BIM features to enhance building permit process activity: a partial least squares structural equation modelling (PLS-SEM) approach

. Construction permit issuance is considered one indicator to measure a country's business. Applying for digital building permits seeks to increase the ease of doing business. Still, over time the performance of the permit process must continue to be improved to increase competition in the ease of doing business between countries and encourage the level of foreign investment. Building Information Modeling (BIM) technology can enhance digital building permits. A quantitative method, based on the partial least squares structural equation modelling, was used to verify the proposed hypotheses. The findings display four process stages of building permits in the form of consultations, zoning assessments, technical recommendations and building assessments that affect the permit time performance. Two significant BIM features to implement in the permit process are 3D model visualization and automated code checking.


Introduction
A construction project requires a building permit from an authorized institution and is an essential milestone of a project [1].Building permits in urban areas are required to start a construction project in the form of an official document provided by a public authority that grants permission as an administrative decision to commence construction work under applicable laws, regulations and building requirements [2].The building permit process is a significant component of the institutional factors that influence the success of a construction project [3].A fundamental process even though it is a brief stage compared to the planning, construction and operational phases of a building [4].
Construction permit issuance is considered one indicator to measure a country's business [5].As technology develops, public authorities are replacing the paper-based conventional building permit process with the electronic building permit process through digital submission [6].Utilizing web-based information technology in the form of online submissions and uploading supporting documents is an electronic permitting (e-permitting) system proven to increase the efficiency of the building permit process [7].Since the beginning of 2018, the City Government of Jakarta has implemented digital building permits to improve the ease of doing business.However, increasing the performance of permitting times is increasingly needed to increase competition for ease of doing business between countries and encourage the level of foreign investment [5].
E-permitting is divided into three levels based on the automation and technology applied.Level 0 is a traditional or conventional permit process that uses paper without implementing automation, even though all planning preparations, submission and inspection use computers and related software.Level 1 is the basic e-permitting permit process where submission is made digitally or online, and supporting documents such as 2D building plan drawings are uploaded.Level 2 e-permitting implements digital filing using digital model files and various levels of process automation [6].

Fig. 1. E-permitting category [6]
BIM's innovative technologies have changed how construction projects are conceived, designed, built and operated [1].Various BIM technological capabilities are able to assist the efficiency of the permitting process in line with the application of BIM to the building permit process following the planning and design stages of the construction project life cycle [8] and the implementation of BIM-enabled basic e-permitting to advance to the next level [6].
This study aims to obtain a significant relationship between BIM features that can be applied to the building permit process to improve the performance of building permits.There was no past study that used quantitative methods through PLS SEM statistical analysis, and this has driven the study to use PLS SEM approach to prove the hypothesis of the significant relationship in the context of Indonesia and especially the city of Jakarta.
The discussion and findings of this study will provide an overview of the most influential BIM features according to needs that can be applied at the stages of the building permitting process to improve the performance of permitting times while simultaneously increasing the level of digital building permits.

