The BIM Quality Problem: What’s Broken and How We Can Fix It

I’ve discussed BIM quality with contractors, quantity surveyors, and software developers. Invariably, they complain about the quality of BIM models they receive, especially from architects. If we want to automate digital workflows, the information quality of models is critical and should be ensured to a much higher standard than it is today.

Poor Modeling Is a Common Dilemma

A few days ago, I interviewed Martin (Marty) Rozmanith, co-founder and CTO of Skema.ai (I will release the interview video soon). Skema is the world’s first BIM Knowledge Reuse Engine. The software allows architects to create new designs from a catalog of dynamic model blocks.

Skema ingests existing Revit models as the raw material for the catalogs. I asked Marty about the quality of the architects’ models they use as the starting point. He admitted that they often find many issues, such as messy wall joints and mislabeled spaces.

Initially, Marty’s team cleaned and repaired the models manually; now, they utilize AI to automate the process.

I have heard similar comments about BIM quality from other software developers and companies that want to use architectural models for quantity takeoffs. One large general contractor informed me that they have to remake the models to make them usable for QTOs.

Why Model Quality Matters

Construction is the most networked industry. Any deviation from the required BIM data quality has a ripple effect across the value chain. Here are a few examples of processes that rely on quality data:

• Building permitting: Automating building permitting, as Finland and some other countries are starting to do, relies on impeccable geometrical and metadata compliance.
• Quantity takeoffs: If models have unnoticed duplicates, missing objects, or incorrect metadata, an erroneous QTO can lead to dire business consequences.
• Sustainability calculation: ESG reporting, the EU’s taxonomy requirements, and national CO2 emission targets require data about a building’s materials and energy consumption. Such data can be best obtained from a quality-managed model.
• Procurement: To automate supply chain processes, project data must be reliable.
• Digital twin creation: BIM models are essential for creating digital twins used in property and facilities management, and they must be trustworthy.

What Leads to Quality Issues

Model creators, mainly designers, are naturally responsible for achieving BIM quality. The problem is that design agreements do not specify the BIM quality that the client requires.

Furthermore, those who purchase services are not experts in BIM quality issues and don’t budget funds for more detailed design work or data consistency. For many, a model that looks visually fine is good enough.

A key issue is that many stakeholders still rely on printed drawings, which are also the legally binding documents. Model errors are not necessarily visible on paper.

I also suspect that architects are not always trained to consider the data quality of the models, and that the tools they use don’t support it sufficiently.

How to Fix This

Solving the BIM quality dilemma starts with accountability. Every project should have precise, contractually binding BIM requirements that define acceptable data quality standards.

An engineering firm CEO told me that “we are basically PDF factories.” Architects and engineers must treat model data as critical project infrastructure, not just a tool for efficiently outputting drawings.

Training programs should emphasize data literacy and regular quality assurance practices, rather than leaving them to chance or relying on AI cleanup after the fact. BIM coordinators should utilize their expertise to perform more value-added tasks than merely identifying model errors.

Technology also needs to play a bigger role earlier in the process. Integrated quality checking tools and smart automation, such as those offered by Solibri, should be standard components of every modeling workflow.

Most importantly, clients must start valuing BIM quality properly. Investing in BIM quality yields a high return on investment in the long run.

BIM is no longer optional. It’s infrastructure. It must be built right.