Clash detection in BIM is the process of identifying conflicts between building systems inside a digital model before construction begins. When an architect, structural engineer, and MEP (mechanical, electrical, plumbing) designer each build their own 3D model, those models inevitably contain elements that overlap, intersect, or interfere with each other. Clash detection catches those problems on screen instead of on the job site, where fixing them costs significantly more time and money.
How Clash Detection Works
On any building project of moderate complexity, separate teams produce separate models. The architect models walls, floors, and ceilings. The structural engineer models beams, columns, and foundations. The MEP teams model ductwork, piping, conduit, and equipment. Each model is accurate on its own, but none of them “know” about the others.
Clash detection starts by combining these individual models into a single federated model, essentially layering every discipline’s geometry into one coordinated view. Software then runs automated checks across the combined model, flagging every location where elements from different disciplines conflict. The output is a list (often hundreds or thousands of items long) that a BIM coordinator reviews, prioritizes, and assigns back to the responsible discipline for resolution. After fixes are made, the detection runs again to confirm the clashes are gone.
The standard workflow follows six steps: create and aggregate the discipline models, run preliminary quality checks for compliance issues, execute automated clash detection, identify and prioritize clashes in a detection log, resolve each clash and track updates, then recheck the model to verify clean results.
Three Types of Clashes
Hard Clashes
A hard clash is a physical collision: two objects occupy the same space. A structural beam runs through the path of an HVAC duct, or a plumbing pipe passes directly through an electrical conduit. These are the most straightforward clashes to detect because the geometry literally intersects. They’re also the most expensive to fix during construction, often requiring demolition and rework of already-installed components.
Soft Clashes
A soft clash, also called a clearance clash, occurs when an element doesn’t have enough space around it to function properly. The object itself doesn’t intersect with anything, but a required buffer zone does. Think of a valve that needs 18 inches of clearance for a maintenance wrench, or a fire damper that requires a minimum access gap for inspection. Soft clash detection works by adding a virtual envelope around specified elements and checking whether that envelope overlaps with nearby geometry.
Workflow Clashes (4D Clashes)
A workflow clash is a time-based conflict rather than a spatial one. It shows up when the construction schedule puts two activities in the same location at the same time. If electrical conduit installation is scheduled during the same window as a concrete pour in the same area, crews will either block each other or one trade’s work gets buried under another’s. These clashes are called 4D because they add the dimension of time to the 3D model, linking each building element to a schedule task.
Who Manages the Process
The BIM coordinator is the person most directly responsible for running clash detection and managing the results. This role sits at the intersection of every discipline on the project. The coordinator sets up the clash matrix, which is essentially a grid defining which model categories get checked against which others. Structure versus MEP is a common pairing. Architecture versus MEP is another. Not every combination needs checking, and a well-designed matrix prevents the software from generating thousands of irrelevant results.
Once clashes are identified, the coordinator has to determine which discipline owns each conflict. That’s not always obvious. A duct running through a beam could be the MEP engineer’s problem (reroute the duct) or the structural engineer’s problem (relocate the beam), depending on project priorities and which element was placed first. The BIM coordinator triages these decisions, escalating unresolved disputes to the BIM manager or project lead. Quality assurance on the overall clash detection process typically falls to the BIM manager, who reviews the coordinator’s work for completeness.
Software Used for Clash Detection
Autodesk Navisworks has been the industry standard for years. It imports models from various authoring tools, combines them into a federated view, and runs interference checks across disciplines. Its strength is handling large, complex models with millions of elements. Autodesk Revit also includes built-in clash detection for projects modeled entirely within its environment, though it’s less suited for multi-software coordination.
Solibri offers both a desktop application (Solibri Office) and a cloud-based option (Solibri CheckPoint) that works with Revit and IFC models and connects to platforms like Autodesk Construction Cloud or Procore. Its rule-based checking goes beyond simple geometry conflicts to evaluate code compliance and design standards. For teams focused specifically on MEP coordination, plugins like BAMROC automate not just the detection but the resolution of mechanical-versus-structure clashes. Reporting tools like CRDR BIM streamline the creation of clash resolution documentation within Navisworks.
The choice of software often depends on the project’s authoring tools and collaboration platform. IFC (Industry Foundation Classes) files serve as the common exchange format when teams use different modeling software, allowing models from various vendors to be federated in a single detection environment.
Why It Matters for Construction
Without clash detection, conflicts get discovered when a plumber cuts into a beam or an electrician finds ductwork blocking a cable tray run. At that point, the fix involves tearing out installed work, redesigning on the fly, ordering new materials, and delaying downstream trades. Each of those steps costs money and pushes the schedule.
Clash detection compresses that discovery process into the design phase, where changes are made by editing a model rather than ripping out steel. The cost of resolving a clash in software is essentially the labor time of the designer making the change. The cost of resolving the same clash on site can be orders of magnitude higher, factoring in material waste, crew downtime, change orders, and schedule delays.
The process also forces coordination between disciplines that might otherwise work in isolation until their systems collide in the field. Regular clash detection sessions, sometimes called coordination meetings, keep teams aligned throughout design development rather than revealing surprises at the end.
Machine Learning in Clash Filtering
One persistent challenge with automated clash detection is the sheer volume of results. A federated model can generate thousands of flagged intersections, and a large percentage of those are false positives or low-priority items that don’t need human attention. Sorting relevant clashes from noise is time-consuming and relies heavily on the coordinator’s experience.
Recent work has applied machine learning to this filtering step. Researchers have developed plugins that extract metadata from clashes (element types, discipline, location, severity) and train classification models to replicate a BIM coordinator’s decision-making. The goal is to automatically separate true clashes that require design changes from false or trivial ones that can be dismissed. Testing on real BIM models, validated by industry professionals, has shown that these tools can reliably replicate how an experienced coordinator would filter and classify results. The approach works within existing BIM software environments, meaning teams wouldn’t need to overhaul their workflows to benefit from it.