current text, the BIM method has the greatest use throughout the life cycle. That is, from the construction design through its implementation and subsequent administration. There are a number of examples that confirm BIM as a potential life-cycle method [9,10]. Within the design of constructions, there are several dozen successful case studies, and not only from the European environment. However, interesting information seems to focus on the individual "less-feasible" steps in the project design. In particular, it is the implementation of systems based on the Common Data Environment (CDE), where all participants in the construction design must learn to work with and use modern tools for their routine activities. Alternatively, additional information associated with coordination activities between several construction objects is available if more than one design program is used and "meets" in one data environment (primarily a matter of line constructions) [20–23]. These findings are valuable for considering how to realize the implementation process of the BIM method, not only in the stage of construction design.

The next stage of the construction life cycle associated with monitoring the progress on a construction site during the construction phase is crucial. Construction monitoring using the BIM method enables digital tools primarily associated with construction costs and schedules. An interesting observation in this regard is the use of Field Data Capture Technologies (FDCT) and Communication and Collaborative Technologies (CT), where colleagues applied digitization in a real construction project from the Spanish environment. Other articles also present other modern technologies digitizing processes in construction, such as Dynamic Site Layout Planning (DSLP) and Development and Framework for Integrating Unmanned Aerial Vehicles (UAVs) [24–27]. The conclusions of these articles and studies have common denominators in the form of three pieces of information. The first is that setting up a project for these digital processes should ideally follow the previous design stage. Furthermore, "data discipline" is required, where the data in all systems must be constantly updated, and, last but not least, the key is the support of the construction management, which determines the direction of the construction process.

By far, the longest stage in the construction life cycle is its management. There are several professional publications dealing with the benefits of involving the BIM method in the operation of both buildings and line objects. Leading publications present these benefits in specific projects. In particular, they deal with the integration of data from BIM models into various Computer-Aided Facilities Management (CAFM) systems with an emphasis on the greatest possible automation of information transfer. In a real operation, this information has a key role in the maintenance and servicing of individual machines/structures, including their location, etc. [28–30]. The result of this "lifelong" data interconnection is a Digital Twin, which represents the target state of the digitization of the building environment.

#### *1.3. Life-Cycle Assessment (LCA) and Complex Building Quality Assessment (CBQA)*

Another important methodology recently highly used in the AEC sector is Life-Cycle Assessment (LCA). This method can evaluate the negative environmental impacts of the building and help optimize it. The LCA method is described in the ISO 14044:2006 norm [31] and can be adjusted according to the national standards, which can vary by 20% [32,33].

BIM and LCA can be combined in various use cases, as shown in recent articles [11,12,34,35]. The purpose of all studies is to develop a workflow that can be (i) precise enough for the building design process, (ii) possible for building optimization (in terms of structure, materials, environmental impacts, etc.), and (iii) replicable for another project with minimum effort. Wastiels [36] wrote an overview of possible different workflows [37], and Obrecht et al. processed a systematic literature review [13].

The whole LCA method is very complex and not perfectly suitable for everyday practice. It also only covers the environmental part of the building quality. Therefore, certification tools and methods for Complex Building Quality Assessment (CBQA), such as the Building Research Establishment Environmental Assessment Method (BREEAM), Leadership in Energy and Environmental Design (LEED), Deutsche Gesellschaft für Nachhaltiges Bauen (DGNB), and the SBTool were developed [38]. These tools simplify the LCA method, and moreover, they evaluate other aspects of the building quality: social, economic, and local. As a result, complex quality assessments can be processed [36,39]. In the reality of the Czech AEC market, the mentioned complex quality assessment methodologies are commonly used. In addition, the Czech national methodology SBToolCZ [40] was developed for the Czech market by the Czech Technical University in Prague together with The Technical and Test Institute for Construction Prague and the Building Research Institute—Certification Company. This assessment framework is based on the international SBTool and adjusted for national economic, social, and natural conditions. It is accessible for free for the following types of buildings: family houses, apartment houses, offices, and schools. The current target is to transfer the manual assessment process to the BIM-based workflow. The study's objective is to identify possible data workflows for combining BIM and SBToolCZ as an example of a complex quality rating system.

