Gateway 3 isn’t just a milestone on the programme; it’s the moment the project must stand up to scrutiny with credible, retrievable evidence that the building has been constructed as designed and meets the specified performance. Reality capture — laser scanning, 360° imagery, photogrammetry and mobile LiDAR — has become the practical way to prove it. When it’s procured and integrated properly, it closes the gap between model intent and site reality, underpins the golden thread, and reduces the friction of late-stage queries. When it’s left to the last fortnight, it becomes another scramble that strains access, budgets and tempers.
TL;DR
/>
– Treat reality capture as a scoped deliverable from RIBA Stage 4 onwards, not a late add-on at completion.
– Specify accuracy, frequency, formats and metadata so point clouds and photos link directly to the asset information model.
– Appoint a lead for survey control and model alignment; unclear responsibility is the biggest source of unusable evidence.
– Use deviation analysis to drive corrective work before handover, not just to document what went wrong.
– Measure value by query response time, proportion of spaces verifiably captured, and the number of issues closed ahead of sign-off.
Specifying reality capture for Gateway 3 evidence
/> If the brief doesn’t spell out what “evidence” actually looks like, you’ll get pretty pictures and point clouds that can’t be relied on. Set out the scope in plain terms: which areas, how often, and for which purposes — dimensional verification, services routing, firestopping, façade interfaces, penetrations and shafts, ceiling voids, plantrooms and risers. Define acceptable modalities: static laser scanning for accuracy-sensitive zones, mobile LiDAR or SLAM for progress sweeps, and 360° imagery for context and visual proof. State expected tolerances for key packages, recognising that a plantroom verification has a different accuracy requirement to a corridor.
File formats and alignment matter. Mandate open formats alongside vendor-native: E57 or LAS for point clouds, IFC for model exchange, JPEG/PNG for imagery, and CSV for issue logs. Require registration to a shared survey control network with coordinates and level datums agreed early, and insist on a documented alignment to the federated model. Give the CDE location for uploads, naming conventions that relate to zones and levels, and metadata tags that connect imagery or scan data to asset IDs and room numbers. If your asset information model is in place, state how the reality-capture outputs must map to it.
Don’t forget the basic governance. Confirm who owns the data, who can access it, and how long it’s retained as part of the golden thread. Include minimum image quality and scan density guidelines, frequency of capture during fit-out, and the requirement for a deviation report against the design model with clear colour maps and issue pinning. Flag sensitive areas and privacy constraints, and align the method with site safety procedures.
UK scenario
A 22-storey residential block in Manchester is six weeks from completion. The main contractor’s project manager has a façade snag list to close, the MEP coordinator is chasing penetrations sign-off, and the clerk of works is challenging visual evidence for firestopping in concealed spaces. A scanning subcontractor is available two days next week, but the lifts are on test and access to risers is narrow. The BIM manager has a federated model, but grid control in the basement plantroom differs from the fit-out levels. Programme pressure is intense: bring trades back twice and the margin is gone. The team must agree a focused capture plan that proves compliance in risers, plantrooms, and sample flats, integrates photos and point clouds into the CDE, and triggers fixes before the façade mast climbers come down.
Interfaces and risk across design, trades and data
/> Reality capture touches design assurance, QA, and digital delivery. The easiest way to lose value is to leave responsibility floating between the surveyor and the BIM team. Name a lead for survey control and registration, and another for model comparison and issue raising; on smaller jobs this may be one role, but it must be explicit. Package your trade interfaces: MEP terminations and clearances, firestopping locations, structural openings, and façade tolerance checks. The outputs should drive an issues process with owners and dates, not just sit on a server.
Access and logistics are the next choke points. Scanning requires clear lines of sight and sometimes quiet floors. Coordinate capture windows with plasterboard closures, ceiling tiles, and door hung dates so you’re not trying to scan through finishes. For 360° imagery, plan routes and room states; photos are worthless if services are hidden and a firestop label isn’t visible. For external façades, confirm safe vantage points before access equipment is off hire.
Data integrity is a risk. Establish how new scans supersede previous ones, and maintain a consistent coordinate framework. If multiple suppliers are capturing data, set out who validates registrations and how discrepancies are escalated. Agree a naming convention that helps retrieval during Gateway 3 queries: zone-level-room, date, and capture type. For sensitive areas (security or privacy), be clear on masking requirements and who signs them off.
