Under-floor scanning is the process of capturing a three-dimensional point cloud of the space between the structural slab and the raised floor tiles using LiDAR or SLAM-based sensors. The scanner records millions of individual points, each with an X, Y, Z coordinate and an intensity value representing return signal strength. It fires a laser pulse and measures time-of-flight or phase shift to calculate the distance to each surface it hits. A tripod-based unit like the Leica RTC360 completes a full 360-degree scan in under two minutes, capturing up to two million points per second.
The coverage scale depends heavily on plenum height. A standard 24-inch raised floor provides technicians with enough clearance to belly-crawl a tripod scanner between pedestals, capturing setups every 8 to 12 feet. Plenums under 12 inches require SLAM units or crawler robots because a tripod simply won’t fit.
3D laser scanning provides a way to capture accessible underfloor conditions and transform them into measurable point-cloud data that engineering, BIM, VDC, and facility teams can use for planning and design.
What Is Beneath a Data Center Raised Floor?
The raised-floor system creates a hidden layer between the finished floor and structural slab.
Depending on the facility, this space may contain:
- Power cabling and pathways
- Network and communication cabling
- Cable trays
- Conduit
- Cooling infrastructure
- Piping
- Floor pedestals and supports
- Structural elements
- Equipment connections
- Other mechanical and electrical infrastructure
As data centers evolve, these spaces can become increasingly complex. New equipment may require additional cabling. Cooling systems may be modified. Infrastructure may be rerouted during renovations. Equipment may be replaced while older infrastructure remains. Over time, the conditions below the floor may no longer perfectly match the original construction documents. That is where existing-condition scanning becomes valuable.
Read More About – 3D Laser Scanning for Data Centers in 2026: AI Infrastructure, Liquid Cooling & Rack Mapping Guide
Why Existing Drawings May Not Show Current Conditions
A data center may have excellent original drawings, but the physical facility continues to change after construction. Consider a facility that has undergone several upgrades:
Original Construction ➡️ Equipment Replacement ➡️ Additional Cable Runs ➡️ Cooling Modifications ➡️ Infrastructure Rerouting ➡️ Current Field Conditions
The drawings available to the engineering team may represent only part of that history. Before designing another modification, teams need to understand what is actually there today. Instead of manually measuring individual components and relying on photographs alone, reality capture can create a measurable digital record of accessible existing conditions.
How 3D Laser Scanning Captures Underfloor Infrastructure
Underfloor scanning involves positioning reality-capture equipment so that the sensor can see the infrastructure beneath the raised floor.
A laser scanner measures visible surfaces and creates millions of XYZ points. Together, these points form a 3D point cloud representing the captured environment. Your existing draft correctly identifies point-cloud creation and registration as the foundation of the process. A typical workflow may look like:
Site & Safety Planning ➡️ Raised-Floor Access ➡️ Strategic Scanner Positions ➡️ 3D Data Capture ➡️ Point-Cloud Registration ➡️ QA/QC ➡️ Registered Point Cloud ➡️ CAD / BIM / Existing-Condition Documentation
Multiple scan positions may be necessary because cable bundles, supports, trays, piping, and other equipment can block the scanner’s line of sight. The objective is not simply to collect as many points as possible. It is to create sufficient coverage of the infrastructure required for the client’s project.
The Scanning Methods You’ll Actually Encounter
Tripod-Based Terrestrial Laser Scanning (TLS) – Overview
Terrestrial laser scanning uses a stationary, tripod-mounted laser scanner like the Leica RTC360 to capture a complete spherical point cloud from a fixed position. It’s the industry baseline for accuracy in confined mechanical spaces. Originally built for construction and forensic documentation, it’s become the default for under-floor LiDAR work where floor height allows.
Technical Specifications
A single RTC360 setup captures roughly 2 million points per second at ranges up to 130 meters, with registration accuracy around 2- 3 mm at typical plenum ranges. A dense underfloor scan of a mid-size data hall generates 8- 12 GB of raw PTX data before registration.
Best For / Use Cases
Facility engineers documenting cooling plenum baffle placement, cable tray layout for Scan to BIM conversion and structural pedestal spacing verification ahead of a floor load assessment.