#### *1.4. Industry Founded Classes*

BIM can be used in any of the use cases. However, with higher demands of automation (e.g., analysis, simulations, model data mining, dashboards, e-building permit, etc.), the importance of standardization grows. At a certain level, it is already implemented in the mentioned norm ISO 16739-1:2018 [41]. The standardized data structure can be (i) included in the primary model developed in the BIM authoring tool or (ii) exported from the native format to the Industry Founded Classes (IFC). This open international format is developed by the Building Smart organization.

The IFC format represents a standard that the construction market must fully implement and adopt. Several publications deal with this standard and emphasize its purpose in achieving the most open construction process possible [42–44]. This activity is called OpenBIM. The conclusions of the publications also call for the continued development of this format, including related key support software. IFC can also be used during the building operation period [45].

As Obrecht et al. [13] state, there are many examples and case studies integrating BIM and LCA, and different workflows for combining BIM and LCA proposed by Wastiels [36] can be used. LCA represents a simplified assessment process because of the limited number of required parameters and set of properties compared to complex quality building assessment and its complexity. Integrating BIM and CBQA represents new challenges and raises new research questions. This article aims to share the experience gained with semi-automatic BIM data-driven workflows and a suitable BIM data structure set up for complex building quality assessment (CBQA) on a pilot case study of a residential building combining BIM and SBToolCZ.

#### **2. Methodology**

The main problem of an automatic complex quality building assessment is that current methodologies and assessment schemes do not follow the data structure of a Building Information Model. Due to a different data structure, it is impossible to use automatic algorithms.

The problem, on the one hand, is a completely different set-up of the assessment schemes, which use specific quality rating scales, often based on verbal or visual rating scales. The other issue is the Level Of Information (LOI) of the assessed Building Information Model (BIM). For an efficient automation process in complex quality assessment of the building in the design phase, it is necessary to first understand the data structure of the IFC, and second, to understand the data structure of the assessment methodology used for the specific assessment.

#### *2.1. Data Format of a Building Information Model (BIM)*

The Building Information Model (BIM) is a database of graphical and nongraphical data stored as (i) an open and vendor-neutral data file format or (ii) a proprietary data file format. Using an open and vendor-neutral data file format is crucial for public procurement projects, as well as for wide collaboration between different branches of the construction sector and between different stakeholders in the design and construction process. Thus far, the only widely accepted representative of data schemas in the construction industry is IFC (Industry Foundation Classes).

#### *2.2. IFC Data Structure*

IFC is a widely accepted, used, well-structured, standardized, open, and vendorneutral data structure (or schema) of a built environment, including buildings and civil infrastructure. The IFC is maintained and developed by BuildingSMART [46], the nonprofit organization, and it is an official standard ISO 16739-1:2018 [41]. The most important schemas of the IFC data format involve:


The IFC data can be physically encoded in various formats. The most common is the STEP format (.ifc). An example of the data structure is shown in Figure 1.

#### 2.2.1. Semantics and Classification of Data Objects

The object definition in IFC is anything perceivable or conceivable that has a distinct existence, albeit not material. It is a broad term. Therefore, there is a logical classification of all objects in IFC. All of the classified objects are rooted entities; they are rooted in a hierarchical classification tree. This classification integrates entities in a logical and characteristic inheritance manner. The subordinate (lower branch) entity inherits all characteristics of the superordinate (higher branch) entity and adds some of its own characteristics.

