Quality Assurance for BIM models.
Swissroc Building Intelligence
Technological innovation at the service of real estate.
?Customers don't form their opinions on quality from marketing.(…) They form their opinions on quality from their own experience with the products or the services.”
Steve Jobs
1??Introduction.
Having a consistent and high-quality information model can save money and time during the construction and operations of an asset. Finding and removing potential errors and information deficiencies or visualizing risks in a virtual environment is much cheaper than dealing with their consequences on site and in subsequent years of use.
Although market awareness of BIM is still largely shaped by colorful visuals and PR, ultimately clients will test the gains of implementing the methodology. Properly set objectives, well-chosen BIM Uses and documents (e. g. EIR followed by BEP) certainly help. Ultimately, however, we need competent specialists capable of implementing procedures and verifying the effects of BIM related works.
In Swissroc Building Intelligence we specialize in coordination and validation of ?BIM models. We often conduct the verification for mature and BIM-aware Clients who deliver Exchange Information Requirements (EIR) and negotiate BIM Execution Plans (BEPs) with the designers. The designers themselves are often elite multinational companies. Even so the models we check usually have hundreds of collisions, thousands of elements lacking agreed information and multiple other discrepancies which disqualify them when it comes to execution works and incorporating the BIM database with facility management software – see. Figure 1 below.
Even the best planned and implemented process based on recognized standards will therefore not bring value without continuous monitoring. Cyclic verification of models but also methods of their creation and use are critical factors for correct implementation of the methodology. In this respect BIM follows methods used in modern process management systems.
I describe issues related to quality assurance through BIM methodology. However, we must remember the obvious – namely that achieving high quality of the design, construction works, and operations require other processes than those directly related to information models. Here I focus on the general practices and methods of working with information models, without extensively describing the other process like verification of building code adherence, safety, efficiency of maintenance, etc., where nevertheless BIM can help a lot.
Figure 1 MEP Elements lacking proper information in accordance with BEP. Source: Swissroc Building Intelligence.
2 The role of QA in BIM.
Quality Assurance (QA) (not to be mistaken with Quality Control) is based on a Total Quality Control (TQC) system linked to the Lean philosophy. The system is based on the belief that if the results obtained at each stage of the process are of high quality, the final product will also be of high quality. If this is the case, the process is said to be stable. If not, the causes of the defects must be found, a strategy to eliminate them must be developed and implemented. The system includes issues such as employee training, hiring high-level specialists, purchasing the right software, or choosing the right standards.
To start with, appropriate requirements for QA should be an essential part of the BIM documents (EIR/BEP). As always it is necessary to start from the final product, which is the model used for maintenance or at least for the execution works, so that the QA is focused on the end goal and not particular interests of the BIM modeler. The BEP should include a detailed QA procedure with methods for verification and analysis of the models within the implemented BIM Uses.
But why do it at all? Isn’t design coordination enough? What are the actual objectives of QA? The primary objective of QA is to track and resolve inadequacy before product release (documentation delivery, commissioning, commencing refurbishment).
Among the other objectives of QA we can count:
It must be noted here that when talking about information models control, we are not only talking about BIM models, but also about PIM and AIM models, aggregated during the full investment cycle – preferably on the Common Data Environment.
We can distinguish 3 main stages (see Figure 2) of information model’s control:
This is very important. Especially when discussing complex BIM models where a person modelling has little opportunity or time to validate a model in a broader context of other disciplines and/or project standards. Below I describe general good practices on each stage of the validation to avoid problems like the ones described in the introduction.
Figure 2 Subsequent stages of quality control.
3?QA on the leveL of task teams.
The person responsible for producing a particular model will primarily check the model in a native software used for modelling. This is because regardless of the required coordination formats and final deliverables submitted to the client, only a properly developed model in native format will allow for efficient export and communication with other parties. The control should be the responsibility of a designated person or role in the team because a certain mindset and skillset is needed for this task. This said, other team members should also feel responsible for checking the effects of their work. Apart from substantive aspects (proper selection of materials, loads or equipment efficiency) everyone should be involved in the ongoing control of compliance with the standard imposed by the BIM Manager of the project.