Building Permit Process
Building permits are required for property owners to apply for building permits by filling out forms and providing supporting documents.In contrast, public authorities are required to check applications against various requirements and regulations and require coordination among various departments [9].The building permit process is a process that begins with the application for a building permit and ends with the issuance of a building permit by the public authority [10].
Empirical studies through data collection and evaluation from 2016 to 2019 show the flow of the building permit process in general, which is divided into three levels, namely level 1, the primary process of the building permit authority, level 2a is an assessment of submitted information is divided based on zoning regulations and building regulations.Level 2b is a process information checking with the involvement of the relevant technical government department [11], as seen in figure 2. Fig. 2. Schematic illustration of building permit process [11] The stages of building permit process activities before construction in European Union countries were studied by [12], and the Netherlands, the United Arab Emirates and Egypt were reviewed by [13].In general, the building permit process was divided into four stages consist of the consultation stage, the zoning assessment stage, the technical recommendation stage and the building assessment stage.
The consultation stage involves consulting the permit applicant with the public authority before submitting a formal building permit application [13].Consultations are carried out to obtain initial information about development plans and building functions on certain lands in accordance with urban planning regulations and to obtain information on specific development needs regards to the land [12].
The main process of building permit stages begins with a formal application through the system, and the inspection of permit content is divided into two processes which comply with zoning regulations and building regulations [11].The zoning assessment stage will be the basis for planning buildings according to spatial regulations but does not bind public authorities to grant building permits and not as a permit to start construction on the ground [12].Zoning regulations explain the conformity of building planning with spatial regulations such as building functions, the area used, permitted height, relationship with the environment, relations within the area and various other things [13].
The building assessment stage is where an examination of the proposed building plan is carried out in accordance with the applicable building requirements and regulations [11].This stage is a review by the permit authorities of the building planning drawings prepared by the planning consultant in the architectural, structural and mechanical electrical plumbing [13].Building planning that complies with the building requirements and regulations becomes the basis for building permit approval [11].
The technical recommendation stage is a statement and approval from the technical authorities related to the permit authorities [13].Technical recommendations related to the conformity of building plans with regulations according to the authority of technical agencies such as the building's impact on the surrounding environment, traffic generated, flight operational security and various other matters [10].The procedure for technical recommendation stages varies in different permit authorities, whether before or after obtaining zoning permits, depending on the regulations [11].

Building Information Modelling
The use of Computer-Aided Design (CAD) in several industries in the world began to recognize three-dimensional models in the early 1970s, which was the initial concept of Building Information Modeling (BIM) [14].BIM is a technological enhancement of CAD that aims to increase the productivity of planning, building construction and building management in the construction industry [15].BIM was created due to integrating information and communication technology in the developing construction industry in the early 2000s [16].
BIM is an approach that enables information technology to implement an integrated digital representation of a building and capture extensive information about different phases of the project life cycle in the form of data stores [17].BIM can be utilized for many purposes, such as integration of design and construction, cost estimation, scheduling, coordination, energy simulation, safety and facility management [1].Planning, design, building, and operation comprise the four stages of the BIM application in the life cycle of a construction project [8].Fig. 3. LoD stage criteria [18] The concept of Level of Development (LoD) was introduced in BIM to enable construction workers to define a high level of content clarity and the reliability of 3D models at various stages [19].LoD distinguishes between geometry and property information.A model can be developed well through geometry without having much information about the properties of the model, or the model can have a low level of property and a high level of geometry [20].Utilization of LoD 100 is used at the conceptual design level, LoD 200 at the design development level, LoD 300 at the documentation level, LoD 400 at the construction level and LoD 500 at the management level [19] [20].
3D model visualization is able to display actual modelling in real-time, which can facilitate visualization and help planners find problems with drawings, effectively increasing the accuracy and depth of architectural designs, which were previously limited by 2D drawings [8].Adding room names and other information to 3D models facilitates area calculations [21].Using specific LoD criteria, the 3D model can be applied to the entire building permit process based on the information requirements of the 3D model [20].
Improved accuracy of 2D planning drawings through 3D models that can be converted and directly facilitate the calculation of the area of a room or an area [21] while simultaneously establishing work standards and information templates developed in a single 3D model by various disciplines [21].
The benefits of automated code checking on BIM by using the property values of BIM elements to meet specific building requirements and regulations can be achieved by translating building requirements and regulations into machine language [21] [23].It can be used for independent verification by planners of building plans against building requirements and regulations [20].
Based on the review literature above, four phases of the building permit process influence the performance of building permit time, and a faster processing time for building permits will result in a higher permit time performance.Digital permits can be improved through the application of BIM through various technological capabilities.BIM can help improve the efficiency of the permit process in line with the application of BIM in the building permit process [8], and the application of BIM at all permit process stages was able to elevate the basic digital permitting level to the next level [6].These BIM capabilities are visualization of 3D models, collaboration, online platforms and automated code checking [22][23] [24].
The relationship between the permit stages process and the BIM features applied to it forms the basis of this research, as seen in figure 4. The research hypothesis is proposed as follows:

H1
The consultation stage has a significant effect on the permit time performance.