The detailed explanation of the IFC structure that is relevant for data modeling for complex quality building assessment is as follows:

	- - Subordinate entity: IfcProduct: The IfcProduct is an abstract representation of any object that relates to a geometric or spatial context. The IfcProduct consists of annotations (cotes, comments, etc.), structural items (actions and reactions), spatial elements (site, building, story, space, spatial zone), Architectural, Engineering, and Construction elements (AEC products), etc.
		- Subordinate entity: IfcSpatialElement is the generalization of all spatial elements that might be used to define a spatial structure. It contains spatial zone (IfcSpatialZone), space (IfcSpace), story (IfcBuildingStorey), building (IfcBuilding), and site (IfcSite).

 Subordinate entity: IfcElement is a generalization of all components that make up an AEC product. It is IfcProduct with the addition of material nature.

IFC objects are specific modeled products in the building information model, such as columns, slabs, boilers, parking places, and zones for disabled users, which carry information about the properties, quantity, and quality that can be used for complex quality building assessment.

**Figure 1.** An example of the IFC structure [46].

#### 2.2.2. Properties

Properties are, in general, described in IFC in two ways:


Some alphanumerical values are enumerated.

As mentioned above, rooted entities are classified in the inheritance tree. This also applies to the inheritance of properties of rooted entities.

Example (Property set—property):

• Superordinate entity:

IfcElement: Has property set describing warranty (Pset\_Warranty), that consists of 8 properties.

• Subordinate entity:

IfcActuator: Has a Pset\_Warranty and, in addition, many other property sets: e.g., for electrical device common properties.

#### 2.2.3. Relationship

Relationships among entities in IFC are realized by objectified relationships. There are two different types of relationships: one-to-one relationships, and one-to-many relationships. Objectified relationships express:


Objectified relationships are inverse attributes of the modeled object.

As mentioned above, rooted entities are classified in the inheritance tree. This also applies to the inheritance of rooted entity relationships. For example:

• Superordinate entity:

IfcSpatialStructureElement: The spatial decomposition is a logical structure of spatial elements ultimately assigned to the project. The order of spatial structure elements being included in the concept for building projects are, from high to low level, IfcProject, IfcSite, IfcBuilding, IfcBuildingStorey, and IfcSpace.

• Subordinate entity:

IfcSpace: Space is part of spatial decomposition and, in addition, has a relationship to covering elements. Those coverings may represent different flooring or tiling areas.

#### *2.3. BIM Data-Driven Workflows*

The aim of data mining from the information model is to analyze the building and optimize its design. There are several possibilities of BIM data-driven workflows depending on the level of digitalization in the building sector, on the development of national BIM data standards, and assessment methods. The goal is to achieve a fully automatic assessment workflow that allows real-time design optimization. The degree of automation highly depends on data standardization, on the level of integration of assessment methodologies in the BIM software, etc. The levels of model data structure and data-driven workflows are: (i) low structured model data, and manual workflow, (ii) utilized IFC data structure and semi-automatic workflow, and (iii) highly structured model data and automatic workflow.

#### 2.3.1. Low Structured Model Data and Manual Workflow

In this case, model data are manually read, and parallel calculations and/or analysis are carried out in an independent calculation software (Figure 2). The results are manually inserted as model properties of a particularly added property set. The assessment information is not connected to modeled entities. BIM can exist without modeled entities, used just as a data format. This approach is inconvenient; design optimization is time-consuming and generates a high risk of calculating errors and errors in communication processes between the author of the building information model and authors of the specific calculations.

#### 2.3.2. Utilized IFC Data Structure and Semi-Automatic Workflow

A semi-automatic workflow (Figure 3) uses calculations and analysis outside the data model in specific calculation software and tools, but the information needed for the assessment is exported in a predefined structure and can be automatically imported into the calculation software. Model data can be exported straight from the native model to the calculation tool in case the calculation software can read the data directly. This allows keeping the original data quality from the model. The model data from the native model can be exported in IFC format and then imported through the utilized database (e.g., .xls, .csv) or directly to the calculation software. Particular property or property sets and/or entities that are not common parts of the data model can be manually added in the calculation software. The results of the calculation and assessment can be uploaded back to the native model automatically in case of nongraphical information using sufficient data bridge or manually in case of graphical information. In the case of semi-automatic workflow, the time-consuming and manual part of the assessment process still remains, but communication processes and data exchange are efficient, and the risks of errors decrease.

