By Pierre Monico, Sortdesk
This article explores how the Information Delivery Specification (IDS) enables structured, automated checking of IFC data – moving the industry beyond unreliable visual inspections toward repeatable, specification-driven quality assurance. It reflects on practical lessons from applying IDS in real workflows, the challenges of poor specifications, and the opportunities automation creates for earlier feedback and clearer accountability in digital construction.
I’m Pierre Monico, co-founder of Sortdesk. We’re a software agency focused on the AEC industry, delivering high-quality services like BIM automation, data integration, and productivity tooling for design and construction firms. Alongside client work, we develop openBIM tools that support structured data validation, including IDS-based checking and tabular data QA. A version of this thinking was first presented during our session at Digital Construction Week 2025.
The Problem With Visual Checking
In many BIM workflows, information validation is still driven by what can be seen in the model viewer. Geometry is reviewed, elements are inspected visually, and models are often signed off based on aesthetic or spatial accuracy. But this misses a significant portion of the deliverable: the structured data behind the model.
Too often, IFC files that “look fine” are later found to be missing key data fields, misusing property sets, or failing to meet naming conventions. These issues only surface late – after submission, after coordination, or worse, after delivery. That lag causes rework, uncertainty, and frustration. Relying on visual inspection alone simply isn’t scalable for projects with structured information requirements.
What IDS Brings to the Table
The Information Delivery Specification (IDS), developed by buildingSMART, introduces a standardised, human- and machine-readable way of specifying expected data in IFC deliverables. Rather than relying on ad hoc documents or spreadsheets, IDS formalises requirements in a structured format: properties, classifications, data types, value constraints, and more.
When used in conjunction with a checking engine, IDS enables automated validation directly against the IFC model. This means we can test whether a deliverable conforms to its contractual information requirements without interpretation, manual review, or custom scripts. It also means that the same requirements can be reused across multiple appointments and projects.
At DCW, I showed examples of how IDS checks can catch missing or malformed data early – before submission. The key insight: if we can define the requirement clearly, we can check it automatically.
Lessons from Automation in Practice
In applying IDS-based checking workflows, several lessons emerged:
- Specification quality is everything. Many models fail not due to poor modelling but because the requirements were vague, incomplete, or inconsistent. Writing good IDS files takes rigour – and sometimes negotiation.
- Instant feedback changes behaviour. When designers can self-check their models and see precisely what’s missing, the feedback loop shortens dramatically. QA becomes embedded, not reactive.
- Transparency challenges assumptions. Automated checks expose inconsistencies that manual review might ignore or interpret away. This can be uncomfortable – but it’s ultimately healthier for project outcomes.
Automation, when paired with clear specifications, doesn’t replace human oversight – it elevates it by making manual reviews more focused and strategic.
Reflections from DCW
The response at Digital Construction Week was encouraging. Attendees were especially interested in how IDS can serve both as a contract tool and a technical tool – linking the legal side of project delivery with the digital side.
Questions from the audience often centred on adoption: Who should write IDS specifications? How do we manage change across appointments? How do we ensure consistency across authoring tools? These are fair and open questions – and point to a growing maturity in the conversation around structured data.
The broader theme was clear: the industry is ready to move beyond fragmented checking toward integrated, spec-driven assurance.
Conclusion
Geometry still matters – but it’s no longer enough. As digital construction matures, structured data is becoming central to how we measure quality, compliance, and readiness. IDS offers a practical way to embed data accountability directly into model deliverables.
The real opportunity isn’t just in catching errors. It’s in shifting responsibility upstream, enabling earlier feedback, and building trust in data through transparent, repeatable, automated checks.
At DCW, we took a step forward in that direction. Now it’s up to us to make these tools part of our everyday practice.