H11
The BIM 3D model visualization has a significant effect on the technical Recommendations stage.

H2
The zoning assessment stage has a significant effect on the permit time performance.

H12
The BIM 3D model visualization has a significant effect on the building assessment stage.

H3
The technical Recommendations stage has a significant effect on the permit time performance.

H13
The BIM collaboration has a significant effect on the technical recommendations stage.

H4
The building assessment stage has a significant effect on the permit time performance.

H14
The BIM collaboration has a significant effect on the building assessment stage.

H5
The BIM 3D model visualization has a significant effect on the permit time performance.

H15
The BIM online platform has a significant effect on the consultation stage.

H6
The BIM collaboration has a significant effect on the permit time performance.

H16
The BIM online platform has a significant effect on the technical recommendations stage.

H7
The BIM Online platform has a significant effect on the permit time performance.

H17
The BIM online platform has a significant effect on the building assessment stage.

H8
The BIM Automated code checking has a significant effect on the permit time performance.

H18
The BIM Automated code checking has a significant effect on the zoning assessment stage.

H9
The BIM 3D model visualization has a significant effect on the consultation stage.

H19
The BIM Automated code checking has a significant effect on the technical recommendations stage.H10 The BIM 3D model visualization has a significant effect on the zoning assessment stage.

H20
The BIM Automated code checking has a significant effect on the building assessment stage.

Methodology
The research questions stated at the beginning of the study became the basis for exploring the literature to determine the variables that became the content and construct of the research to be validated by experts through two rounds of the delphi method, which is a gradual process used to collect and filter expert judgment using a series of questionnaires interspersed with feedback to answered the research question [25].Then a questionnaire survey was carried out to the respondents, the results of the questionnaire will be analyzed using PLS SEM whose results will be discussed to become research findings.As seen in Fig. 5.
Validation related to content and constructs, as well as research indicators, was carried out by ten experts from individuals who are experts with specific experience and knowledge in the field of building permits and BIM.An expert is generally considered to be more experienced, more proficient, and more significant in the work environment [26].Expert validation defines as many as 34 indicators as questionnaire statements.There are 21 indicator activities in the building permit process and 13 indicators in the BIM feature that will be use in questionnaire survey.

Sample and measurement scales
This study aims to review the existing building permit process and identify BIM features that can be applied to the permit process.Therefore, the individuals from departments authorized to issue building permits will be the population of this study to respond to the survey appropriately.Criteria for respondents with experience working in a permitting agency for a minimum of four years, which is presumed that years of experience is needed for employees to know and be familiar with building permit process activity and have knowledge about BIM.
The required number of samples is determined accordingly to the 10 times rule for determining the minimum sample size in a PLS-SEM analysis [27].The proposed research model consists of eight independent variables and one dependent variable.Therefore, according to the 10 times rule, the minimum number of respondents must be at least 80. Respondent questionnaire data collection was carried out online using the Google Form link, which was sent via the WhatsApp application to all respondents.The questionnaire had two sections, demographic details of respondents, including education and years of experience in the first section.The second section consisted of statements on the selected constructs in the form of rating scales using a Likert scale anchored between strongly disagree (indicated by 1) and strongly agree (indicated by 5).The sample of 80 respondents successfully collected from building permits departments, research data provided in table

Data analysis technique
Data were analyzed using SmartPLS version 4.0 software which developed a conceptual measuring model for analysing observable properties [28].Variance-based PLS-SEM approach was adopted because PLS-SEM was based on its ability to simultaneously estimate causal relationships among all latent constructs while accounting for measurement errors in the structural model [27].Also, the benefit of employing this approach is that it can handle actual issues in the nature of the data, like data noise, missing data, and skewness [29].A separate evaluation will be conducted first for the measurement model and continued for the structural model.

