Detecting clashes is the easy part: the software finds them in minutes. Resolving them is where programmes slip, because every fix needs someone with the engineering judgement to decide what moves, who signs it off and what else it disturbs. The strategies that work are the same on every scheme: classify clashes, filter out the noise, give each one a named owner and a deadline, fix it in the model, then re-test. Teams that run short of engineering hours at that point can embed engineers into the coordination workflow, and that is how we support UK consultancies at Xponexus Engineering.
TL;DR
- Clash detection tells you where the problem is. Resolution decides how to fix it and proves the fix worked, so budget time for both.
- Sort clashes into hard, soft and sequencing types, then rank them by consequence: structure and life safety first, cosmetic overlaps last.
- Every clash needs one named owner by discipline and a deadline, otherwise it circulates between meetings.
- Run a repeating cycle: federate, test, meet, fix, re-test. Prevention (modelling hard-to-move elements first) is cheaper than any fix.
- When coordination work outgrows your team, embedded remote engineers can take the calculations, drawings and BIM coordination without a permanent hire.
Table of Contents
- Detection and resolution are different jobs
- Start by classifying what the software found
- Prioritise before anyone starts fixing
- Give every clash an owner and a deadline
- A repeatable resolution cycle
- Reduce the number of clashes you create
- Speed up the loop with the right information exchange
- Where resolution stalls, and what to do about it
- How we help UK teams clear the clash backlog
- FAQ
- Sources
Detection and resolution are different jobs
A clash test is a rule run over federated models, the combined discipline models of a project. It returns a list. It does not tell you whether an overlap between a duct and a beam is a genuine design error, a modelling artefact or something the contractor will simply route around on site.
Resolution is the work that follows: reviewing each flagged item, deciding whether to move, resize, reroute or redesign an element, and then confirming the change did not create a new conflict somewhere else. One BIM clash resolution guide frames the difference as “where is the problem?” versus “how can the problem be solved?”, and notes a project can have excellent detection and still carry unresolved clashes.
That gap matters commercially. Industry estimates quoted by Symetri put the cost of each coordination issue rectified on site at between £3,000 and £4,500. Fixed in the model, the same issue costs a fraction in engineer time. The aim of every strategy below is to push fixes upstream, into the design stage.
Start by classifying what the software found
A raw clash report mixes three different problems, and each is resolved differently.
| Clash type | What it is | Typical example | Usual resolution |
|---|---|---|---|
| Hard | Two solid elements occupy the same space | A pipe running through a beam | Move or resize one element, or add a sleeve or penetration agreed with the structural engineer |
| Soft (clearance) | Elements do not touch but breach required clearance or access space | A valve too close to a wall to be serviced | Adjust position, or set the tolerance in the test and re-run |
| Sequencing (workflow) | Programmes conflict rather than objects | Drywall scheduled before the ductwork behind it | Change the programme, found through 4D BIM |
These three categories are set out in Procore’s UK guide to clash detection, which also notes that sequencing clashes only show up when the model is linked to the programme. Some teams add a fourth category, the duplicate: the same object modelled twice by different authors. Removing duplicates is a housekeeping fix, but it can clear a surprising share of a long list.
Write the tolerances down before the first test. A soft-clash test with no agreed clearance either floods the report or misses real problems. A clash matrix, one table listing each test, the models involved, the clash type, the tolerance and the priority, written once and re-run at every model issue, keeps the rules consistent. Imhora’s guidance on clash types and tolerances recommends running the highest-priority tests, those protecting structure and life safety, at every model drop.
Prioritise before anyone starts fixing
A first federated test on a medium-sized building can return thousands of items. Assigning them all at once guarantees that the important ones get lost. Three steps bring the list under control.
1. Group and filter
Grouping clashes by the element or system that causes them turns a thousand rows into dozens of decisions. Symetri describes scripting that groups clashes by the impacted element using Revit categories or Uniclass codes, which cuts the number of review tasks. Matterport suggests filtering out minor overlaps, which can reduce a list to around 150 high-priority clashes, though the reduction depends entirely on the project and the filters.
2. Rank by consequence
A workable three-level scale:
- Critical: conflicts with columns, beams, cores, foundations or fire-stopping strategy. These can change the structural design and its calculations.
