A superintendent in Columbus once told me, “We spent two Saturdays cutting and re-hanging duct because the riser was six inches off the wall we thought we had.” That job burned $28,400, three RFIs, and a week of float. The as-builts weren’t malicious; they were just wrong. And wrong is expensive.
The hidden bill for bad as-builts
Bad existing-conditions data doesn’t just frustrate teams. It multiplies risk across design, estimating, and construction.
- Design rework: Engineers add 10–20% safety to account for unknowns. Owners pay for the fat.
- RFIs and change orders: We routinely see 15–40 RFIs on mid-size renovations tied to missing or wrong dimensions. Each RFI averages $1,200–$2,500 in staff time before a single change occurs.
- Fabrication waste: One off-axis pipe rack or mis-located curb can scrap a full spool. A 200 LF misfit duct run can be $12,000–$25,000 in re-fab and labor.
- Schedule slip: A single uncoordinated chase or beam depth surprise can push drywall or MEP rough-in by 3–10 days.
On a $4.2M interior renovation in Cleveland, the drawing set showed 9'-4" floor-to-deck. Reality was 8'-11" under a truss bay. The team lost two weeks and $38,000 between re-design and rework. A measured model would have cost under $15,000 and been in hand in 10 business days.
“The six inches we didn’t account for cost two Saturdays and $28,400. The scan would have cost $9,200.”
>
— Superintendent, Columbus GC (2025)
Where the errors creep in
- Legacy PDFs with “field-verified where shown” scribbles that never got field-verified.
- Tenant fit-outs layered on top of base-building drawings without true elevations.
- Laser tape snapshots and cell phone photos stitched into guesswork.
- Missed slab steps, cambered beams, and bowed masonry walls that 2D plans flatten.
- MEP systems drifted off centerline after years of adds and repairs.
We see it weekly. None of it is malicious. All of it is preventable.
What “good” looks like: measurable, verifiable, shared
When teams start with a measured point cloud and a model tied to control, the ambiguity disappears.
- Capture at walking pace with a NavVis VLX3 for large interiors; augment with a Leica RTC360 or FARO Focus for tight tolerances and exterior tie-ins.
- Establish survey control. We set and shoot control targets and check critical tie points with a total station.
- Register cloud-to-cloud with control constraints, then QA against independent checks. Our interiors typically register to ±6 mm; point density lands around 5–10 mm at 10 m.
- Deliver in formats your team actually uses: E57, RCP, LAS, plus Revit models (LOD 200/300/400) and 2D CAD plans and elevations.
If you’re new to this, start with our primers: Building 3D Laser Scanning, Scan to BIM, and Scan to CAD. File type questions? See Point Cloud File Formats Explained: E57, RCP, LAS, and PTS.
A field-tested workflow (what we actually do)
- Pre-scan scope meeting: define tolerances, LOD, and deliverables. Set acceptance criteria.
- Control: set targets; tie to local or project coordinates. Verify with backsights.
- Capture: VLX3 walk-throughs for coverage; RTC360 static scans in congested MEP rooms and roof. Typical interior capture rates: 80,000–120,000 sq ft/day with VLX3.
- Registration: 1–3 days per 150,000 sq ft for QC’d registration, depending on complexity.
- Modeling: LOD 200 shell in 3–5 days for 100,000 sq ft; LOD 300 MEP in 10–20 days depending on scope.
- QC and delivery: viewpoints, section cuts, and clash-check snapshots included.
Turnaround examples:
- 120,000 sq ft office: 2 days capture, 3 days registration, 7 business days to LOD 200 model.
- 300,000 sq ft hospital wing: 4 days capture, 6 days registration, 20–30 business days to LOD 300 select systems.
The dollars-and-days comparison
| Scenario | Upfront Cost | Typical RFIs | Change Orders | Schedule Impact | Net Outcome |
|---|---|---|---|---|---|
| Rely on legacy as-builts | $0–$2k (site walks) | 20–40 | 3–8% of contract | +10–20 days | Higher contingency, more rework |
| Limited tape/laser measure | $3k–$8k | 12–25 | 2–5% of contract | +5–12 days | Gaps remain; surprises mid-build |
| Verified scan + model (VLX3 + RTC360) | $0.25–$0.60/sq ft all-in | 2–8 | <1–2% of contract | 0–3 days | Fewer unknowns, prefab confidence |
On a 200,000 sq ft adaptive-reuse project in Mansfield, scanning and LOD 200 modeling ($84,000 total) kept change orders under 1.1% and cut expected RFIs by 30+. Compare that to projects we’ve benchmarked without verified data, which typically land 3–6% in scope creep. Read the city-scale version in our case study: Downtown Mansfield: 180K Sq Ft in 3 Weeks.