Result and analysis 4.1 Analysis of measurement models
Considering the guidelines of [27], constructs with measurement models, i.e., Consultation Stage, Zoning Assesment Stage, Technical Recommendations Stage, Building Assesment Stage, BIM 3D Model Visualization, BIM Collaboration, BIM Online Platform, and BIM Automated Code Checking, were analyzed for their reliability and validity.All constructs were assessed for their composite reliability (CR) and Cronbach's alpha values, higher than 0.70 critical level.The value of Average variance extracted (AVE) for all constructs was also higher than the critical value of 0.50 [27].Complete results of the validity and reliability of all constructs are presented in Table 2.These results meet all requirements for establishing the validity and dependability of measurement models.The test for the discriminant validity of measurement models was performed in two steps, first by the Fornell-Larcker criterion presented in Table 3. Bold values show the square root of AVE, which is higher than the estimated correlation values, thus demonstrating the discriminant validity of constructs included in the proposed measurement models by evaluating all cross-loading values of reflective constructs indicators.The rule of thumb was indicators of reflective measurement models should have the highest loading value on their underlying latent construct than other constructs involved in the structural model [27].
Table 4 provides a comprehensive rundown of the cross-loading values for each and every indication that was incorporated into the various measurement model constructs.The results demonstrate that the discriminant validity of the measurement models is supported by robust evidence and demonstrate that the evaluation criteria for cross-loadings have been met.Based on the above discussion, the suitability of formative constructs is established and the overall assessment of measurement models demonstrates acceptable results to evaluate the structural model.

Analysis of the structure model
The structural model was assessed for the overall explanatory power of constructs through R² value and path coefficient β-values.The findings of the structural model are presented in Fig. 5.A summarized overview is shown in Table 5.
The result indicates that the proposed model with R²=0.543 has 54.3% of explanatory power for permit time performance.It is found that H1, which is the relationship between the consultation stage and permit time performance (=-value=2.918;=.004), is positive and significant.Moreover, H2 is the relationship between the zoning assessment stage and permit time performance (=-value=3.852;=.000) is positive and significant.H3 which the relationship between technical recommendations stage and permit time performance (=-value=3.101;=.001) is positive and significant.H4 which the relationship between the building assessment stage and permit time performance (= -value=3.453;=.000) is positive and significant.H5 which relationship between BIM 3D model visualization and permit time performance (=-value=3.615;=.000) is positive and significant.H6 which relationship between BIM collaboration and permit time performance (=-value=0.087;=.931) is positive but not significant.H7 which the relationship between BIM online platform and permit time performance (=-value=1.024;=.306) is positive but not significant.H8 which relationship between BIM automated code checking and permit time performance (=-value=3.115;=.000) is positive and significant.These were the relationship between independent and dependent latent constructs.As shown in Fig. 6, the R² value of the structural model is 0.543, which indicates that the proposed conceptual model has adequate explanatory significance [30].R² value is considered the primary way to evaluate the explanatory power of the model [31].Although, PLS-SEM do not generate overall Goodness of Fit (GoF) index to assess the model fit.The goodness of Fit (GoF) is measured by using the geometric mean value of the average communality score (AVE values) and the average R² values (for endogenous constructs).It is calculated using the following equation, (GoF=√(AVE×R²)).The cutoff values for assessing the results of the GoF analysis: GoFsmall=0.1;GoFmedium=0.25;GoFlarge=0.36 [32].The calculation of the Goodness of Fit (GoF) index is presented in table 6.The research model used in this study produced a Goodness of Fit (GoF) index value of 0.608, which suggests a very good (GoFlarge) model fit.