- Moderate: services misaligned in ways that need a reroute but not a redesign.
- Low: aesthetic overlaps and items the contractor can resolve on site without cost or risk.
On the A1(M) Leeming to Barton scheme, the team tracked design clashes and construction and maintenance tolerances by level of significance, federating over 330 information models fortnightly. Autodesk’s case study reports that the process mitigated more than £400k of additional costs. Few projects are that large, but the principle scales down: focus the meeting on what is significant.
3. Apply the sequence
Sort by priority first, then by location, so each discipline can clear a zone at a time. Fixing structural clashes before service clashes avoids the common trap where a services reroute is made, then undone when the beam moves.
Give every clash an owner and a deadline
A clash with no owner is a clash that survives to site. Assign by discipline, following the logic Matterport sets out:
- Structural engineers resolve beam and column conflicts, foundation clashes and anything load-bearing.
- MEP designers adjust ductwork, pipework and cable routes.
- Architects deal with clearance, accessibility and design inconsistencies.
Where two disciplines are in conflict, the rule that saves arguments is simple: the element that is hardest to move stays, and the more flexible system moves. Procore’s UK guidance notes that main contractors typically populate structural components and hard-to-move items first, such as sloped drainage, so flexible systems are modelled around them. It also describes UK teams aggregating models in a common data environment early and assigning issues to design leads for prompt action.
For each issue record four things: the owner, the deadline, the proposed resolution and the status. The deadline is what most teams forget. A fortnightly cycle with a five-working-day return on critical items keeps pace with a typical model drop.
A structural engineer’s fix often has consequences beyond the model. Moving a column or adding a penetration through a slab changes the load path, and the calculation and the drawing have to follow. Keeping the two in step is the subject of our write-up on structural drawing and calculation coordination.
A repeatable resolution cycle
The most reliable pattern is a loop, not a one-off exercise. McBains, a UK consultancy, describes its approach in a BIMcollab customer story: upload the models, run automated clash tests between design portions, hold clash resolution workshops, resolve the issues in the models, then re-test. Its rulesets for clashes between disciplines are updated fortnightly, with progress reports from the BIM lead to project directors.
Translated into a routine any team can adopt:
- Federate. Each discipline issues its model to the common data environment on an agreed date, so the test runs on a consistent snapshot. Our guide to a BIM coordination workflow mapped to ISO 19650 covers how this fits the information management standard UK clients now expect.
- Test. Run the clash matrix. Group, filter and rank the results, as above. Our note on actionable Navisworks clash detection shows how we set up the clash groups.
- Meet. Hold a short coordination meeting for critical and moderate items only. Agenda, roles and follow-up are laid out in our piece on design coordination meetings.
- Fix. Each owner amends their own model, never another discipline’s, and notes what changed.
- Re-test. Run the same tests on the next drop. Close an issue only when it disappears from the report and no new one has appeared beside it.
The discipline here is closing the loop. A resolution that is agreed in a meeting but never verified in the next test is the most common source of a clash resurfacing weeks later.
The same logic applies outside engineering. A practical guide to managing a whole house renovation from brief to handover makes the point that a project is only as smooth as its decision cycle, and a clash log is simply that cycle applied to the model.
Reduce the number of clashes you create
Resolution is cheaper when there is less to resolve. Procore’s UK guidance argues that avoiding clashes saves more than detecting them, and that detection need not be all-or-nothing: model the congested areas such as corridors and plant rooms in 3D, and use 2D coordination to find where the congestion is. The practical habits that follow:
- Agree the model order. Structure first, then primary services, then secondary routes and fit-out.
- Reserve zones. Set aside space for risers, ceiling voids and maintenance access before anyone routes through them.
- Collaborate early. Contractor Jarvis lists early and frequent collaboration between architects, engineers, contractors and subcontractors, and prefabrication, as key clash-avoidance strategies.
- Check drawings before issue. Many clashes begin as inconsistencies between a structural drawing and an architect’s layout. A structured pass, such as our structural drawing checklist for UK projects, catches them before they reach the federated model.
- Plan the trades. Sequencing conflicts are avoided the same way on any site: by agreeing who works where and when. A practical guide on decorating project management shows how much finishing trades depend on that ordering.
