BIM coordination and clash detection: the software finds the conflict, the meeting decides who moves

Who does it, and the mistake of assuming

The single most common failure is that nobody is clearly responsible. Contracts frequently refer to coordination without naming who runs it, who models what, at what level of detail, and who decides when trades disagree.

• The coordinator — often a BIM manager or VDC lead employed by the GC — runs the process and the meetings.
• Each trade models its own scope to the agreed level of development, and is responsible for its accuracy.
• The design team resolves conflicts that arise from the documents rather than from installation choices.
• The GC decides when trades cannot agree, applying the priority order in the execution plan.

All of that should be written down before the first model is exchanged, in an execution plan that is part of the subcontracts.

The execution plan is the whole job

• Which areas and systems are coordinated, and which are excluded.
• Level of detail expected from each trade, including whether hangers, supports and access space are modeled.
• Priority order — typically gravity systems and large ducts first, then plumbing, then conduit and small piping — so disputes have a default answer.
• Meeting schedule and attendance, with the requirement that attendees can make decisions.
• File formats, software and exchange method, agreed before anyone starts.
• Sign-off process, defining what "coordinated" means and who signs before fabrication begins.
• Change handling, since a single design revision can invalidate weeks of coordination.

Coordinate for maintenance, not just for fit

A model can be clash-free and still produce a building that is miserable to operate. Geometry says two objects do not touch; it does not say a technician can reach a valve. The items most often forgotten:

• Access space around equipment, panels, valves and dampers, including room for a person and a ladder.
• Filter and coil pull space, and the path to actually remove a component from the building.
• Ceiling access points, coordinated with the reflected ceiling plan rather than discovered afterwards.
• Hangers and supports, which occupy real space and are frequently left out of models.
• Insulation thickness, which turns a fitting clearance into a clash.
• Code clearances for electrical equipment, which are not negotiable regardless of what the model shows.

Coordination for contractors without a BIM department

Most construction companies in the country do not employ a VDC team, and the choice is not between full coordination and nothing. There is a middle ground that captures a large share of the benefit:

• Coordinate only the dense areas. Mechanical rooms, main corridors, ceiling spaces above lobbies and anywhere four systems converge. Most of a building never needs it.
• Hire coordination as a service. Independent providers run the process for a defined fee, which is often cheaper than the first mistake it prevents.
• Let the largest trade lead. On many projects the mechanical subcontractor already models their scope and can host coordination for the others.
• Use sections and overlays when full modeling is not justified. Two-dimensional overlays of ceiling systems still catch the obvious conflicts.
• Walk it before it is closed. A joint site walk with all trades before ceilings are enclosed remains a real, low-tech form of coordination.

What cannot be skipped, regardless of the tools, is the decision process: somebody must have the authority to say which system moves, and that has to be agreed before the disagreement happens.

What the owner gets out of it

Owners increasingly ask for models at handover, and contractors who treat that as a checkbox miss a commercial opportunity. What actually has value to the people who will operate the building:

• Accurate as-built geometry of concealed systems, which turns a future renovation from exploratory demolition into planning.
• Equipment data attached to the model, including model numbers, capacities and warranty dates.
• Maintenance access documented, so the facility team knows where the panels and valves are and how to reach them.
• A record of coordinated clearances, which protects the owner when a future contractor wants to run something through that space.

Discussing this early, and pricing it honestly, separates contractors who are chosen for negotiated work from those who are invited only to bid. It is also a legitimate reason for a higher number that an owner can understand and defend internally.

Running a coordination meeting that works

• Weekly, same time, area by area, following the construction sequence rather than the alphabet.
• Decision makers only. A participant who must call the office to agree to move a duct wastes the room.
• Resolve live. Move the element on screen and confirm it; do not leave with a list of unresolved items.
• Record decisions with area, system, what moved and who confirmed.
• Sign off by area, releasing it for fabrication, so progress is visible and irreversible.
• Escalate quickly. Conflicts that require design input should leave the room as an RFI the same day.

