From 40 TB to 10 TB: how CIMA+ shrank its point cloud footprint by 75%, without losing a single point

The problem

A pricing model that punished fieldwork, and a format they couldn't escape

Every reality capture project at CIMA+ starts the same way: scanners deployed at dozens, sometimes hundreds, of positions across a site, terrestrial units on tripods and mobile systems in motion. Each position produces a structured scan: its own file, its own panoramic imagery, its own overlap with the neighbouring stations.

Here's the trap. Their previous platform billed by the scan position. Not by value delivered, not even by data stored. A position weighing 100 MB cost exactly the same as one weighing 1 GB. And to keep the imagery matched to the point cloud, the platform required structured scans: the heaviest possible way to carry the data, locked in with no alternative. The invoice was chained to how the team worked in the field, and the format was chained to the platform.

The math was merciless: the more the team scanned, the better they did their job, the more they paid. Forty terabytes of structured data online, and every new campaign made it worse. Doing excellent work had become the most expensive thing the team could do.

The challenge, in five pain points

• Billed by the scan position

  Every tripod setup, every station was an invoice line. Costs tracked field effort, not business value.

• Format lock-in

  Structured scans were mandatory to keep images and points connected. The heaviest format in the industry, imposed by contract. Zero flexibility.

• Massive redundancy

  The same geometry was captured and stored from multiple overlapping stations, project after project, 40 TB deep.

• Heavy, slow sharing

  Reviewers and clients waited for transfers of structured data instead of opening a link in a browser.

• A model that couldn't scale

  With terrestrial and mobile campaigns already running and drone capture on the horizon, every new sensor meant more positions, more lock-in, more cost.

The workflow

Step by step

The entire conversion runs through VisionLidar desktop. No third-party tools, no format gymnastics. Here is exactly what a technician does, file in, file out.

1. 1

   Import the raw structured scans into VisionLidar desktop

   Drop the field data in as-is: FARO projects as .lsproj (the series of FLS station files) and E57 files from terrestrial, mobile and SLAM scanners. A structured scan means one file per scanner position with the panoramic photos embedded inside, and that's exactly what VisionLidar reads. No pre-processing, no conversion utility, no re-registration.

2. 2

   Open and validate the project

   The assembled project opens in the full 3D environment. The technician checks registration, colorization, station positions and the spherical imagery station by station. What you validate on screen is exactly what gets published.

3. 3

   Export one unified LAZ

   One export command merges every station into a single, deduplicated point cloud and writes it as compressed LAZ. Concretely: structured E57 / .lsproj in → one unified LAZ out. Overlap between stations is resolved; coordinates, density, intensity and classification are untouched.

4. 4

   Export the trajectory file with the JPGs

   In the same pass, VisionLidar extracts the panoramic photos from the structured files as standalone JPGs and writes the trajectory file: the exact coordinates and orientation of every scanner position. That pair lets VL365 re-attach each photo to its position in the unified cloud: click a station, stand in the 360° view, points perfectly aligned.

5. 5

   Publish to VL365

   Upload the LAZ + trajectory + JPGs to VL365 (unlimited upload, no per-gigabyte toll), and anyone with a browser can stream the project, stand in the 360° photos, measure, annotate, vectorize and export. No desktop licence required for viewers.

The results

The quarter of the weight, without compromising the data

Time

Publishing a project no longer means shuttling hundreds of station files. One LAZ, one trajectory set, one upload. Field-to-shared went from a background chore someone had to babysit to a routine the team barely thinks about. And reviewers open projects in a browser instead of requesting file transfers.

With 75% less data to stream, projects open faster in VL365, navigation stays fluid even on big sites, and jumping from station to station is near-instant. A quarter of the weight means the viewer has a quarter of the work to do, and the whole team feels it every day, on every project.

And notice the step that's missing: no mesh. No converter. No processing queue. Drag and drop into the browser, native Leica and FARO formats included, and the data is there. Usable, immediately. No downtime, nobody waiting for a platform to “prepare” the project. Engineering teams open it and work. That's the value.

Money

The old model billed by the scan position: every station, every setup, another line on the invoice, regardless of whether that position weighed 100 MB or 1 GB. The cost of hosting was welded to the intensity of the fieldwork. VL365 severs that weld. Scan as densely as the project demands; the platform doesn't count your tripod setups. Combined with a 75% lighter archive, CIMA+ turned a cost that scaled with their ambition into one that finally behaves.

Lighter data, lighter bill. And yes, lighter footprint

This isn't a green slogan; it's arithmetic. Every terabyte in the cloud is powered, cooled and replicated 365 days a year. Store 30 TB less and the bill follows, and so does the footprint: the cheapest, cleanest terabyte is the one you never store.

What CIMA+ gained

Beyond the storage cut

The migration was about terabytes. What the team discovered afterwards is what makes them stay.