**Figure 2.** Low structured model data with manual workflow.

**Figure 3.** Utilized IFC data structure and semi-automatic workflow.

2.3.3. Highly Structured Model Data and Fully Automatic Workflow

The fully automated workflow (Figure 4) uses calculations and analysis as an integral part of the software for data modeling (as a plug-in calculation module) or as an independent calculation software and uses model data for calculations and analysis. External data input, if needed, is inserted automatically and directly to the model or as boundary conditions for the calculations. The calculation algorithms are prepared in the software (as an integral part or as a third-party plug-in), and they are calculated or analyzed in the model as well. As in the previous case, the results of the calculation and assessment can be uploaded back to the native model automatically in case of nongraphical information using sufficient data bridge or manually in case of graphical information.

**Figure 4.** Highly structured model data and fully automatic workflow.

In this case, a high-level model data structure is needed. All modeled data are general and usable for other purposes, e.g., housing unit area quantities, housing unit quantities, etc. Assessment calculations are generated automatically. The main advantage is that in case of any design change, information take-off, calculation, and assessment are automatically generated in real-time with minimized risk of errors and misunderstandings, allowing efficient optimization of the designed building.

This approach requires powerful HW, on the one hand, and continues software maintenance and control according to changing standards, legislation, software version, etc.

#### *2.4. Data Structure of SBToolCZ*

#### 2.4.1. SBToolCZ

To set up the architecture of the data model of the building, it is necessary to understand the data structure of the assessment methodology. This section describes the data structure of the Czech national assessment methodology SBToolCZ for sustainable and complex building quality assessment. The data structure of SBToolCZ is, in general, very similar to other assessment methods, such as LEED and BREEAM. Therefore, the approach of using data structures analysis can also be used for other assessment methods [37]. Data analysis was carried out on the assessment version for residential buildings.

SBToolCZ [40] is a national sustainability rating system developed for the Czech Republic that respects local building tradition, the social and economic environment, and its specifics. SBToolCZ is based on a generic framework SBTool (Sustainable Building Tool) for evaluating the sustainable performance of buildings, and projects developed by iiSBE (International Initiative for a Sustainable Built Environment) have been available on the market since 2010.

Currently, four certification schemes for new buildings in the design phase have been developed according to building purposes: office buildings, residential buildings, family houses, and schools. The assessment framework for kindergartens is in progress, along with adjustments of current schemes for other project phases such as the operation (in use) phase and refurbishment (renovation).

Criteria for assessing the built environment are divided into groups: environmental, economic, and management criteria, plus the location of the building, including social criteria. Each criterion has an algorithm for setting up the indicator value, i.e., they are weighted according to importance. Criteria also differ according to the phase of the life cycle of a building (design, start of operation, operation).

2.4.2. Data Structure of the SBToolCZ Assessment Scheme for Residential Building

Detailed data analysis of the assessment tool SBToolCZ showed that more than 500 parameters are required as inputs for the complex quality and sustainability assessment of the residential building. The data sources differ according to the character of the information. The following data sources for the required information have been recognized:


The BIM-based data model is a relevant source for almost half of the building parameters required for complex quality and sustainability assessment using SBToolCZ. The data structure of the model consists of the following data types:

	- - Areas and surfaces—used as general project information, e.g., total floor space, total gross floor space, total heating volume, etc., or specifically for used methodology, e.g., green roofs area, green facades area, area for rainwater collection, external relaxing areas as balconies, terraces, etc.,
	- - Lengths and dimensions—used as general project information, e.g., clear height or specifically for used methodology, e.g., dimensions of communication corridors for disabled users, etc.

