What changed in reality capture this fall?
Three announcements in a few weeks tell the story.
- 360 video to 3D. Ricoh says it can turn 360-degree video into a detailed, measurable 3D model in roughly half a day, and Tokyo Electric Power Grid is using it to keep building records current after construction or equipment changes (pttl).
- Phones with survey control. DroneDeploy made Ground Pro generally available, turning a 360 walk into floor maps, ceiling maps and a measurable 3D model, and says a LiDAR iPhone or iPad Pro with RTK can tie a scan to survey control at about 1.5 inches (DroneDeploy).
- Scanning on the move. The NavVis MLX combines laser scanning and photogrammetry while the operator walks, cutting out the stop-and-set-up at every station (Foro3D).
If capture is cheap, where does the value go?
To the record. When anyone can make a 3D model, a model by itself is a commodity. The scarce thing is a record you can trust: measured to a known accuracy, tied to the right building, dated, and still true next month.
More capture without a system creates a new problem. Every walk makes another file. Files pile up in different tools, at different accuracies, and nobody knows which one is right. That's data entropy at a faster speed. The winners won't be the teams with the most scans. They'll be the teams with one living record that every new capture makes better.
Who sets the ground truth?
Skilled field teams. Look closely at Ricoh's own test. It judged gaps in the video models against reference laser-scan data, and the reviewers note that ranking hasn't been independently validated (pttl). Even the fastest video method is measured against a laser scan that someone set up, ran and checked.
That's the point. Fast capture tools are only as good as the ground truth they're checked against. A trained scan technician knows where accuracy matters, where a blank wall will fool software, and when a site needs a tripod instead of a walk. That craft doesn't get cheaper. It becomes the anchor for everything else.
How do fast capture and field crews work together?
As one system. We think about it in two layers:
Managed Capture
Our field teams scan the site and set the measured baseline, the ground truth every later capture is checked against.
Enabled Capture
Your store and facility teams take a phone video or photo in the Robotic Imaging mobile app to keep the record current between visits.
One record
Every capture lands on the same building in roboticimaging.ai instead of becoming another file.
Data Correction Layer
Keeps each record right as sites change, so the newest capture improves the record instead of competing with it.
The crew sets the truth. The phone extends its reach. People stay accountable for the measurement.
Does cheaper capture make field crews less important?
No. It makes them more important. When anyone can walk a building with a phone, the number of captures goes up and the number of trusted captures doesn't. Somebody has to decide which capture is the reference, catch the drift between versions and know when a site needs real survey work.
That is a skilled job. It takes knowing how light, glass, blank walls and moving people fool software. It takes knowing which dimensions a design team will build from and which ones only need to be close. And it takes someone willing to put their name on the measurement.
We build for those crews, and for the store and facility teams who now share the work. A store manager filming a quick walk after a reset is doing real documentation. A scan tech setting the baseline is doing the craft that makes that walk worth something. Both deserve tools that respect their time.
What should owners ask before buying more capture?
- What is this capture checked against, and who checked it?
- Where does it live after the project ends?
- Will next month's capture update this record, or start a new one?
- Can my team ask a question across every site, or only open one file at a time?
Hardware will keep changing. Cameras, scanners and phones get better every year, and we welcome all of them. What should not change is the record underneath. Speed from capture to answer only matters if the answer is right.
See how it works on the Enterprise page, or open roboticimaging.ai.
Frequently asked questions
- Can 360 video replace a laser scan?
- For many updates, video and phone capture are fast and useful. A laser scan by a trained field team still sets the measured baseline those captures are checked against.
- What is ground truth in reality capture?
- A measured capture of the real site, at a known accuracy and date, that other data is checked against.
- What is the difference between Managed and Enabled Capture?
- Managed Capture is Robotic Imaging field teams scanning a site. Enabled Capture is your own team keeping it current with phone video and photos in the Robotic Imaging mobile app.
- Where does each capture go?
- Into one record of the building in the roboticimaging.ai web app, so every new capture updates the same record.
Your buildings. Fully documented.
Book a capture