What do a trillion LiDAR points look like?

What do a trillion LiDAR points look like?

A trillion LiDAR points captures extraordinary detail about the world around us. But what happens when the people who need that information can't easily access it? Sampo explores the growing accessibility gap in spatial data and how browser-based visualisation is helping make complex point clouds easier to explore, understand and use.

Published on


Do not index
Why the next step forward in spatial data is making the detail accessible to everyone who needs it.
Nobody commissions a LiDAR survey because they want a trillion measurements of a surface.
They want to understand something about a real place. The condition of a road, the clearance around an asset, how a construction site has changed.
The point cloud makes those things tangible. Yet when the results are delivered, the most detailed datasets can be the one the customer is least equipped to explore.
The surveyor has the tools, hardware and expertise to work with it. The person who commissioned the survey might have none of those things.
In some cases, they often end up relying on static images, simplified models or someone else to navigate the data for them.
Here’s the big disconnect. We capture granular levels of detail, but rarely consider how we can make the underlying value of that data accessible consistently across varying levels of end-user technical capability.
And the larger these datasets get, the more important that accessibility problem gets.

With great detail, comes great complexity

In June 2026, Leica Geosystems introduced a new range of terrestrial laser scanners, with models capable of capturing up to two million measurements per second.
At that maximum measurement rate, you could accumulate a billion points in just over eight minutes. That's not the time needed to produce a finished survey, but it shows how quickly the data adds up.
Now imagine what happens to those measurements afterwards.
Stored as a standard colourised LAS file, a billion points would occupy roughly 36 GB before compression. That's too large for an ordinary email attachment, can be slow to transfer and open, and is far beyond what you'd want to load into a browser all at once.
At the trillion-point scale, we're talking about roughly 36 terabytes of uncompressed data.
Compression can reduce the file size considerably, but it doesn't solve the fundamental problem. Nobody should need to load a trillion points just to inspect one building, a particular stretch of road or a small section of a construction site.
💡
It makes me wonder; how much value is being lost if the people who actually need spatial data can't quickly and easily explore it?

The problem with the first look

A quick example: a city planner has commissioned a detailed LiDAR survey.
The data is delivered, the technical work is complete, and the survey team has produced everything requested.
Now, the planner wants to inspect a particular street. Perhaps they want to understand the relationship between the road, surrounding buildings and vegetation.
What’s next?
They could contact the GIS team to open the dataset, navigate to the right location and prepare an extract. They might receive a static image showing the relevant area.
There's nothing inherently wrong with either approach.
Specialist teams are essential for processing, validating and analysing spatial data. But there's a difference between needing specialist expertise to interpret complex measurements and needing it simply to look at the results. All that time the planner spent just enabling their own access is - ultimately - lost time.
When information access has such a high barrier to entry, it's easy to imagine how often the original point cloud goes unexplored, and underused.

A point cloud shouldn't be difficult to look at

This is where browser-based visualisation offers compounding benefits.
The goal should never be loading a billions of LiDAR points into someone's browser all at once. That simply shifts the technical problem somewhere else.
Instead, a spatially indexed dataset allows a viewer to retrieve the relevant area at an appropriate level of detail. Additional detail can be streamed when it's needed. All that underlying data remains large and complex, but the experience of exploring it no longer will be.
It’s not radical thinking; we already do this with digital mapping apps.
Nobody expects to download every street, building and terrain feature in a country before looking up an address. We navigate to the location we're interested in, and get where we need to go.
I think we should expect something just as logical and intuitive from 3D spatial data.
Not because point clouds are simple (they're not), but because that complexity shouldn't be passed on to the person who wants to explore the results.

What changes when more people can see the data?

I’ve personally already seen what this can mean in practice.
⚙
In a pilot with the City of Helsinki's GIS Centre, traditional desktop workflows could take 10–15 minutes just to open large point-cloud files.
Moving to browser-based access made data exploration instantaneous, including for planners who didn't even use specialist point-cloud software.
The point isn't the improvement in loading time, but what becomes possible when accessing the data is no longer a substantial task in itself.
  • A construction manager can inspect the latest site capture during a meeting.
  • An infrastructure team can explore a surveyed corridor without first arranging a file transfer.
  • A city planner can look at the surrounding 3D environment before deciding whether further technical analysis is needed.
These aren't replacements for professional surveying or GIS analysis. The specialists still provide the accuracy, processing and interpretation that make the information trustworthy. Only now, more people can engage directly with the results.
That freedom opens the door for more conversations, questions, decisions and opportunities that might otherwise never have happened.

Visualisation is becoming part of the value of a survey

I think there's an big implication here for companies providing spatial data.
Historically, the emphasis has understandably been on the quality of the capture and the accuracy of the deliverables. Those things remain fundamental in the industry.
However, the experience of accessing the results matters too.
If a customer can easily return to a surveyed site, explore different areas and understand what has been captured, the point cloud becomes something they can use throughout a project.
It's no longer something they only encounter when a technical specialist opens it for them.
At LumiDB, this is one of the problems we're solving: making large spatial datasets available through browser-based visualisation while keeping them connected to the GIS and other professional workflows that depend on them.
The aim isn't to simplify the data itself or remove the need for specialist tools.
It's to make the information accessible at the level each person needs.
Because a trillions measurements should amount to more than an impressive technical achievement.
They should provide a clearer view of the world around us.

Want to see what your own point clouds look like in the browser? Try LumiDB.

Written by

Sampo Lappalainen
Sampo Lappalainen

CEO, Co-founder