Properties of specific real or virtual objects of a distinct value predefined by specific enumeration types according to the assessment methodology, such as bicycle parking place being (i) non-guarded and non-covered, (ii) non-guarded and covered, (iii) covered and guarded, and (iv) covered and guarded by a central security system.

#### *2.5. Connecting BIM and Complex Building Quality Assessment Method (CBQA)* 2.5.1. Specifics of Complex Building Quality Assessment Methods

As shown in the previous chapter, the complexity of complex quality building assessment methods is compared to other calculations in the design phase much higher mainly because of the variety and diversity of required data. As mentioned in Section 2.4.2, in the case of SBToolCZ, only about 40% of required data can be exported from a BIM model.

About 60% of the information needed for the CBQA is still included in other sources, such as GIS, documents and technical reports, process description, or other external sources, and has to be mined manually.

A common BIM model corresponding to current BIM standards was used for the case study. The data structure of the model enables common data processing and analysis based on direct or derived geometric properties (areas and surfaces, lengths and dimensions, volumes).

The BIM model structure did not cover many parameters and data linked to the existence of a specific real or virtual item in the model predefined by the assessment methodology and number of those specific items, such as the number of parking places with charging stations, number of bicycle parking places, etc., and linked to properties of specific real or virtual objects of a distinct value predefined by specific enumeration types, e.g., non-guarded and non-covered, non-guarded and covered, covered and guarded, or covered and guarded bicycle parking place.

Some of those items were added to the model to use automatic data export, but still, some are missing, as the modeling was complicated and time-consuming compared to manual reading.

The fully automatic process for CBQA, under those circumstances, is currently not achievable, and the semi-automatic approach combining automatic data export with manual model reading shows a suitable workflow.

#### 2.5.2. BIM Data Workflow for Complex Quality Building Assessment

The sustainability evaluation of a building or building project is a very complicated and time-consuming process and requires detailed information on building technologies, materials, energy systems, and processes and deep knowledge of the assessment methodology used for the assessment. Automation and the use of BIM in this process help to optimize the evaluation process and can improve the quality of the building project.

Considering current software possibilities and the complexity of quality assessment methods, the semi-automatic workflow described in the previous chapter was chosen as the appropriate strategy to combine BIM and complex quality assessment. Calculations and analysis for the complex building quality assessment are carried out outside the model in specific calculation tools, and the information needed for the assessment is exported in a predefined structure and can be automatically imported into the assessment software. A particular property or property set or entities that are not a common part of the data model are added manually in the calculation software.

2.5.3. BIM Data Modeling, and Use for Assessment Methodology

In the BIM-based data model, these parameters, such as lengths, dimensions, surfaces, areas, volumes, are gained from:


The assessment methodologies generally use:


The crucial part of the process enabling automation in the assessment process is the BIM model set up and its architecture according to the required information. The data structure of the assessment methodology is described above.

However, many different software applications enabling BIM are used in design practice; it is recommended to prepare the BIM model architecture based on IFC as an open and vendor-neutral data file format that allows cooperation, collaboration, and data use across different fields and specialization and using different software and tools for supporting analysis. The required data according to the data structure can be modeled by using the following IFC modeling tool:

	- - Areas—direct (or derived) attributes of IfcSpacialElements, such as spatial zone (IfcSpatialZone), space (IfcSpace), story (IfcBuildingStorey), building (IfcBuilding), and site (IfcSite).
	- - Surfaces—direct (or derived) attributes of IfcElement as a generalization of all real or virtual components that make up an AEC product.
	- - Lengths and dimensions—direct or derived attributes of IfcElement as a generalization of all real or virtual components that make up an AEC product or their annotations.
	- - Volumes—direct (or derived) attributes of IfcProducts (IfcElement and/or IfcSpacialElements) or/and real or virtual components that make up an AEC product.