Speed up the loop with the right information exchange
Tools do not resolve clashes, but they shorten the loop. In Stoke-on-Trent City Council’s Autodesk case study, a cloud common data environment and weekly face-to-face meetings over 3D models cut redesign time from three weeks to an average of three days. The council also reported cost savings of 3.2%, delivered on time and under budget. The lesson is not the specific platform. It is that issues were visible to everyone, annotated in the model and discussed against the model rather than against screenshots.
Whatever platform you use, check that it logs each issue with an owner, status and history. That log is your audit trail if a late change is later questioned.
Where resolution stalls, and what to do about it

The point at which resolution stalls on most UK schemes is not the technology. It is capacity. A structural engineer who is also finishing calculations, responding to RFIs and attending design team meetings cannot turn around thirty critical clashes in five days. The list ages, the next model drop arrives, and the new test returns the same items plus new ones.
There are three typical responses: stretch the existing team, recruit, or bring in extra capacity. Stretching the team risks the quality of the calculations. Recruiting is slow and creates a fixed overhead that a pipeline of uneven workload may not justify. Extra capacity that works inside your own processes, to your standards, and leaves the client relationship with you, is the response that matches a fluctuating clash backlog.
Our view on clash resolution
We see clash resolution as an engineering task wearing a coordination label. The software flags an overlap, but the decision about whether a beam can take a penetration, or whether a foundation detail has to change, is a structural judgement, and it has to be reflected in the calculation and the drawing, not just the model.
That is why we embed qualified civil and structural engineers directly into a client’s workflow rather than running a separate service alongside it. They work to UK standards and Eurocodes, they handle calculations, drawings and BIM coordination, and the outputs arrive review-ready. The client keeps direct control of the relationship with their own client.
Our advice to any UK consultancy: decide your priority rules and owners up front, keep the cycle short, and make sure someone has the hours to close every critical item before the next model drop. If that someone is not on your payroll yet, you do not need a permanent hire to get there.
Mohammed
How we help UK teams clear the clash backlog
We provide remote civil and structural engineering support to UK consultancies, construction firms and architects. Embedded engineers take on the calculations, technical drawings and CAD and BIM modelling that clash resolution generates, and a project can kick off within 48 hours. Work is covered by NDA and non-compete protection, and deliverables are aligned to UK standards and Eurocodes.

If your coordination meetings keep producing actions that nobody has the hours to close, see our CAD and BIM services, our structural engineering offer, or remote engineering support for dedicated overflow capacity.
FAQ
What is the difference between clash detection and clash resolution?
Detection is the automated test that finds conflicts between models. Resolution is the engineering work of deciding how to fix each one, amending the models and verifying the fix in the next test. A team can detect thousands of clashes and still resolve none of them.
Which clashes should be resolved first?
Start with hard clashes that affect structure and life safety, such as conflicts with columns, beams, foundations and fire compartmentation. Then move to moderate service reroutes, and leave aesthetic overlaps until last. Agree the ranking in your clash matrix before the first test.
Who is responsible for resolving a clash?
The discipline that owns the element that needs to change. Structural engineers handle load-bearing and foundation conflicts, MEP designers handle service routes and architects handle clearance and accessibility. Where two disciplines disagree, the hardest element to move normally stays and the flexible one adjusts.
How often should clash tests be run?
At every agreed model issue, commonly fortnightly on active design stages. Run the highest-priority tests at every drop and the lower-priority ones less often if the volume is too high to review.
Can remote engineers help with clash resolution?
Yes. Resolution produces revised calculations, drawings and model updates, all of which can be done remotely by engineers working inside your processes. We embed engineers in your workflow so outputs arrive review-ready and your client relationship stays with you.
Sources
- BIM Clash Resolution: Definition, Process & Methods
- Next Generation Clash Detection, Symetri
- Practical BIM Clash Detection for Constructible Projects, Procore UK
- Clash Types and Tolerances: Hard, Clearance, Workflow
- BIM Clash Detection: How To Avoid Construction Conflicts, Matterport
- BIM on the A1(M) Leeming to Barton, Autodesk
- Adding value in property and construction consultancy, BIMcollab
- Moving to BIM and collaborating in the cloud, Stoke-on-Trent City Council, Autodesk
- Whole House Renovation: Plan It From Brief to Handover
- Decorating project management: a practical guide


Leave a Reply