A dedicated person (coordinator) performs cyclic comprehensive verification of the model before sending the results for coordination to the Lead Appointed Party. The process should take place at least biweekly, so that the number of corrections do not pile up. Typical tasks at this stage may include checking that:
In the case of coordination based on open (IFC) formats, among others we should also check, that:
It is a good practice to have a standardized checklist at this stage. Especially at the beginning of project the number of issues will be huge and it is good to remember what actually needs checking. The result of the control at this stage is a report delivered to the BIM Manager and individual team members responsible for making necessary corrections to the models. It is best to keep the communication on the dedicated information exchange platform.
Table 1 Fragment of a model checklist. Source: Swissroc Building Intelligence.
4?QA on the level of the lead appointed party.
Before the federated models are submitted to the client they need to be properly verified for completeness and correctness both in terms of individual branch models and the interfaces between them. To do this BIM Manager/Coordinator should assemble all transmitted materials into a federated model and perform a number of predefined tests. As mentioned, the verification procedures should be included in the BEP and accepted by the client and model developers.
Importantly, the role of the BIM Manager is not to correct the models submitted by the subsequent disciplines. Even if these are editable files, they should be treated as “read only”. The result of the coordination is a report with proposals for corrections delivered to the responsible designers, e.g., as BCF file transmitted through the CDE platform. The BEP defines several documents that are elements of the quality assurance plan. These include model checklists, clash checking rules, or clash elimination hierarchy matrices (Figure 3).
领英推荐
Figure 3 Simple collision matrix table example. Source: Swissroc Building Intelligence.
Typical tasks at this stage include (see also point 6):
Properly used dedicated software can substantially reduce time needed for the model’s verification at this stage. For example, in Swissroc BI we create validation templates and filters based on the information from BEP. That way we can process many revisions automatically using predefined rules for checking information and geometrical discrepancies. ?Effects of these tests come in various forms. We use BCF file format so that coordination is faster but also provide printouts and pdfs for coordination meetings. This way other stakeholders can take active part in solving of issues. On Figure 4 an example of structured model validation report.
Figure 4 Predefined validation report. BIMcollab Platform. Source: ECCBIM Foundation.
5??Quality assurance at the level of the Client.
In principle, quality control on the part of the contracting authority should be similar in scope to that carried out on the previous steps. After all, the contracting authority is exposed to the highest costs associated with poor quality documentation. In practice however, clients rarely have the necessary software and time/staff to fully assess the quality of the delivered models. Consequently, just as in our case, investors sometimes mature cede this duty to an external consultant, e.g., a company with BIM competences providing the information management services.
Regardless of the manner and scope of verification conducted by the contracting authority, the responsibility for any errors is usually borne by the designer. This seems obvious, but in the complex processes of creating, submitting, verifying information models, the responsibility for the quality of design solutions may seem to be diluted. Meanwhile, the methodology is intended to create a sense of transparency and be a tool for even better control at each stage, which is why it is so important to develop and consult a quality assurance plan in BEP.
The ISO standard links quality control and assurance to the Common Data Environment (CDE). The structure of the CDE, divided into subsequent containers (Work in Progress, Shared, Published, Archive), is intended to ensure proper verification of the products during three main stages described above.
The configuration of the CDE environment, which at least partially ensures the automatic control of the delivered data, may be therefore a key element of the QA procedure. Advanced CDE environments enable the control of e.g. proper naming, metadata assignment, proper administration of processes for approval of documents, file versioning, etc.
6??BIM model control in PRACTICE.
We have discussed the general levels of BIM QA. The remaining question is: “how in practical terms do we validate BIM models received from our contractors or partners?”
For BIM models there are two main methods used for quality assurance: control and analysis. BIM tools provide enormous opportunities for analyzing and controlling the produced information. Here we will focus on methods related to the coordination of models regardless of the BIM Uses proposed for the particular project. They are primarily aimed at verifying that the standard set in the BEP is met. Let us start with the basic methods for model checking applied by our team.
·??????Visual inspection.