#### **3. Case Study—BIM for Complex Quality Building Assessment of a Residential Building**

As a reference case study, the experimental residential building was used. The structural system (columns, shell, and horizontal slabs) consists of a high-performance reinforced concrete structure and is supplemented by timber substructures (façade, internal walls, and partitions). This system is mainly suitable for residential buildings, especially apartment houses, and it was developed in the project funded by the Technology Agency of the Czech Republic [47]. As a showcase, part of the apartment house was built in the experimental field of the University Centre for Energy Efficient Buildings (UCEEB) of the Czech Technical University in Prague (CTU) in 2019–2020. The model of the entire building and the real display are shown in Figure 5. The building itself consists of three independent apartments. Therefore, the whole project was fully under control within the whole design, construction, and operation process. The main goal of this project was described in conference

papers [48]. A building information model of the TiCo project was used. The used model was prepared in Level of Development LOD350 [49].

**Figure 5.** Model and real building of the TiCo project.

#### *3.1. Geometric Data*

For the purpose of this study, one information requirement from the social part of the SBToolCZ methodology is assessed: User comfort—the availability of a bike or stroller parking space. It is completed according to two tables (Tables 1 and 2). Each table assesses the stroller stands in the building from different perspectives. First, the table assesses the overall stroller stand solution:

**Table 1.** Number of credits according to the building design.


**Table 2.** Minimum required bike or stroller spaces according to the number of the apartments.


Additionally, there is a condition on the minimum storage floor area of bike or stroller based on the number of housing units.

The final credit count is calculated as the weighted average between all house units.

According to the IFC schema, we can complete an assessment of stroller stands by three different methods according to Sections 2.3.1–2.3.3. The utilized IFC data structure and semi-automatic workflow were selected. A description of part of the assessment is shown in Figure 6.

**Figure 6.** Bike and stroller spaces in the model.

Geometry is represented by a 2D or 3D object placed in the model. Their properties contain real values, which are compared with pre-set benchmarks, or they are used in the calculation.

#### *3.2. Properties*

#### 3.2.1. Logical Value Type

The Boolean type logical value represents the values "true" or "false" and is used to automatically assess the presence of the specific predefined object in the model. If the object is placed in the model, it is possible to track the total numbers of predefined objects. In the sustainable quality assessment, this tool can be used to track the number of places for waste management, the number of parking places equipped with charging stations, etc.

This approach was used when an automated algorithm indicated that the model contains spaces for bikes and strollers.

#### 3.2.2. Alphanumerical Value Type

This is the most commonly used type of parameter. It includes text values (description of all parameters), numbers (counts, areas, volumes, etc.), or any other property, such as material name, weight, and warranty.

In the proposed use case, the properties used for the description and number of elements are shown in Figure 7.


**Figure 7.** Schedule of spaces for bikes and strollers.

#### 3.2.3. Enumeration Type

Specific properties of real or virtual objects can be defined according to a predefined structure following a quality assessment of specific criteria within the assessment framework. The enumeration type has distinct values that can be compared and assigned, e.g., roomy types, area types, façade types, etc. It is similar to a drop-down menu; this type of information is not possible to incorporate in the model (property of an element cannot contain more values from which the user can select one), and these values have to be written manually (or in a third-party software that allows using this functionality).

#### *3.3. Data Processing*

Complex quality assessment methodologies are more complicated than the standard LCA method already described in the literature and for which only the Bill of Quantities (BoQ) is necessary. A combination of all mentioned data types is needed for assessment; therefore, a more sophisticated data processing workflow is necessary. For SBToolCZ, an online application was developed. The whole assessment was completely online. The limitation is that so far, the application has only prepared the Excel tables. Therefore, it is not yet possible to import the IFC model directly. This is planned for version 2 of the application. The example with the TiCo project description is shown in Figure 8. The application can semi-automatically (with data prepared in the schedules) assess the building project, including different variants. It is suitable for the material and technical optimization of the project.


