Clash Review Agent — the BIM coordination AI agent that prioritises the three clashes you need to look at today.
The Clash Review Agent ingests Navisworks and BIM 360 / ACC clash reports with thousands of issues and ranks each clash by cost-of-late-discovery, trade dependency, and sequence-of-construction relevance. It turns raw clash sets into a short, prioritised list that coordination leads can act on inside existing workflows. It is available now as a Labs engagement under MSA + Appendix.
- Triaged clash report: a three-to-ten item shortlist pulled from a Navisworks or BIM 360 / ACC clash run that originally contained thousands of raw clashes.
- Per-clash trade-dependency call that names the owning discipline, the downstream blocked discipline, and any discipline that can safely defer action until a later phase.
- Coordination-cycle audit log that records what was triaged, what was deferred, and what was resolved between meetings, ready for JV-partner and contractor reviews.
What the Clash Review Agent does.
-
Clash-report ingest from Navisworks and BIM 360 / ACC
The agent reads native Navisworks NWD/NWF clash reports and BIM 360 / ACC Model Coordination clash sets as its primary inputs. It connects to your existing Navisworks integration or BIM 360 / ACC setup so there is no need to rerun clash detection or change how models are published. It treats each clash test, tolerance setting, and viewpoint as a first-class signal when building the prioritised list.
-
Prioritisation by cost-of-late-discovery
Each clash is scored on three dimensions: sequence-of-construction relevance (does this block the next trade on site?), discipline dependency (which trade owns the resolution and who is downstream), and downstream cost if missed. The Clash Review Agent behaves as The Coordinator, surfacing the three-to-ten clashes that matter this week and pushing low-impact penetrations or clearance issues into a deferred queue. Scores are calibrated per deployment so a hospital, data centre, or residential tower can each reflect their own risk profile.
-
Trade-dependency mapping across Revit-based federations
The agent reads the federated Revit model behind the clash report via the firm’s Revit integration to understand which discipline owns each clashing element. It groups clashes into the MEP lead’s queue, the structural lead’s queue, and the architectural-coordination queue instead of handing everyone the same 4,000-row spreadsheet. For MEP clash review workflows, it can further separate mechanical, electrical, and plumbing ownership so each sub-team works its own list.
-
Coordination-cycle tracking and audit history
For every coordination cycle, the agent writes decisions (triaged, deferred, resolved, re-opened) to a per-project audit log keyed to clash IDs and element IDs. Coordination meetings then walk the agent’s prioritised list instead of re-triaging from scratch in Navisworks viewpoints. The log is structured so JV partners, contractors, and insurers can see when a clash first appeared, when it was raised, and when it was closed, supporting claim-ready coordination history on long-running projects.
-
Support for MEP and structural-architectural coordination scopes
The trade-dependency rule pack can scope to MEP-vs-MEP, MEP-vs-structural, structural-vs-architectural, or full federated models. On projects where structural–architectural coordination runs in a different cycle from MEP, the agent maintains separate priority queues per scope. Each Labs deployment tunes naming conventions, workset patterns, and model-breakdown structures so the discipline mapping matches how your teams already organise Revit and Navisworks models.
-
Available now as a Labs engagement under MSA + Appendix
This roadmap-stage agent runs today as a co-built Labs engagement rather than a one-click product. VitruAI calibrates the prioritisation rule pack against your project types, trade-dependency conventions, and clash-detection platform in use, whether that is Navisworks, BIM 360 / ACC Model Coordination, or both. Studios that already run the Code Compliance Agent can reuse shared plumbing for model access, audit logging, and deployment under the same MSA.
Clash Review Agent — FAQ
-
What is BIM clash detection, and where does AI fit in?
BIM clash detection finds geometric conflicts between disciplines in a federated model, typically using tools like Navisworks or BIM 360 / ACC Model Coordination. The Clash Review Agent does not replace clash detection; it reads those clash reports and ranks clashes by cost-of-late-discovery, trade dependency, and sequence-of-construction impact. That means coordination meetings work the three-to-ten clashes that matter instead of paging through thousands of low-risk hits.
-
How does this work with Navisworks?
The agent reads native Navisworks NWD/NWF clash reports and their associated tests, tolerances, and viewpoints as input, using the same pipelines described in the VitruAI + Navisworks integration. It runs on top of the clash-detection runs your team already executes, so there is no change to how you set up tests or export reports. The output is a triaged list that can be reviewed alongside saved viewpoints during coordination meetings or exported into your issue-tracking system.
-
How does this work with BIM 360 / ACC Model Coordination?
For BIM 360 / ACC Model Coordination, the Clash Review Agent connects through the VitruAI + BIM 360 / ACC integration and reads clash sets via the platform’s API. It applies the same prioritisation rule pack used for Navisworks, but respects ACC-specific concepts such as coordination spaces, model versions, and issue links. Each Labs deployment agrees a canonical platform per project, and some studios deliberately run Navisworks on one asset type and BIM 360 / ACC on another while using the same prioritisation logic.
-
When will this ship as a productised release?
The Clash Review Agent is available now as a Labs engagement under MSA + Appendix, which means it can run against live projects that need triaged coordination this quarter. The public, productised Beta release sits downstream of this roadmap programme and will be informed by those Labs deployments. The waitlist you can join from this page is for that future Beta, while Labs engagements are set up directly with firms that have active coordination workloads.
-
How does this differ from the Code Compliance Agent?
The Clash Review Agent focuses on geometric conflicts between disciplines, working from clash reports produced by tools like Navisworks and BIM 360 / ACC. The Code Compliance Agent instead evaluates Revit models against external rule libraries such as local building codes, fire regulations, or internal standards. Many studios plan to run both: Clash Review for coordination between architectural, structural, and MEP teams, and Code Compliance for rule-based checks before permit submission.
-
Does it work for MEP-only coordination, or also structural-architectural?
The agent supports both MEP-only coordination and structural–architectural scopes, with rule packs that can target MEP-vs-MEP, MEP-vs-structural, structural-vs-architectural, or full federated models. For example, an MEP lead can run a focused MEP clash review cycle while a separate team handles structural–architectural coordination on the same project. Each Labs deployment configures which scopes apply at which project phases so the priority queues match how your coordination meetings are actually scheduled.
On the roadmap. Want first dibs?
The Clash Review Agent runs today as a Labs engagement against your project's clash-detection platform and trade-dependency conventions. The waitlist is for the productised Beta release downstream — Labs is for projects that need triaged coordination this quarter.
Join the waitlist