A point cloud is one of the most detailed records of an existing building, facility, or installation that you can obtain through 3D laser scanning. But collecting accurate data is only the first step. What happens with that data afterwards — how it is processed, delivered, and used — determines how valuable it will be for the project.
For some projects, the point cloud is all that is needed. For others, it becomes the basis for CAD or BIM modelling, design work, measurements, or long-term facility documentation. Choosing the right approach at the beginning can save time, avoid unnecessary modelling, and make the resulting data much more useful to the people who need it.
This guide explains what a point cloud is, how it is created, what you can do with it, and how to make sure the data you commission is fit for its intended purpose.
What is a point cloud?
A point cloud is a dataset of millions of three-dimensional points, each with precise XYZ coordinates in space. Depending on the scanner used, each point may also carry intensity data from the laser return and colour information from an integrated camera.
Together, these points form an accurate digital record of an object’s geometry at the moment of measurement. A dense industrial installation, a heritage building, a production hall — all of them can be captured as a point cloud with millimetre-level precision.
What a point cloud is not: an editable model. You cannot “move a pipe” in a point cloud or assign technical attributes to elements. It is a record of reality, not a design tool. That distinction defines everything that comes after.
How is a point cloud created?
Point cloud scanning — also called 3D laser scanning — works by firing a laser beam at a surface and measuring the time it takes to return. A 3D point cloud scanner repeats this operation millions of times per second, building up a dense representation of the surrounding geometry from each scanner position.
A single scan covers one viewpoint. To capture a complete object — especially one with complex geometry and many occlusions, like an industrial installation — multiple scan positions are required. These individual scans are then merged into a single registered point cloud through a process called registration.
The quality of the final dataset depends less on scanner range or resolution than on the number and placement of scan positions. A point cloud with complete coverage from many positions is dramatically more useful than a dense but incomplete dataset from too few. This is one of the most commonly misunderstood aspects of point cloud scanning — and one of the most important to get right when specifying a project.
For more on this topic, read: 3D Scanning Data Quality: Why the Number of Scans Matters
Point cloud formats: what you will receive and what to do with it
The format in which you receive your point cloud determines what software you can use it in and what you can do with it downstream.
E57 — an open format, recommended for long-term archiving. Compatible with most point cloud processing and viewing software. If you only keep one copy of your data, keep it in E57.
RCS / RCP — Autodesk’s native format for ReCap. Imports directly into Revit and AutoCAD. The right choice if your design team works in the Autodesk ecosystem.
LFM — used by AVEVA E3D and other industrial-grade CAD systems. Essential for projects in petrochemical, energy, and process engineering environments.
PTS / PTX — formats associated with older Leica scanners, still widely encountered. Require conversion before import into most design environments.
One practical rule: always archive in E57, regardless of your working format. Software changes — E57 will remain readable.
What can you do with a point cloud?
The answer depends on your project goal. A point cloud is not a single product. Depending on your project goal, it can be the final deliverable — or the starting point for further work.
Working directly on the point cloud — without modelling
For many use cases, the point cloud is sufficient as a final deliverable. Clash detection between a planned installation and existing infrastructure, dimensional verification, condition assessment, volume calculations — all of these can be carried out directly on the point cloud without creating a CAD model.
This approach is faster and less expensive than modelling. It is worth considering whenever the project goal does not require editable geometry. For a detailed discussion of when modelling is and is not necessary, see: Before You Commission a 3D Model — A Strategy That Saves Budget.
CAD or BIM modelling
When data needs to serve as the basis for a modernisation design, component prefabrication, or a digital twin, the point cloud becomes the starting point for modelling. An experienced engineer interprets the data and reconstructs installation geometry as editable objects in a CAD environment (such as AVEVA E3D) or BIM (Revit, AECOsim).
This is the most time-intensive and costly stage — but also the one that transforms raw data into a design tool. Point cloud modelling services cover this process end to end, from data receipt to model delivery in the required format and level of detail.
Mesh models
An alternative to CAD modelling, a mesh model represents object surfaces as a network of triangles. Meshes are generated largely automatically from point cloud data, making them faster and less expensive to produce than CAD models. They are well suited to visualisation, documentation of geometrically complex objects, and volume measurement. They are not suited to industrial installation design — they are not editable and cannot carry technical attributes. For a detailed look at what affects mesh quality and when mesh is the right choice, read: Mesh Models in 3D Scanning – What Affects Quality
Remote access via browser
Regardless of which downstream workflow you choose, the point cloud can be made available to all project participants through a browser-based access platform — so that designers, subcontractors, and maintenance teams can take remote measurements, review 360° panoramas, and verify installation details from any location, without specialist software.
At 3Deling, this is delivered through the Webpano platform.
Want to see how it works in practice? Watch a short Webpano demo — covering the key features from point cloud viewing and 360° panoramas to remote measurement and P&ID integration.
Software for viewing point clouds
If you want to work with point cloud data locally — outside a browser-based platform — you will need appropriate software.
Autodesk ReCap — the standard tool for point cloud work in the Autodesk ecosystem. Supports viewing, measurement, and preparation of data for import into Revit and AutoCAD.
Bentley Pointools — one of the most widely used tools for viewing and analysing large point clouds in industrial projects. Bentley Pointools View is available free of charge in its basic version and handles large files without performance loss — a practical choice for projects where files run to tens of gigabytes.
Leica Cyclone Register / Cyclone 3DR — Leica’s native software for registration and point cloud processing.
CloudCompare — free, open-source software for point cloud analysis and processing. Popular in academic environments and on projects with limited software budgets.
These tools solve one problem: how to work with point cloud data at a technical level. They require installation, specialist knowledge, and capable hardware.
Webpano — access for the whole organisation, not just specialists
Tools like Pointools, ReCap, and Cyclone are built for engineers and surveyors. They leave a different problem unsolved: how to give access to the same data to everyone else in the organisation — project managers, subcontractors, maintenance teams — without CAD licences, without installation, without training.
Webpano addresses this directly. It is a browser-based platform where scans, 360° panoramas, and 3D models are accessible from any location and device. A designer can take a measurement before a client meeting. A subcontractor can prepare a tender without a site visit. A maintenance technician can locate a valve and check its specification before entering a process zone.
Webpano does not replace CAD tools — it works alongside them. Data processed in ReCap or Cyclone is published to the platform and made available to everyone, not only those with a licence and the hardware to run it. A one-off scanning project becomes a long-term technical knowledge asset for the facility.
The most common mistake: data goes on a server and stays there
In many organisations, a point cloud reaches a network drive at the end of a project — and remains there for years, inaccessible to most staff, with no easy way to search it or take a measurement.
This is a waste of an asset that, with the right access infrastructure, could contribute to every subsequent project. Spatial documentation of an industrial plant is cumulative — its value grows with each project that uses it.
Planning a 3D laser scan of your facility, or looking for guidance on what to do with existing point cloud data?
Get in touch — we advise on the optimal processing and access strategy for your project and budget.
