Structural Sizing Agent — preliminary member sizing from the architectural model
The Structural Sizing Agent reads the architectural Revit model, applies the firm’s preliminary-sizing rules, and emits concept-stage member sizes — beams, columns, slabs — for early feasibility, BOQ, and architect-side coordination with the structural engineer. Detailed design and code-compliance sign-off remain with the licensed structural engineer of record. Available now as a Labs engagement under MSA + Appendix.
- Concept-stage beam, column, and slab sizes produced from the architectural model in hours instead of the typical week-long structural-engineer round-trip for early feasibility checks.
- Output feeds early BOQ work, pairs with BIM-to-BOQ automation, and frames the architect-structural coordination conversation before the first formal structural package.
- Detailed design and the engineering stamp remain the structural engineer of record’s responsibility — the agent shortens the architect-side feasibility loop without crossing the engineering-scope line.
What the Structural Sizing Agent does.
-
Reads the architectural model directly
Reads typical spans, floor plates, column grids, and assumed floor loadings from the Revit architectural model, including Generic Models and early massing where needed. For each Labs deployment, the structural sizing agent ingests the firm’s preliminary-sizing rule library — typical spans, span-to-depth ratios, and live/dead load bands drawn from historical projects and pre-design checklists. It aligns with workflows already used in structural sizing for preliminary studies.
-
Emits concept-stage member sizes with rule citation
For each beam, column, and slab, the agent emits a proposed concept size with the rule it applied, such as “primary beam — span 8.4 m, span-to-depth 1:18, depth 460 mm per firm rule SR-12”. Architects drop these members into a Revit Structure concept model for early coordination and BOQ. The structural engineer then takes the same sized concept as a starting point rather than sizing from a blank model, keeping detailed design and sign-off in their court.
-
Honest scope boundary — disclaimer in every output
Every report and exported model begins with the disclaimer: “Detailed design and code-compliance sign-off remain with the licensed structural engineer of record.” The agent only produces preliminary sizing for feasibility, early BOQ, and architect-side checks. It does not run full code checks like the Code Compliance Agent or replace the structural engineer’s design notes, calculation packages, or permit submissions.
-
Outputs to Revit Structure or IFC for structural tools
The sized concept emits as a Revit Structure model aligned with VitruAI + Revit or as IFC for hand-off into non-Revit structural tools via openBIM and IFC integration. The structural engineer imports the concept into their analysis or documentation environment, adjusts member sizes against their own design code checks, and overwrites any preliminary members that no longer fit once detailed loads and deflection criteria are set.
-
Supports architect-structural coordination at concept stage
Architects use the sized concept to check floor-to-floor heights, ceiling zones, and penetrations before formal structural packages exist, reducing late coordination clashes. The agent supports workflows described in structural-architectural coordination by giving both teams a common, traceable starting point. This keeps the early dialogue focused on viable structural zones instead of speculative assumptions that later drive redesign.
-
Available as a Labs engagement under MSA + Appendix
Labs deployments typically run 6–10 weeks under MSA + Appendix, with scope tuned per office and region. VitruAI works with the firm’s structural and design technology leads to calibrate the preliminary-sizing rule pack against a sample of historical projects, then runs a pilot on one live concept design. Each deployment ships a per-project accuracy and run-time report calibrated to the customer’s models and rule library, with typical passes landing in single-digit minutes once tuned.
Structural Sizing Agent — common questions
-
Does this replace the structural engineer?
No. The agent produces preliminary, concept-stage member sizes for feasibility, BOQ, and early coordination only. Detailed design, code compliance, and the engineering stamp remain with the licensed structural engineer of record, who still runs their own analysis and calculations. The value is shortening the architect-side feasibility loop, not removing structural-engineering scope.
-
What design code does the sizing apply?
Each Labs engagement calibrates the structural sizing agent against the firm’s preliminary-sizing rules, usually bands derived from past projects and pre-design guidelines rather than full code clauses. It does not perform full code-compliant detailed design or replace the engineer’s load-takeoff and analysis. The structural engineer of record still applies the governing design code and documents compliance in their own tools.
-
Does this output go into the structural engineer’s tool?
Yes. The agent exports a Revit Structure concept model or an IFC file that the structural engineer can open in their preferred analysis or documentation environment. In many offices this means importing into a Revit-based structural workflow; in others it means consuming IFC for an analysis model. Either way, the engineer treats the output as a starting point to refine, not as a signed design.
-
How does this differ from generative structural design?
Generative structural tools propose multiple structural options against an objective such as cost, weight, or embodied carbon, often exploring dozens of frames or grids. The Structural Sizing Agent instead applies the firm’s pre-agreed preliminary-sizing rules to a chosen architectural concept, giving one disciplined starting scheme. Both sit before detailed design, but in all cases the structural engineer of record signs and owns the final design and code-compliance documentation.
-
Who’s running it today?
The structural sizing agent is available as a bespoke Labs engagement under MSA + Appendix rather than a general product toggle. Current work focuses on calibrating against multiple firms’ preliminary-sizing rules and historical project sets to ensure the rule packs generalise. Broader productisation follows once several Labs cohorts have validated the workflow across different building types and regions.
-
How fast is the sizing pass?
Each Labs deployment ships its own benchmark report calibrated to the customer’s model sizes, rule complexity, and hardware. In practice, concept-stage sizing for a typical multi-storey architectural model lands in single-digit minutes once the rule pack is tuned. Reruns after architectural changes usually complete faster because the agent only re-evaluates affected spans and members.