How Eagle Mapping makes the point cloud the star of every delivery

How Eagle Mapping makes the point cloud the star of every delivery

Eagle Mapping has done photogrammetry and classification work in British Columbia since 1985, and has been flying LiDAR since 2007. Every delivery included the point cloud, but clients usually worked from contours and elevation models instead, because point cloud files were too large for their computers to open. Those point clouds are now hosted on LumiDB, where a 500 GB dataset opens in a browser with nothing to download.

Published on


Do not index
attachment:f2518c7b-6595-4c32-974b-8e0405a73903:image.png

We are trying to make sure the point cloud is the star of the dataset, and that everything else is a derivative of it.

notion image
Matt Aquin IT Manager
Eagle Mapping is an aerial survey and mapping company in Langley, British Columbia, flying its own aircraft over mining, forestry, engineering, and energy projects across North and South America. A delivery goes out as a bundle: the LiDAR point cloud, plus the DEMs, DSMs, contours, orthophotos, tile layouts, and boundaries built from it. Matt Aquin is the IT Manager there, and he led the search for a better way to hand off that data.
 

The challenge

The most valuable part of the delivery was the hardest one to open

The point cloud holds everything the survey captured, down to individual branches, power lines, and easement edges. Everything else in the bundle is a simplification, and it is the only file clients can use to build their own derivatives.
It was also the file clients struggled with most. The LAS and LAZ tiles were too large for their computers to open, so the company started cutting tiles down to 250 meters or smaller just to get them to load. The alternative was to generalize: send a DEM, a DSM, or contours and let the client work from those.
Derivatives are easier to work with, so clients defaulted to them, and months later the point cloud often had not been opened at all. One client built a DEM from contours, not realizing a DEM was already in the delivery, and not knowing that the ground classification in the LiDAR would have let them build exactly the model they wanted.
 
attachment:f2518c7b-6595-4c32-974b-8e0405a73903:image.png

Being able to look at full-fidelity LiDAR sure beats reducing yourself to staring at contours, or at an elevation model binned to five by five meter grids.

notion image
Matt Aquin IT Manager
 
Hardware is part of it. Rendering a point cloud at full density requires GPU and VRAM capacity that most client teams cannot justify buying, and prices have not been moving in their favor. At the same time, governments are putting LiDAR online in volume, through 3DEP in the United States and a province-wide collection in British Columbia. Both give you an index of tiles to download and no way to look at the data first. Aquin saw a service the company could offer on top of its own captures.
 

The solution

The client opens the point cloud before anything gets downloaded

Aquin evaluated several platforms first. Most have been in development for years and have feature sets to match--built for zooming down to ground level and walking or driving through a dataset, priced accordingly. That was more than Eagle Mapping needed.
Data residency decided it. Aquin did not want client data exposed to the US CLOUD Act, and LumiDB stores its data in Ireland, which met the requirement. Being able to move onto the company's own cloud later, if a project ever demands it, mattered as a backstop.
 
attachment:f2518c7b-6595-4c32-974b-8e0405a73903:image.png

Your option for bring your own cloud is a good contingency as well. Break glass in case of crisis.

notion image
Matt Aquin IT Manager
The point cloud now goes to LumiDB with the rest of the delivery. The client gets a link and a permission level, opens it in a browser, and moves through the whole capture at the density it was flown.
 
Contours drawn as vectors, built at Eagle Mapping's request.
Contours drawn as vectors, built at Eagle Mapping's request.
 

The results

Seeing the data no longer takes CAD software and a powerful workstation

 
attachment:f2518c7b-6595-4c32-974b-8e0405a73903:image.png

Being able to take a 500 gigabyte LAZ dataset and visualize it in a web browser is really cool.

notion image
Matt Aquin IT Manager
 
Clients can view the whole capture at full density, instead of relying on a five-meter grid or contour interval. All a client needs is a browser, so CAD licenses and graphics cards stop being a prerequisite for viewing their data. Clients who have received deliveries this way like having the data online.
It also increases the long-term value of a delivery. A dataset that stays accessible is one a client can add to, so a capture from 2014 and a refresh five years later sit in the same project. Aquin uses the compare slider to check one against the other, including calibration between sensors. For a company that reflies the same ground for the same clients, each delivery builds on the last.
 
attachment:f2518c7b-6595-4c32-974b-8e0405a73903:image.png

I would recommend LumiDB to anybody who has a lot of LiDAR collections, especially multi-sensor or multi-temporal.

notion image
Matt Aquin IT Manager
 
Hillshade rendering on a DEM, with light direction, sun elevation, and vertical exaggeration set in the browser. Also added at Eagle Mapping’s request.
Hillshade rendering on a DEM, with light direction, sun elevation, and vertical exaggeration set in the browser. Also added at Eagle Mapping’s request.
 

Want your clients working in the point cloud rather than a derivative of it?

We work with survey companies that want to hand over live, full-fidelity data instead of files their clients cannot open, and with teams delivering across repeat captures and multiple sensors.
Start a trial to see how LumiDB fits your delivery workflow.
 

Written by

Sampo Lappalainen
Sampo Lappalainen

CEO, Co-founder