**Figure 8.** Description of the project in the online SBToolCZ application.

#### **4. Discussion**

#### *4.1. The Data Structure of Assessment Schemes and Methods*

This paper aims to analyze and verify possible BIM data-driven workflows for complex building quality assessment (CBQA) and a suitable BIM data structure set-up for automatic assessment and evaluation.

The development of most currently used assessment schemes and methodologies, such as LEED and BREEAM, including SBToolCZ, which is based on SBTool, started 15–20 years ago, and the structure of criteria and benchmarks was not created while considering the possibility of automatic assessment. Additionally, the IFC standards at that time were not established as a vendor-neutral and open system for data management. One

of the problems that limit automated assessment is that the assessment methodologies do not follow the data structure used for automated processes.

Every single project and every single building are specific according to unique architectural design and structural, technological, and material solutions. One of the crucial aspects of assessment methodologies is to support the architectural quality, diversity, and originality of the architectural design of every building. This is the reason why many assessment criteria and benchmarks are often described in a continuous text, in examples of possible solutions that should motivate and which are not defined as finite enumeration.

For example, one of the quality parameters of a living unit is defined by the quality and size of external relaxation zones, such as balconies and loggias. It is impossible to make a finite enumeration of all spaces in the building that can be used for this purpose.

One of the missions for the future development of complex quality assessment methods is to adjust the criteria and benchmarks structure according to the data types, allowing a higher level of automatized data processing. In this process, it is fully recommended to consider the IFC data structure by adjusting semantic and classification of data objects, their properties, and relationships.

#### *4.2. The Complexity of the Assessment Methodologies*

The aim of using complex quality assessment methods is to cover wide aspects of building quality. It covers a wide range of criteria, including environmental quality and LCA, energy performance of buildings, durability, and resilience, quality of the internal environment, architectural and functional quality, and other social issues in considering the quality of locality. Due to this complexity, the assessment methods are very complicated and include many parameters from different branches. As mentioned above, SBToolCZ for residential buildings requires about 500 parameters for almost 35 criteria for the assessment. The data analysis of SBToolCZ proven within the case study shows that about 200 parameters can be derived from BIM and can be obtained based on current design standards, thus enriching the conclusions contained in the articles of Portuguese colleagues on the topic SBToolPT-H [18,36].

The question is if the ratio of BIM-based data can be higher than almost a half. According to the analysis of SBToolCZ data structure, when keeping the complexity of the assessment scheme, the ratio of data from other sources (GIS, technical reports, databases, processes) will still be very high. One option is to simplify the assessment methods, especially for use in the predesign of the early design phase, to use mainly BIM-based data for automatic assessment and optimization in this project phase. Another option is to adjust the structure of existing criteria in the assessment scheme so that it is better linked to the IFC data structure, and data mining from BIM will be easier.

#### *4.3. BIM Data-Driven Workflows*

Wastiels and Decuypere [36] divided BIM-LCA integration into five types: (i) integrating the tools by exporting the Bill of Quantities (BoQ) from the BIM environment into other tools; (ii) importing surfaces using the IFC format with predefined LCA profiles; (iii) processing information from a BIM tool in a BIM viewer tool, which is then transferred to dedicated LCA software; (iv) using specially developed plug-ins that enable LCA analysis within the BIM tool; (v) including the LCA information in the BIM objects that are used in the BIM model, instead of them being attributed to the building components in a later stage and in separate tools. As Obrecht et al. [13] state, the most adopted is the exchange of information via BoQ from BIM. As presented in the previous text and shown in the case study, the CBQA is more complex and requires a wide range of parameters, so using BoQ but also other mentioned BIM–LCA integration is not sufficient and does not cover all required data. According to this fact, different BIM data-driven workflows have to be used.

Using a semi-automatic workflow with calculations and analysis outside the data model, and data export directly from native model or via export to IFC or utilized database export, is a relevant approach and ensured higher level of automatic workflow, but still, the BIM model corresponds with current BIM standards.

