How To Capture A High-Quality Point Cloud With Nexys

As SLAM-based scanning becomes more commonplace for surveyors and AEC professionals trying to capture accurate 3D models for digital twins, best practices for data collection can get lost in the weeds and new team members can struggle to understand SLAM best practices when capturing data in the field. This can lead to days of lost work and useless scans taking up space on the server. 

Whether you're mapping underground, indoors, or in challenging outdoor environments, Nexys' modularity and real-time data visualization make it easy to capture precise, survey-grade 3D maps. In this post, we're going to walk through some best practices from our engineers to help you optimize your point cloud captures, ensuring smooth, accurate data collection in any scenario.

Step 1: Prepare For Data Capture

Before the scan, you'll want to walk through the environment as much as possible to clear any debris or plan for your scan path and loop closure. If you're planning an autonomous flight with Nexys it might be impossible to observe the area you're looking to explore, but Nexys' rugged aerial autonomy can overcome dust and other obstacles to keep humans out of harm's way while generating a precise 3D map. 

Ensure that the area at the start of your scan is feature rich, with discernible objects the SLAM algorithm can use to locate itself in an unknown environment. To help with enabling zero drift, operators can install and shoot-in ground control points to further constrain local maps to a global coordinate frame. 

As SLAM (red) moves through a new environment it checks for familiar features (green) to help constrain a local map

With autonomous aerial missions, the robot will inherently return home and close the loop to help SLAM constrain a local map. You can think of closing the loop, or loop closure, like spotting a familiar landmark when driving while feeling lost. In that moment you suddenly understand exactly where you are in relation to where you're going.

SLAM platforms have historically struggled in situations where loop closure wasn't possible. But with the right constraints almost any SLAM-based data collection can eliminate or reduce any accumulated drift. If it's impossible to close the loop with a SLAM platform, ground control points can be used to constrain maps over much longer captures. 

Step 2: Start Scanning Mission

With our mission planned, let's initialize our payload. It's important to let Nexys initialize in a stable state to prevent any drift accumulation throughout the scan. 

Activate the Nexys scanner while placed on a stable surface near your starting area and allow 15 seconds for the entire system to initialize. You'll see a notification on the LED screen that ExynAI is up and running and ready to begin mapping. Make sure you’re standing behind the Nexys scanner during startup to give the cameras and LiDAR a clear view of its environment. 

Make note of the orientation of the scanner because you want to try and return to the same position to close the loop.

Step 3: Scanning With Nexys

After the Nexys system has fully initialized, you can freely move the scanner throughout your environment, making sure to pause or walk slowly through specific areas of interest. If carrying by hand, be sure to keep the scanner in front of you and pointed towards any object of interest. You can maneuver the Nexys scanner as needed to capture small areas, just remember that points closer than one meter are rejected. 

Example scanning path to ensure loop closure in a small multi-room environment.

It's also important to remember that your speed as you move through the scan area will influence the detail of the resulting scan. The slower you go, the more data that is collected, resulting in a more detailed point cloud. This is more of a variable if you're mounting a SLAM-based mobile mapper onto a vehicle. 

Step 4: Completing the scan

To complete the scan, you want to return as close as possible to the starting point. This final loop closure will help the post-processed SLAM algorithm further constrain the 3D point cloud for maximum accuracy. 

Tips for Getting the Best Mobile Scan With Nexys

• If you require colorization of the 3D point cloud, make sure the scan area is as well-lit as possible. If the scan is indoors or within a structure, turn on all lighting before the scan.
• When moving from one room or environment to another, it's important to let SLAM see where it's been and where it's going. When you're using Nexys by hand or with a backpack, you can pause in a doorway and turn sideways so the LiDAR sensor can see both rooms through the doorway. Pause for a few seconds and then continue your scan.
• Using the Nexys tablet, you can view real-time point cloud data as it’s captured to ensure you’re getting the scan results you need. This helps eliminate any wasted time or the need to perform a repeat scan.
• Moving slower within the scan area increases the point cloud density. The higher the point cloud density, the more detailed the final 3D map will be. 

Explore Enabling Zero Drift with Nexys

Equipped with Nexys, operators can quickly collect high-density 3D point clouds while reducing or eliminating accumulated drift through additional map constraints. Whether carried by hand, mounted to a backpack or vehicle, or powering an autonomous drone, Nexys is a rugged and reliable tool for capturing high-density 3D models for digital twins. 

Get in touch today to book a personalized demo and learn how to get more done with Nexys.