Why a reliable site record matters
Incident scenes can change within hours. Equipment gets moved, barriers go up, access routes reopen, and repairs begin. By the time a safety consultant or engineer returns, the space may look nothing like it did during the initial visit. A good set of photographs helps, but photographs do not always make it clear how one part of a site connects to another. That is where 3D documentation can make a difference.
Matterport and other 3D scanning systems let investigators preserve a navigable record of the location. Someone reviewing the case later can look around a room, follow an access route, or see how equipment was arranged when the scan was made. That added context is useful, especially for people who were not present during the field visit. A scan cannot tell investigators why something happened. It can, however, help them understand the physical setting in which it happened.
What Matterport actually captures
Not every system described as a “3D scanner” captures information the same way. Matterport is a platform for capturing spaces and turning that information into a digital walkthrough. The result depends partly on the device used. The Matterport Pro3, for instance, uses lidar to collect depth information and a separate imaging system to capture the appearance of the space. Scans taken at different positions are then aligned into a model that people can explore on a screen.
It helps to distinguish the different kinds of data involved. Lidar records distances to visible surfaces and can produce a point cloud: a collection of points located in three-dimensional space. A mesh connects geometric surfaces, while panoramic photographs provide much of the visual detail people recognize in a walkthrough. Measurements and annotations are additional features applied to that record. These outputs serve different purposes. A model may look convincing without capturing every surface well enough for precise measurement.
The Pro3 provides a useful example of why specifications need context. Matterport reports raw depth accuracy of ±20 millimeters at 10 meters under its stated test conditions (Matterport, “FAQ: Pro3 Camera”). That does not mean every dimension taken from a finished model will be accurate within 20 millimeters. Distance, scanning conditions, coverage, and processing all affect what can reasonably be measured. For an investigation, the team needs to know the difference.
Why spatial context matters to incident investigators
OSHA's incident investigation guidance recommends documenting a scene promptly, starting with the overall area and working toward the details (Occupational Safety and Health Administration). Three-dimensional capture fits naturally into that approach. It gives photographs, notes, and observations a shared spatial reference. A reviewer can see where a ladder stood in relation to a platform, or how a machine was positioned relative to a walkway, without having to piece the entire layout together from individual images.
Think about a machinery incident in a crowded plant. Close-up photographs are still needed to show the guarding, controls, equipment labels, and any visible damage. A scan adds a wider view: the machine's location, the available walking space, nearby storage, and the surrounding equipment. Those details may be relevant when a qualified investigator considers access, visibility, or the sequence of work. The scan provides context, not a finding. Interviews, procedures, equipment records, direct inspection, and expert analysis remain necessary.
Applications in construction and industrial safety
Industrial sites present particular challenges. A confined space might be reached through several doors, a hatch, a platform, and a narrow passage beside fixed equipment. A series of aligned scans can preserve that route and show where openings and equipment were located at the time of capture. It cannot show whether the atmosphere was safe, whether energy sources were properly isolated, or whether permit requirements were met. Those questions call for other records, instruments, and technical review.
Construction sites often need more than one capture method. Aerial photographs or drone mapping can show the larger property, roof areas, excavations, and staging locations. Ground-based scanning is better suited to hallways, stairs, mechanical spaces, and areas hidden beneath structures. Used together, these records help connect the overall site to specific work areas. Neither method captures everything, so the choice should follow the needs of the investigation rather than the appearance of the final model.
The same approach can help with crane, rigging, and heavy-equipment matters. Scanning may preserve the positions of nearby structures, travel routes, obstructions, and surrounding machinery. Those relationships are useful to see, but they are only one part of a technical review. Lift geometry, load charts, rigging condition, ground support, manufacturer information, and engineering analysis still have to be considered separately. Spear & Lancaster's work in crane and rigging safety, expert witness services, and incident investigations illustrates where a reliable site record can support a larger investigation (Spear & Lancaster, “Crane & Rigging”; “Expert Witness”).