# Common mistakes
/>
– Treating reality capture as a marketing gallery and skipping the accuracy, alignment and metadata that make it evidential.
– Capturing once, late, and discovering clashes or missing firestops when trades have already left site.
– Assuming the point cloud and model will “just line up” without shared survey control and documented registration.
– Leaving photos in a separate drive with inconsistent filenames, making retrieval painful when a question arises.
– Capture and evidence checklist
– Scope zones with risk-based priority: risers, plantrooms, façades, penetrations, fire compartments.
– Confirm survey control and model alignment in writing; test a pilot area before full mobilisation.
– Lock formats and metadata: E57/LAS, IFC, JPEG; zone/level/room tagging and asset IDs.
– Agree capture cadence aligned to programme gates: pre-close, pre-ceiling, and pre-handover for each area.
– Define deviation reporting with thresholds and an issues workflow in the CDE.
– Set data ownership, access rights and retention as part of the information protocol.
Value measures that matter at completion
/> Measuring the value of reality capture at Gateway 3 is about proof and speed. Start with completeness: what proportion of high-risk areas has verifiable capture linked to the asset information model? Then look at issues closed: how many deviations identified by scanning were resolved before practical completion rather than after? Track retrieval time: how quickly can the team locate evidence for a specific room, penetration or asset tag when asked? If it takes days, your metadata and naming conventions need work.
Dimensional accuracy is another proxy for quality. For key tolerances — plant clearances, door widths, firestopping thicknesses — can you demonstrate compliance within reasonable bounds using the captured data? Deviation colour maps against the design model are useful, but they must be accompanied by a clear narrative of what was fixed and when. For façades, orthophotos from photogrammetry can provide a clear record of bracket locations, fire barriers at floor lines, and panel set-out, provided early capture is planned before access is removed.
A mature setup turns capture into decisions. The BIM or digital engineer compares scans to the federated model, pins issues to locations, and assigns them to trades in the CDE. The clerk of works and package managers validate visual evidence of remediations. By the time Gateway 3 questions arrive, the project can pull up a room ID, show the pre-close image, the post-install scan, the deviation report, and the sign-off — all in one chain.
What to watch next? Expect more clients to ask for structured, queryable evidence packs rather than bulky data dumps, and to test retrieval in progress meetings rather than at the end. The sharp questions for your next project review: Who owns survey control and model alignment? Which areas will be captured before they vanish behind finishes? How will a regulator or client retrieve proof for any room or asset within minutes, not days?
FAQ
# Who should be responsible for reality capture on a UK project?
/> On most jobs, the main contractor appoints a specialist surveyor for scanning and a digital/BIM lead to integrate outputs into the CDE. What matters is clarity: a named owner for survey control and registration, and a named owner for model comparison and issue management. Smaller projects may combine these roles, but responsibility must be explicit in appointments and programme.
# How often should scanning and 360° imagery be carried out during fit-out?
/> Frequency should follow risk and programme gates. High‑risk areas benefit from capture before closure and again at pre-handover, with interim sweeps where trades overlap. Low‑risk spaces can be sampled or captured once if the design is repetitive and access is easy. Agree the cadence in the information requirements and align it to key dates like ceiling closures and plant energisation.
# What file formats and standards should be specified to avoid lock‑in?
/> Ask for open formats alongside any vendor-native deliverables. E57 or LAS for point clouds, IFC for model exchange, and standard image formats with consistent naming and metadata are a pragmatic baseline. Specify survey control details and alignment documentation so future teams can reuse the data, and require issue logs in CSV or similar for portability.
# How is data ownership and privacy handled for captured site data?
/> Ownership should be defined in the appointments and information protocol, with rights for the client to retain data as part of the golden thread. Access controls in the CDE need to reflect sensitivity: some images may include occupants, security systems or personal items on retrofit jobs and require masking or restricted sharing. Ensure site inductions cover privacy zones and that capture teams follow the project’s photography policy.
# Can reality capture replace traditional inspections and sign-offs?
/> It complements, but doesn’t replace, competent inspections and package sign-offs. Scans and imagery reduce disputes and help verify that concealed works match the design, but someone still needs to interpret results and confirm compliance against specifications. The best setups pair capture outputs with targeted inspections and a structured issues process so findings lead to action before handover.