Getting it from the model to the field

Coordination that stays in the office changes nothing. The connection to the installation is what produces the savings:

• Field layout from the model, using total stations or layout tools, so hangers land where they were coordinated.
• Spool drawings and prefabrication derived from the coordinated model.
• Tablets on site with the current model, not a printed version from six weeks ago.
• A rule about deviations. Any field change to a coordinated element has to be reported, because it invalidates someone else's coordination downstream.
• As-built updates, which are what makes the model useful to the owner after handover.

The arithmetic on a mid-size project

Round numbers for a $22 million healthcare renovation:

• Coordination effort across all trades and the GC: roughly $95,000, including software, personnel and meeting time.
• Clashes resolved before installation: several thousand, of which perhaps 200 would have caused real field problems.
• Assuming those 200 average $1,200 each in rework, delay and change order cost: about $240,000 avoided.
• Prefabrication enabled by trusted dimensions: additional labor savings in the mechanical and plumbing scopes.

The numbers vary enormously by project type, which is why the decision should rest on system density rather than on a percentage of contract value.

Why coordination fails when it fails

• Starting too late, after fabrication has already begun.
• Trades modeling at inconsistent detail, so the comparison is meaningless.
• Sending representatives without authority to the meetings.
• Treating the report as the deliverable rather than the decisions.
• Ignoring design changes that invalidate completed coordination.
• Leaving supports, insulation and access space out of the model, then discovering them in the field.
• No consequence for deviating from the coordinated model during installation.

Where to start

If your company has never run coordination formally, start on one project with dense systems and one area of it — a mechanical room, a main corridor. Write a two-page execution plan naming the coordinator, the level of detail, the priority order and the sign-off rule. Hold four weekly meetings with people who can decide, and measure exactly one thing: how many conflicts were resolved on screen that would otherwise have been discovered by a crew with a ladder. That number, presented to your own project team, ends the debate about whether it is worth doing on the next job.

Frequently asked questions

Is BIM coordination worth it on every project?

No, and the honest test is density of systems rather than size or contract value. Strong candidates are hospitals, laboratories, data centers, industrial plants, multifamily buildings with tight ceiling spaces and anything combining structural steel with heavy mechanical work. Mid-size commercial projects with generous plenum space and simple systems are marginal, and small tenant improvements or simple wood-frame residential work rarely justify the effort. The deciding question is what happens when two trades meet unexpectedly in the field: if the answer is a few hours of adjustment, coordination is overhead, but if it means a delayed inspection, a cut structural member or reordered fabricated material, it is inexpensive insurance. Companies without a VDC team can still coordinate the dense areas only, or hire the process as a service.

Who is responsible for clash detection on a project?

It must be named in writing before any model is exchanged, because the most common failure in coordination is that nobody is clearly responsible. Typically a coordinator, often a BIM manager or VDC lead employed by the general contractor, runs the process and the meetings; each trade models its own scope to an agreed level of development and is responsible for its accuracy; the design team resolves conflicts that come from the documents rather than from installation choices; and the general contractor decides when trades cannot agree, applying the priority order set in the execution plan. That plan should also define which areas are coordinated, the file formats and exchange method, the meeting schedule, the sign-off that releases an area for fabrication and how design changes are handled.

Why do coordinated models still produce field problems?

Usually because the model was checked for geometry and not for use. A model can be entirely clash-free and still produce a building where a technician cannot reach a valve, pull a filter or open an electrical panel. The elements most often missing are access space around equipment, panels, valves and dampers, component removal paths, ceiling access points coordinated with the reflected ceiling plan, hangers and supports, insulation thickness and code-required clearances. The practical fix is to model clearance zones as solid objects so the software protects them automatically. The other frequent cause is field deviation: when a crew moves a coordinated element without reporting it, everyone else's coordination downstream quietly becomes wrong.