Key Advantages
- Highest point density of any method on this list
- Built-in HDR imagery captures cable color coding alongside geometry
- Native registration accuracy suitable for sub-centimeter as-built drawings
- Works reliably in low-light plenum conditions without external lighting rigs
Compatible with standard point cloud registration software like Leica Cyclone
Limitations
- Requires enough plenum clearance to physically position a tripod, typically 18 inches minimum
- Setup time per station (60-90 seconds scan plus repositioning) adds up across large floor plates
- Struggles with reflective conduit surfaces that scatter the laser return unpredictably
When to Use Tripod TLS
Use this format when you need sub-centimeter accuracy, have more than 18 inches of plenum space, or require high-quality color imagery for cable identification.
SLAM-Based Handheld and Backpack Scanning
Overview
Simultaneous Localization and Mapping (SLAM) scanners like the GeoSLAM ZEB Horizon or Leica BLK2GO build a point cloud continuously as the operator moves through the space. There’s no tripod, no fixed stations, just continuous capture while crawling or crouch-walking through the plenum.
Technical Specifications
SLAM units typically deliver 8- 15 mm accuracy over a 20-30 minute continuous capture session, generating point clouds in the 3- 6 GB range for a data hall-sized plenum. Drift accumulates over long capture paths, so operators loop back to known reference points to close the trajectory.
Best For / Use Cases
Crawl spaces under 18 inches, plenums with heavy obstruction where tripod repositioning isn’t practical, and rapid documentation passes ahead of an emergency cable pull.
Key Advantages
- No minimum clearance requirement beyond what a crawling technician needs
- Captures continuous coverage without gaps between tripod stations
- Dramatically faster than TLS for large, obstructed floor plates
- Works in complete darkness with onboard illumination
Limitations
- Lower point density means fine details like small conduit fittings can be missed
- Trajectory drift accumulates without loop closure, degrading accuracy over long runs
- Less suited for deliverables requiring sub-centimeter precision
When to Use SLAM
Choose this when plenum height is too low for a tripod, when speed matters more than absolute precision, or when the space has too many obstructions for practical tripod repositioning.
Robotic Crawler-Based Scanning
Overview
Purpose-built crawler platforms mount a compact LiDAR unit or camera array on a wheeled or tracked chassis designed to fit under extremely low-raised floors. These platforms exist because human access simply isn’t possible in some retrofitted facilities with 8 to 10-inch plenums.
Technical Specifications
Crawler units typically deliver 5- 10 mm accuracy depending on sensor payload, with capture sessions limited by battery life and cable-tether length. A typical crawler pass covers 200-400 square feet per battery cycle.
Best For / Use Cases
Legacy facilities with sub-12-inch plenum heights, hazardous environments with standing water or exposed conductors, and situations where insurance or safety policy prohibits human entry.
Key Advantages
- Accesses spaces no technician can physically enter
- Reduces safety exposure in plenums with live electrical infrastructure
- Repeatable capture paths for before/after comparison during retrofits
Limitations
- Slower coverage rate per hour compared to handheld SLAM
- Limited maneuverability around dense pedestal grids and cable bundles
- Higher equipment cost per project relative to tripod or handheld methods
When to Use a Crawler
Choose this when plenum height falls below 12 inches, when safety protocols restrict human entry, or when a client needs repeat scans over time to track infrastructure changes without disturbing cabling.
Photogrammetry as a LiDAR Supplement
Overview
Photogrammetry reconstructs geometry from overlapping photographs rather than laser returns. On its own, it can’t match LiDAR’s metric accuracy in a confined, low-light plenum. Paired with a LiDAR pass, it fills in texture and label detail that intensity values alone don’t capture.
Technical Specifications
A supplemental photo pass adds roughly 200-500 high-resolution images per data hall section, aligned to the existing point cloud coordinate system for texture mapping.
Best For / Use Cases
Cable label verification, color-code auditing for circuit tracing, and stakeholder presentations where photorealistic texture communicates plenum condition better than a raw point cloud.
