BIM Coordination Software: Why Twelve Thousand Clashes Become a Meeting Instead of a Decision

Why coordination breaks even when the clash detection works fine

Wednesday, 9am, the weekly coordination call. The BIM manager has run the federated model overnight and Navisworks reports 11,842 clashes. Most are noise: a hanger passing through a ceiling void, the same duct hitting the same beam counted 40 times because the duct is drawn in segments, insulation intersecting a wall by four millimetres. Perhaps 400 matter. The team spends 55 minutes talking about level 3 mechanical against structure, agrees that the electrical contractor will drop the tray, and someone types that into the minutes. Next week the tray has not moved, the electrical modeller says nobody told him, and the minutes are in a folder nobody opens.

The tools in the room are good at what they do. Navisworks federates models from Revit, Tekla, Civil 3D and IFC exports and finds geometric interference reliably. Solibri applies rule based checking and catches quality issues that pure geometry misses. Revizto gives everyone a fast viewer and an issue list. Newforma Konekt covers issue tracking across firms. Each is competent. None of them holds the thing a coordination process actually runs on: a decision, with an owner, a due date tied to a procurement or fabrication milestone, and a record of which model version it was made against.

The gap costs real money in a specific way. A clash resolved verbally and not remodelled becomes a field conflict, and a field conflict on ductwork above a hospital corridor is a change order plus a schedule impact plus an argument about who owns it. Firms we work with can usually name three or four of these per large project. The frustrating part is that the clash was found. It was on the list. It was even discussed. What was missing was the mechanism that made someone accountable for it against a date.

Problem 1: 12,000 results is not a work list

Raw clash output has no relationship to how work is actually divided. Coordination happens by zone and by discipline pair, on a rhythm set by when each area needs to be signed off for fabrication. A results list sorted by clash test tells you nothing about which zone is behind.

Navisworks lets you group clashes and save selection sets, and skilled BIM managers do exactly that, by hand, every week. That labour is the product. It does not survive a model update, because the next run generates new clash identities and the grouping has to be redone or painstakingly matched. Solibri's rule sets are more durable but they answer a model quality question rather than a coordination workflow question.

What a custom build does: apply your firm's grouping rules automatically and persistently. Group by element pair type, by system, by zone and by level, then dedupe against a stable identity so that the same conflict recognised in last week's run is the same issue this week with its history intact rather than a new row. That single behaviour, persistent clash identity across model versions, is the thing off the shelf tools handle least well and it is what turns a list into a backlog. The output your coordination lead needs is not 11,842 clashes. It is nine zones with a count of open issues, an owner and a days to gate number.

Problem 2: models arrive at different maturity on different cycles

The structural model is at a level of development suitable for fabrication. The mechanical model is design intent with generic equipment. The electrical subcontractor has not uploaded since the last milestone. Coordinating those three as though they were equally mature generates thousands of false conflicts and, worse, gives everyone permission to dismiss real ones as premature.

Standard tools federate whatever is in the folder. They do not know that a given model is stale or that a given system is not yet supposed to be coordinated at this phase. So the BIM manager either coordinates against a model they know is out of date or spends the morning chasing uploads before the meeting.

What a custom build does: a model register that records, for each authoring firm and each discipline, the expected level of development by phase, the required update cadence and the last received version with its date. The federated run is assembled from declared versions, so every clash carries the exact model versions it was found against and any issue raised against a superseded model is flagged automatically for revalidation. Late uploads become a visible service failure rather than a rumour, which matters when a consultant later argues that coordination delay was not theirs.

Problem 3: the coordination decision has no home

A resolved clash needs four facts recorded: what was agreed, who is doing it, by when, and against which milestone. Meeting minutes hold the first and sometimes the second. Issue trackers hold an assignee and a status. Neither ties the resolution to the date it actually has to be done by, which is driven by fabrication release, procurement or the concrete pour, not by the coordination meeting cadence.

What a custom build does: attach every issue to a coordination zone, attach every zone to a gate date pulled from the construction schedule, and drive the whole view from days remaining. Assignment goes to a person at a named firm rather than to a discipline, because a discipline does not answer email. Sign off becomes explicit: a zone is coordinated when the responsible parties have each accepted it, with a captured snapshot of the model versions at that moment. If someone later modifies a model in a signed off zone, the system reopens the zone and tells the people who signed it. That is the control that prevents the most expensive category of failure, which is not an unresolved clash but a resolved one that got undone.