Considering the complexity of the assessment methods and their fast development, on the one hand, and the complexity and diversity and development of standards and legislation in the building sector, on the other hand, the semi-automatic strategy brings many advantages. Using an independent calculation tool outside the data model enabling data import in predefined data structures from BIM allows better control over the assessment process and the BIM model and data management. These findings are also confirmed by colleagues from other leading universities [6,7,12].

To be able to process all the necessary steps, it is important to have the BIM in very good shape (e.g., classification system, proper model health, correct and precise naming, BoQ). These requirements must always be known at the beginning of the project and accepted by all stakeholders.

In a utilized IFC data structure and semi-automatic workflow, the amount of manual assessment is significantly lower than in low structured model data and manual workflow. Thus, to be able to achieve a high degree of automation, it is necessary to move the whole BIM environment to a highly structured model data and fully automatic workflow. There are many obstacles and challenges (e.g., lack of legislation, low availability of skilled manpower), but the direction for the BIM environment to the higher level of automation and usage of robotics seems to be clear [50].

A literature review showed that specifically for the Czech Republic, SBToolCZ methodology is not digitalized, and legislation and standardization of the BIM are not finished; therefore, it is not yet possible to use highly structured model data and fully automatic workflow. These conclusions aim to advance, not only in the Czech Republic, the discussion regarding the necessary completion of the above-mentioned standardization process.

#### *4.4. Level of Information of BIM*

As was mentioned before, the SBToolCZ methodology needs around 500 inputs for the full assessment, and only around 200 of them can be taken relatively easily from the BIM. Due to the fact that the IFC format is very capable in terms of storing data, all the other parameters can be stored directly in the model as well. However, it is not clear how it would help to speed up the process of assessment. Currently, the IFC format is being developed rapidly, and no software can use its full potential of it. The only option to use all the possibilities is manual data import in the IFC file, which is inefficient. Further research showing the optimal amount of data taken out of the model is necessary for every mentioned complex quality assessment methodology. Similar conclusions regarding the IFC format have been reached in other scientific studies [41,43,45]. It is a question of how long this state will take before there is automation between individual data formats.

#### **5. Conclusions**

Using Building Information Modeling for complex building design and optimizing building quality in a design phase becomes an important part of integrated and advanced building design. Using data from an information model in the design phase allows efficient assessment of different design strategies and structural variants and a higher quality of final design. This paper aimed to analyze and verify possible BIM data-driven workflows for Complex Building Quality Assessment (CBQA) and a suitable BIM data structure set-up for an automatic assessment and evaluation. For an efficient automation process in complex quality assessment of the building in the design phase, it is necessary first to understand the data structure of the Industry Foundation Classes, which is widely accepted and used for buildings, and second, to understand the data structure of the assessment methodology used for the specific assessment. This article describes the potential of the automatic assessment of complex buildings based on the experience gained in a project where the complex building quality was tested on a real pilot project of a residential building using SBToolCZ, the Czech national assessment method. The outcomes present the experience with semi-automatic data workflow and recommendations for setting up the data model of a building for automatic assessment. The outcomes also present the experience gained by combining BIM and automatic complex building quality assessment based on semiautomatic data workflow combined with data export from a structured BIM model with the independent calculation tool. The case study proved this method as an efficient approach to automatic assessment.

**Author Contributions:** Conceptualization, J.R. and J.V.; methodology, Z.R.; investigation, S.V.; resources, S.V.; writing—original draft preparation, J.R., J.V., Z.R. and S.V.; writing—review and editing, J.V.; supervision, P.H.; project administration, J.V.; funding acquisition, J.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Technology Agency of the Czech Republic, grant number TN01000056, research project NCK CAMEB Centre for Advanced Materials and Efficient Buildings.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The research outcomes have been achieved within the research project BIMIP Building Information Modelling—Integrated Whole Life Cycle Process in the framework of the research project number TN01000056 NCK CAMEB Centre for Advanced Materials and Efficient Buildings.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

