At the core of 3D SLAM mapping technology is an exported file known as a point cloud. A point cloud is created during the digital mapping process and is the foundation of creating an accurate digital twin. Fully understanding what a point cloud is, how it's created, and how best to capture one will aid surveyors and engineers throughout their data capture workflows.
Below, we’ll break down exactly what a point cloud is and how it’s created so you can get more out of this evolving technology.
A point cloud is a 3D representation of a physical space, such as a building or the inside of a mining cavity. It’s called a point cloud because the 3D representation is made up of millions of “points”.
Each point in the cloud represents a tiny piece of the physical environment being scanned. The data for each of these millions of points consists of X, Y, and Z coordinates along with the reflectivity and color.
When these points are combined, they create a full 3D representation of the space. You can think of a point cloud as similar to a printed newspaper image that uses tiny black and white dots of varying intensity to create an instantly recognizable image that’s full of detail.
A point cloud is similar except that each point represents a 3-dimensional location in space and is highly accurate. This allows you to zoom in or virtually move through the 3D point cloud as if it were the real environment.
The number of points within a cloud, referred to as point cloud density, impacts the detail of the resulting 3D virtual image. The more points in the cloud, the more detail.
Our Nexys autonomy and mapping payload creates survey grade accurate up to ±1cm point clouds that allow engineers, architects, geologists, and others to use accurate 3D models for highly detailed project planning, exploration, and management.
Point clouds are created by a technology known as LiDAR (Light Detection and Ranging). LiDAR is a type of computer vision and many of its features mimic human vision. Laser light emitted by the LiDAR system is reflected back to sensors which work like a human retina. Each point of light reflected and captured by the sensor becomes a point in the point cloud to make up the final 3D image. 
For mobile and autonomous digital mapping, LiDAR is combined with SLAM (Simultaneous Localization And Mapping).
With LiDAR as the vision system, SLAM is the “brain” that allows the mapping unit to know where it’s located and where it’s going by interpreting the LiDAR data in real-time.
The original way to create a point cloud was by using static scanners known as terrestrial laser scanners (TLS). These stationary tripod-mounted units are placed at precise locations and then scan a small slice of an environment.
Static scanning is accurate, but it’s extremely slow and tedious as the TLS needs to be moved and precisely placed to capture each new area.
Mobile scanning takes the standard TLS unit and incorporates LiDAR SLAM technology so it can capture a point cloud while freely being moved throughout any environment. This greatly speeds up the scan time as now an operator can simply walk through a structure and capture the entire point cloud.
One early drawback of mobile scanning is that it’s sometimes not as accurate as static scanning. But as ExynAI has matured as a SLAM solution across mining, construction, and industrial inspection, our Nexys payload can collect survey-grade accurate point clouds while on the go.
Part of this is done through our on-site-post-processing handled by the ExynAI technology stack. During this post-processing, any drift that was accumulated in the point cloud is corrected and aligned with any provided global coordinate frames.
The result is a 3D point cloud that perfectly aligns with any existing spatial data you already have, such as GPS or satellite imagery.
Autonomous scanning allows for mobile mapping in any environment without the need for a surveyor or drone pilot.
For example, our Nexys system can control an aerial drone and uses the same point cloud for autonomous exploration as it scans an environment. Our customers have even used that data to find instant return on investment during a single flight.
This is crucial for hazardous environments or situations where updated point clouds are needed regularly, such as in construction to measure progress.
When capturing point clouds with any mobile scanner, there is a trade-off between speed and accuracy. Since a scanner is essentially capturing data, the longer it captures data, the more accurate it is.
Think of it like studying for a test, the longer you study, the more detail and complexity you’ll understand about the subject.
With mobile scanning, this means the slower the the payload moves throughout it's environment, the more dense the point cloud will be.
To help you optimize this tradeoff, our Nexys system allows real-time viewing of the point cloud on the included rugged tablet computer. Surveyors can check the point cloud anytime and make any adjustments needed to capture the most accurate data possible.
At Exyn, we’re constantly developing new features and technology to improve the accuracy of point clouds captured by our Nexys system. These features allow you to use mobile mapping in situations previously only suitable for static laser systems.
This includes the use of available GPS resources, retroreflective coordinate markers, and advanced machine algorithms that constantly error-check the point cloud as new data is collected.
The best part is that these features are virtually all automated so you can focus on the surveying task at hand and not the underlying technology.
Contact us today for a personalized demo of our Nexys mobile mapping system to experience the speed and accuracy this technology can bring to your next surveying mission.
