Building information modeling (BIM) is a powerful tool for civil engineering projects, but it also requires careful management of data quality. Poor data quality can lead to errors, delays, conflicts, and rework, affecting the project performance and outcomes. How do you ensure consistent BIM data quality throughout a project? Here are some tips to help you achieve this goal.
Before starting a BIM project, you need to define the data standards and specifications that will guide the data creation, exchange, and validation. Data standards include the formats, structures, naming conventions, classifications, codes, and metadata of the data. Data specifications include the level of detail, accuracy, completeness, and consistency of the data. Data standards and specifications should be aligned with the project objectives, scope, and requirements, as well as the industry best practices and regulations.
Liaise with the client to understand goals first. Requirements must be realistically adjusted to the uses it will have, such as coordination and asset management, and the supply chain capabilities i.e. client must be aware that a higher requirement will have a higher cost, supply chain selection and management resourcing. Extra requirements “just in case” without real knowledge of why and how are likely to become a burden of time and cost. And similarly, too lax requirements are likely to bring teething issues and a missed opportunity for efficiency.
During the BIM project, you need to implement data quality checks to ensure that the data meets the standards and specifications. Data quality checks can be done manually or automatically, depending on the complexity and volume of the data. Manual checks involve reviewing and verifying the data by human experts, using tools such as checklists, spreadsheets, or visual inspections. Automatic checks involve using software tools or scripts to run data validation, analysis, or reporting, using rules, algorithms, or models.
Design the checking process so it can be automated as much as possible from the beginning, but be careful with making it so obscure (or so onerous) that only one person can run it. It is important that the process and results are clear and actionable for everyone involved. If you have to choose between clarity or precision, choose clarity, as it may mean the difference between the team checking something or dismissing the process at all.
To facilitate the data quality checks, you can use various data quality tools that are available in the market or developed in-house. Data quality tools can help you perform tasks such as data conversion, extraction, transformation, loading, cleansing, enrichment, integration, comparison, or visualization. Data quality tools can also help you identify and resolve data issues, such as errors, gaps, duplicates, inconsistencies, or conflicts. Data quality tools should be compatible with the data standards and specifications, as well as the BIM software and platforms.
There are notable digital tools for BIM data quality checks. Examples include Navisworks, Solibri, Simplebim, TrimbleConnect and Excel, among others. These tools perform some or all various functions such as data conversion, extraction, transformation, cleansing, enrichment, integration, comparison, and visualization. From experience, the use of two or more digital tools is crucial to ensure that BIM data complies with standards. Specifically, I use Navisworks for clash detection and visualization, while Solibri is useful for identifying inconsistencies in BIM model, conducting data validation for compliance with standards, and for data extraction. Simplebim proves instrumental for data cleansing, enrichment, manipulation, and integration.
To ensure that the data quality checks are carried out effectively and efficiently, you need to establish data quality roles and responsibilities within the project team. Data quality roles include the data owners, producers, consumers, managers, and controllers. Data owners are the stakeholders who define the data requirements and expectations. Data producers are the team members who create and deliver the data. Data consumers are the team members who use and benefit from the data. Data managers are the team members who coordinate and oversee the data quality processes. Data controllers are the team members who perform and report the data quality checks.
It is essential to define this at the beginning of the project/appointment agreement. It can be set, foe example, through a RA(S)CI matrix or roles and responsibilities table, with these roles then assigned to individuals. Having this clarity at the beginning and ensuring it is understood by everyone prevents conflict later on.
To ensure that the data quality is maintained and improved throughout the project, you need to monitor and evaluate the data quality performance and outcomes. Data quality performance can be measured by using indicators such as data completeness, accuracy, timeliness, consistency, and usability. Data quality outcomes can be assessed by using criteria such as data satisfaction, value, impact, and risk. Data quality monitoring and evaluation can help you identify and address the data quality issues, challenges, and opportunities, as well as provide feedback and recommendations for data quality improvement.
Set milestones for data checks, clarify what is expected at each one of them and get them agreed with the team. The data assignment may not necessarily happen in parallel with the design evolution, especially if the process is under time pressure - having reasonable milestones helps ensure it is not left behind.