Key Advantages
- Captures readable text on cable tags and conduit labels
- Adds photorealistic texture for non-technical stakeholder review
- Relatively inexpensive to add to an existing LiDAR capture plan
Limitations
- Requires consistent lighting, which most plenums don’t have without added rigs
- Not a substitute for LiDAR’s metric accuracy on its own
- Adds processing time for photo-to-point-cloud alignment
Comparison Table: Underfloor Scanning Methods at a Glance
| Method | Typical Accuracy | Best For | Key Technical Advantage | Use When |
| Tripod TLS | 2-3mm | Static plenum documentation | Highest point density per setup | Open, accessible plenum with clear sightlines |
| SLAM Handheld | 8-15mm | Continuous crawl-space capture | No tripod setups needed | Tight clearance, obstructed sightlines |
| Robotic Crawler | 5-10mm | Sub-8-inch plenum height | Camera access below human clearance | Extremely low raised floor height |
| Photogrammetry Supplement | 10-20mm | Cable color/label verification | Captures texture LiDAR misses | Combined with any LiDAR pass for QA |
Data Centers as an Environment: Why Underfloor Scanning Matters
Data centers are expanding faster than almost any other segment of real estate and infrastructure investment, and the underfloor plenum is where that expansion pressure actually lands. Every capacity upgrade adds cable runs. Every cooling retrofit reroutes ductwork or reconfigures baffles. Every refresh cycle, now down to 3 to 5 years as hardware turns over faster, means new equipment gets tied into power and cooling infrastructure that was designed for a previous generation of the room.
None of that work can be planned safely against a floor plan that was accurate the day it was built and has been drifting from reality ever since. Facility teams end up making decisions about airflow, cable capacity, and load paths in a space nobody has fully mapped in years, which is exactly the gap underfloor scanning closes.
Why Choose Arrival 3D?
Experience Across Underfloor Environments
We’ve scanned raised floors in colocation facilities, hyperscale campuses, and 20-year-old enterprise data halls with plenum heights ranging from 8 to 36 inches. That range means we’ve built field protocols for nearly every clearance scenario a facilities team throws at us.
Right Technical Method for Your Plenum
We don’t force every project into the TLS tripod just because it’s our default. Plenum height, obstruction density, and safety conditions determine whether we bring a tripod scanner, a SLAM unit, or a crawler platform to your site.
Structured Project Management
Our PMP-certified project managers issue fixed-price proposals after a site walk confirms plenum access conditions. Most underfloor scans deliver a registered point cloud within 48 hours of the final field day, with quality assurance built into every registration pass.
Frequently Asked Questions
Can I scan an under-floor plenum without disconnecting live cabling?
Yes, LiDAR and SLAM scanning are entirely passive and require no cable disconnection or power interruption. Technicians work around live infrastructure the same way an electrician would during a routine inspection. The only requirement is enough physical clearance for the scanner or technician to move through the space safely.
What’s the technical difference between tripod TLS and SLAM for underfloor work?
Tripod TLS captures higher point density per setup, typically 2- 3 mm accuracy, but requires enough clearance to position the tripod at each station. SLAM sacrifices some accuracy, usually 8- 15 mm, in exchange for continuous movement through tight or obstructed spaces. Most large facilities end up using both methods across different plenum zones.
How much clearance do I need for a robotic crawler versus a human technician?
Crawler platforms typically operate in plenums as low as 8 inches, while a crawling technician with a SLAM unit generally needs at least 14-16 inches for safe, sustained movement. Below 8 inches, even crawler access becomes difficult depending on cable tray height and pedestal spacing.
The infrastructure beneath a raised floor can play an important role in data-center upgrades, retrofits, cooling modifications, and equipment changes. When that infrastructure is accessible, 3D laser scanning can transform visible underfloor conditions into measurable digital information that project teams can use long after field capture is complete. And when underfloor scanning is coordinated with white-space and above-ceiling reality capture, teams can gain a much more complete picture of the existing facility. Planning a data center renovation, expansion, or existing-condition survey? Arrival 3D can help determine the appropriate reality-capture approach and deliverable for your project.