Discussion
The findings of this study show that the building permit process, which consists of four stages, i.e., the consultation, the zoning assessment, the technical recommendation and the building assessment stage, has a significant and positive influence on the performance of the permit time.These findings are in line with the previous studies conducted in a few countries about building permit process stages and their activity [12] [13].Although the terms at each stage of the process and activity can be different, in essence the process is needed to assess a building plan according to the regulations that apply to the context of the research location.According to research [12], the building permit process for each country may have different procedures and terms depending on local authority regulations, the complexity of the building, and the building permit applicant.
The four features of BIM can be applied to the building permit process to improve permit time performance, i.e., 3D model visualization, collaboration, online platform and automated code checking [22][23] [24].However, the findings stated that only two features had a significant positive effect which is 3D model visualization and automated code checking.Visualization from 3D models as part of submitting building permit documents is beneficial for ensuring a more precise and efficient process [33].Application of BIM 3D model visualization in the zoning assessment stage by the capability of information take-off that could be applied in calculating building intensity to comply with city zoning regulations will make it easier for permitting authorities to carry out the assessment.The existing permitting process requires manual checking with the support of two files, consisting of AutoCAD 2D planning drawings and Microsoft excel calculation tables, while BIM through Autodesk Revit only requires one file, which includes planning drawings and calculation tables, thereby reducing manual calculation errors.This support the study that stated BIM could facilitate the calculation of the area of a room or an area according to the needs of the assessment carried out by the permitting authorities [21].BIM LoD requirements can be adjusted to the needs of the building permit process assessment [20], LoD 200 3D model is considered sufficient to help understanding and assessment process in the consultation, the zoning assessment and the technical recommendation stages in contrast to the 3D model requirements at the building assessment stage, which must be more informative and on a par with the project's documentation level or at least in LoD 300.
Although there is no building permit public authorities have not yet applied fully for an automated code checking feature, but recent studies indicate that much research is moving toward it [20][21] [23].One of the automated code checking recommendations is how the building meets the green building requirements conducted in the building assessment stage, such as the value fulfilment of OTTV (Overall thermal transfer value), at the LoD 300 3D model level, BIM has the ability to accurately predict OTTV value by using visual scripting to extract thermal and physical parameters from the BIM database to facilitate the calculation of thermal transfer value [34].This can replace the manual calculation method which can be time-consuming, to improve the permit time performance.
The BIM features that doesn't have a significant impact is collaboration, it might happen because the target of the questionnaire respondents came only from government agencies where the use of BIM did not yet exist in permitting authorities, and BIM just started to be known in Indonesia the past five years, henced with an understanding of the current building permit process which employs 2D CAD drawings that have implemented collaboration features from various disciplines.It is aligned with research that explains the knowledge of BIM in permitting agency employees is less than professional office employees [2].The BIM online platform also has no significant effect.It is probably due to the respondents' understanding regarding the condition of current building permits that have implemented the digital permitting process with online platform technology, in line with the statement that standardized electronic permitting with digital applications and a centralized online platform connecting government and applicants [6].The study also revealed that respondents who were employees of permitting authorities still needed to understand BIM and its capabilities thoroughly to implement the BIM-based building permit process.Following the gradual application of BIM at each stage of the building permit process, a strategy to increase BIM knowledge for building permit authorities is required.

Conclusion
The four stages of the building permit process, i.e., consultation, zoning assessment, technical recommendation and building assessment stages significantly influence the permit time process, and technological improvement through BIM based process in these stages, namely, 3D model visualization and automatic code checking, will enhance permit time performance.This study's findings suggest practical implementations for Jakarta's building permit authorities to concentrate on the significant BIM features in developing BIM-based building permits, provide the policy for BIM permitting submission guide and implement a strategy to improve the comprehension of BIM for the authorities' human resources.The BIM-based building permit development should be in the simplest and fastest BIM process, through BIM 3D model visualization information take-off based assessment to gradually replace the manual assessment and encourage automatic code checking to advance digital permitting to a higher level, furthermore achieving the efficiency of the building permit process to boost Indonesia's competitiveness among countries for ease of doing business and to improve the degree of foreign investment.

Table 2 .
Validity and reliability of latent constructs.
* Values in the bold are Square root of AVE.

Table 4 .
Cross loadings among reflective measurement scale items.
Note: Bold values are loadings for items above the recommended value of 0.5.

Table 6 .
Calculation of Goodness of Fit (GoF) index.Based on a thorough analysis of the measurement model and structural model, it is concluded that both models, i.e., measurement model and structural model, are validated.Moreover, These results indicate that this study's proposed theoretical model has moderate explanatory power.