Measurement accuracy is a project requirement, not an automatic feature
One of the easiest ways to misuse a 3D model is to assume that a displayed measurement must be exact. Scan positions, distance from the sensor, line of sight, reflective glass or metal, moving objects, and gaps in coverage can all affect the result. Thin features are a particular concern. A handrail may be clearly visible in a panoramic image while its geometry is only partially captured in the depth data.
For that reason, the team should decide early whether the scan is primarily for orientation or whether certain dimensions will support technical opinions. If an investigation depends on a clearance, elevation, or distance between two components, those measurements should be planned before the scene is captured. Additional scan positions, independent tape or instrument measurements, control points, or a survey-grade method may be appropriate. NIST has published work on evaluating terrestrial laser scanner performance for forensic use, reinforcing the need to understand the method rather than treating every scan as equally reliable (National Institute of Standards and Technology).
A manufacturer's accuracy figure is a useful starting point, but it is not proof that a particular site model meets that accuracy. The field record should identify the equipment used, the scanning and processing approach, any measurement checks, and the limitations found during review. That makes it much easier for another professional to decide what the model can and cannot support.
Scene changes, missing data, and the risk of false certainty
Timing matters as much as the scanning method. A scan records what was present when the team arrived. If a damaged machine has already been removed or a guardrail installed, the model documents those later conditions. It should never be described as though it shows the scene untouched at the moment of the incident. The record needs to include the capture date, known changes, and any areas the team could not access.
There are also details a walkthrough may miss. Small fasteners, worn labels, cracks, damaged cables, and surfaces hidden behind equipment may need dedicated close-up photography or hands-on inspection. Shiny and transparent materials can be difficult to capture accurately. A digital model can give an impression of completeness even when some areas are missing. Investigators should be clear about which conditions were observed, which were measured, and which were inferred later.
A defensible capture and handoff workflow
A useful capture plan starts with the questions the investigation needs to answer. The field team should document the site's condition and any changes that occurred before arrival. Wide photographs and overall scans establish orientation. More focused scan positions can then cover access routes, machinery, openings, and other important features. While on site, the team should also collect close-up photographs, field notes, identifiers, and independent measurements wherever those details matter.
The work does not end when the scanning is finished. Original files should be retained, along with the device information, capture dates, personnel, processing steps, and quality-control notes. The finished model needs a careful review for gaps, alignment problems, and features that should not be measured from the available data. A shareable walkthrough makes remote review convenient, but the team should also preserve the underlying records and explain any annotations or changes made during processing. Evidence retention and access should follow the requirements of the particular matter.
Depending on the scope, the handoff might include panoramic views, a navigable model, point-cloud data when the platform supports it, site photographs, measured details, annotated images, and a short explanation of limitations. A link to a virtual tour can be part of that package. It should not be the only record retained.
Where spatial capture fits alongside expert analysis
Spear & Lancaster describes incident investigation as a process of gathering facts, examining causes, and considering measures that could prevent a recurrence (Spear). A 3D site record belongs primarily in the fact-gathering stage. It helps reviewers understand the layout, locate observations, and communicate site conditions. It does not, by itself, establish the root cause or decide whether a standard was violated.
That distinction becomes more important when several specialists are involved. A safety professional may be focused on guarding and work practices, an engineer on equipment or dimensions, and an industrial hygienist on exposure conditions. Attorneys and other reviewers may need to understand how the documented spaces connect. They can all benefit from the same scan, even though each is asking different questions and relying on different evidence to reach a conclusion.
Preserving the site is the technical advantage
The real advantage of Matterport and 3D scanning is not the virtual tour itself. It is the opportunity to preserve a site in a form that people can revisit after conditions change. When that record is supported by photographs, measurements, field notes, and the original source files, it becomes much more useful than a standalone visualization.
SterFlies can provide the spatial documentation that helps investigators return to the scene and understand what was visible at the time of capture. Spear & Lancaster and other qualified professionals can then use that information alongside the rest of the evidence. Good scanning does not replace good investigative work. It gives that work a clearer picture of the place being examined.