Problem 4: joint ventures need access without handing over native models

On a joint venture or a design and build with several tier one contractors, firms are contractually cautious about native model exchange. Sharing a Revit file is sharing intellectual property and inviting someone to edit your work. Sharing nothing means coordination happens in a viewer nobody trusts.

What a custom build does: standardise on IFC for geometry exchange and BCF for issue exchange, both open formats supported by the major authoring tools, and expose a browser viewer over derived geometry rather than native files. Permissions run by firm and by zone, so a subcontractor sees the federated context around their scope without downloading anyone's model. Issues round trip into each firm's own tools through BCF so modellers work where they already work instead of logging into a portal they resent. Firms that get this right stop arguing about model sharing and start arguing about the actual conflicts, which is the argument you want.

What this costs and how long it takes

Across the 2,000 plus projects Digital Heroes has delivered, this is the honest shape. A first release covering the model register with version tracking, automated clash ingestion and rule based grouping with persistent identity, assignment and zone level sign off gates, and a coordination dashboard runs $60,000 to $140,000 and ships in 12 to 16 weeks. A full platform adding a browser based federated viewer, two way BCF exchange, joint venture permissioning, schedule linked gates and owner reporting runs $160,000 to $400,000 phased over 6 to 12 months.

What pushes the number up: the browser viewer, which is the single most expensive component if you want smooth navigation of a large federated model on a laptop over site wifi. Deep integration with a specific authoring environment such as a Revit add in for round tripping issues. Point cloud overlay for renovation and retrofit work. And running clash detection yourself rather than ingesting results from Navisworks, which is a substantial engineering effort and rarely the right first move.

What keeps it down: ingesting clash results from the tool your team already runs, using IFC and BCF rather than proprietary formats, and starting with one live project as the pilot. A coordination platform designed in the abstract without a project pushing on it will be wrong in ways you cannot predict from a workshop.

Build versus buy, and when the existing tools are enough

Buy if you are coordinating a single building with two or three consultants and a stable team. Navisworks for detection, Revizto or Newforma Konekt for issues, a clear BIM execution plan and a competent BIM manager will handle that well, and the licence cost is trivial next to a build. Buy also if your firm's coordination process is genuinely conventional and you have no contractual complexity around model sharing. Do not build software to replace a process that works.

Build when two or more of these are true. First, you run several concurrent large projects and your coordination quality depends on which BIM manager is assigned. Second, your firm has coordination standards, naming conventions and grouping rules written down, and no tool enforces them. Third, you work in joint ventures where native model exchange is contentious. Fourth, you have had a resolved clash reappear in the field because a model changed after sign off. Fifth, the owner or main contractor wants coordination status reporting that you currently produce by hand each month.

The real argument for building is repeatability. A firm that coordinates well because of specific individuals has a hiring problem disguised as a technology problem. Encoding your model breakdown, your grouping rules, your level of development expectations by phase and your sign off gates into a system is how that knowledge stops walking out of the door, and it is why design technology leads sponsor these builds rather than project teams.

How to choose a developer for BIM coordination software

Ask them what BCF is and how they would use it before you explain it. If they do not know the BIM Collaboration Format, they will build a closed system and your modellers will refuse to use it because it forces them out of Revit.

Ask how they will keep a clash identity stable across model versions. This is the technical heart of the product and the answer should involve deriving an identity from the participating elements and their guids rather than from the clash test result index. A vague answer here means you are getting a to do list with a 3D picture attached.

Ask what they have shipped that renders large geometry in a browser. Federated models on a real project are heavy, and there is a large difference between a demo of a single Revit export and a viewer that stays usable on a site laptop with 12 disciplines loaded. Ask for the project and the model size.

Ask who owns the code and settle it in the contract before kickoff. You should own the repository, the cloud accounts and the right to hire anyone else to continue. At Digital Heroes the client owns the code from the first commit. Coordination standards are your firm's competitive knowledge, and encoding them into software somebody else controls defeats the purpose of the exercise.