This seems to be the simplest method of checking for basic errors and inconsistencies. It’s enough to open a discipline or federated model in a dedicated viewer to detect model positioning errors, information gaps or incorrectly modelled geometry. The visual check should be the responsibility of each modeller but is especially important during federation of models before Data Drop. A random check (so-called statistical check) of individual models, groups of components or sets of information can be surprisingly effective – especially in the case of sloppily executed models. We apply standardized procedures for visual inspection by using visibility filters, lists of parameters, level checks, etc. But inconsistencies pop right away when you turn on and off different sets of elements or information. Sometimes this method is the only way to check for model inconsistencies – e.g., continuity of the MEP systems.
·??????Compliance with the standard
This consists of checking the adherence to the information exchange standards set in EIR and BEP. Elements such as naming convention, file formats, level of detail of the individual components, their visibility, relations, connections, etc. are validated through proper checklists. ?The process is especially important before submitting the materials for sharing with other parties. It can be supported by a properly configured CDE solution where information standard is being validated during data upload- e.g., naming convention or file formats are rejected if not set correctly.
·??????Clash detection.
This involves using dedicated software to automatically test for geometric collisions (such as overlapping or intersecting model components or their incorrect distance to each other). It should be performed according to predefined rules defining the range of checked elements and possible tolerances. Apart from detecting the collisions they should be grouped and labeled in accordance with:
Again, when preparing the coordination standard, at SRBI we always create predefined rules for the collision check procedure. This enables for quicker assessment of the models on the subsequent stages of development.
·??????Control of non-geometric parameters of models
The verification of non-geometric parameters is performed based on predefined filters and lists developed in dedicated software. These filters allow for the automatic location of unfilled or wrongly filled parameter values (e.g., a “size” parameter without a numerical value or a "manufacturer" parameter without a name value). In addition, it is possible to carry out a cyclic export of component lists in the form of spreadsheets, where discrepancies can be found by filtering the components and their values accordingly. The control serves to maintain the completeness of information required by the ordering party and included in the BEP.
The main objective for this is to fulfill the information needs of the Client – e.g., for the Facility management purposes.
7?analyses
Depending on the goals applied to the project the Contracting Authority may agree on the scope of necessary analyses. In principle, they will be influenced by the BIM Uses selected for the project and in practice also by the competence of the designers, the available software, and the project budget. Managers often do not have time to go into technical details related to the implementation of subsequent analyses but It is important to limit those to absolute minimum for achieving the goals set by the Client.
Obviously, the BIM models may become one of the key sources for such reports. Digital tools (e.g. Design Builder, Microsoft PowerBI, or even Excel) can analyze and integrate data in the form of cost schedules, reports, charts, maps and diagrams. The advantage of this type of solution is the possibility to update an audit quicly ??once it has linked to model elements. The disadvantage is the need for a competent person to prepare the report, often during an intense period. However, the extra work may be worth it, because a well-developed dashboard will allow quick identification of risks to the project. Figure 5 shows an excerpt from a sample visual report produced from data systematically exported from a BIM model.
Figure 5 Power BI dashboard generated from the BIM model. Based on the template published on ProvingGround.io.
8??Summary,
A well-designed, agreed-upon and implemented quality assurance plan help in eliminating issues and discrepancies almost as soon as they occur. This is very important because correcting errors later takes a lot more time and may even be impossible due to the decisions made in the process. Roles and responsibilities should be established but also a culture for the everyday control of models need to be in place. CDE platforms can also help in proper validation of information and distribution of found issues. The quality assurance process needs to be the responsibility of all people involved in the project - not only BIM enthusiasts. But to make it work, the procedures need to be simple and the end results legible for all engaged parties.
Coming back to the Clients – the quality assurance should always be based upon their needs and goals. Well prepared EIR should be the first source of information about the quality standards. Therefore, owners and clients should make sure that established standards are being checked on every stage of model control, because only properly validated work can bring real benefits of BIM during the execution and operational phases. In practical terms – clear rules, standardization and repetitiveness are the key factors of efficient control on every stage of the process.