A Cleveland story: the column that “wasn’t there”
We scanned a 1920s warehouse on the west side for an adaptive reuse into lofts. The paper plans missed a hidden steel column boxed in by shelving. The architect had a corridor laid out at 50 inches clear. Reality was 42 inches. We flagged it during registration review with a section cut at Station 12. Cost avoided:
- $19,500 in corridor rework and finish changes
- 14 days of schedule hit tied to door reordering
- Four RFIs that never had to be written
The scan captured 165,000 sq ft in two days with the VLX3, plus 42 static RTC360 scans in tricky spots. Registration wrapped in four days. The point cloud (22.4 GB E57) and LOD 200 Revit model went out on day 11. The developer didn’t love the surprise. They loved seeing it before demo even more. For similar projects, see Adaptive Reuse of a 1920s Mansfield Warehouse into Loft Apartments and Scan to BIM for Adaptive Reuse.
What it actually costs (and saves)
Rules of thumb we share with owners and GCs:
- Capture only (point cloud): $0.08–$0.15/sq ft, depending on complexity and travel.
- Capture + LOD 200 model: $0.25–$0.35/sq ft.
- Capture + LOD 300 architectural/structural + select MEP: $0.35–$0.60/sq ft.
Savings show up fast:
- Fewer site visits: architects using cloud viewers and slices cut travel by 50–80%. See How Architects Use 3D Scanning to Eliminate Field Visits.
- Prefab confidence: better dimensions mean more offsite work, fewer Saturday fixes.
- Estimating accuracy: quantities from the cloud reduce takeoff guesswork. See Preconstruction Estimating with Point Cloud Data.
On a $3M interior project, a $45k scan-to-LOD-300 package is less than 1.5% of construction cost. Cut change orders from 4% to 1.5%, and you’ve paid for the scan twice.
Set expectations in the spec
Bad scan scopes cause their own pain. Fix that in the RFP with five items:
- Accuracy and density: state registered accuracy (±6 mm interiors) and target density (≤10 mm at 10 m). Define what areas need higher density (e.g., shafts).
- Deliverables: list formats (E57, RCP, LAS), coordinate system, level naming, and sheet standards. For models, define LOD and modeled systems clearly.
- Control and QA: require control targets, independent checks, and a QA report with deviation heat maps.
- Access and sequencing: note areas that must be scanned before demo, overnight access windows, and safety constraints.
- Review checkpoints: add a 60% registration review and a 90% model review to catch misses early.
Two helpful references: How to Write a 3D Scan Deliverable Spec That Protects Your Project and the Scope of Work Template for Scan to BIM Projects.
When a full scan is overkill
Not every job needs full coverage. For simple TIs under 5,000 sq ft, selective measurement plus a targeted scan of problem areas can do the job. For historic preservation, we pair scanning with hand measurement and photographic detail to capture profiles and ornament the sensors can’t resolve. Reality capture is a support tool for the conservator and architect, not a replacement. If that’s your world, start here: Preservation and How 3D Scanning Supports Historic Preservation Documentation.
What ZEALOT brings to the table
- Tools that fit the site: NavVis VLX3 for speed; Leica RTC360 and FARO Focus for tight spaces and exterior control; total station for survey-grade tie-ins.
- Deliverables that drop into your workflow: E57, RCP, LAS point clouds; 2D/3D Floor Plans; As-Built Documentation; Scan to BIM at LOD 200/300/400.
- Registration and QC you can trust: control targets, cloud-to-cloud registration with constraints, deviation reports, and spot-checks logged by station ID.
- Turnarounds that protect schedule: 100,000–150,000 sq ft per day captured with VLX3; registration in 2–5 days for most buildings; models delivered in 7–30 business days based on LOD and scope.
- Ohio know-how: we’ve worked active hospitals in Columbus, plants in Mansfield, office towers in Cleveland, and distribution in Cincinnati. See what that looks like in our portfolio and our university sweep, Capturing a 12-Building University Campus in 9 Days.
The bottom line
You can pay for certainty once, or pay for uncertainty every week. We prefer the first option because it’s faster, cleaner, and cheaper — and it keeps your RFIs, change orders, and headaches where they belong: near zero.