Last Wednesday, November 9 we hosted a webinar entitled “How to Make Survey-Grade 3D Models with the Elios 3 and GeoSLAM Connect.”
The webinar was one of our most popular ever, with almost 1,000 signups and the most questions ever asked at a single webinar we’ve hosted.
We were unable to answer all of the questions asked during the webinar, so we’ve put this article together to provide answers to all the questions we didn’t get to—scroll down to see all of the answers or skip around using the menu below to view answers by category. (By the way, we've changed the wording of some questions for readability.)
Here is a menu to help you navigate these questions:
- Questions related to our accuracy and precision testing
- GeoSLAM and data processing questions
- Elios 3 operations and best practice questions
- LiDAR-related questions
- LiDAR data collection questions
Want to learn more about the two tests we highlighted in the webinar?
In the webinar, we shared the results from two tests we recently completed for the Elios 3’s LiDAR data processed with GeoSLAM connect. View those test results in these two white papers:
- Testing Global Accuracy and Georeferenced Accuracy in 3D Mapping with the Elios 3 and GeoSLAM Connect
- Testing Local Accuracy and System Precision in 3D Mapping with the Elios 3 and GeoSLAM Connect
Want to watch a recording of the webinar?
Fill out the form on this page to view a recording of the webinar.
Questions Related to Our Accuracy and Precisions Testing
Q: Is the accuracy reported from the global accuracy and georeferenced accuracy test only valid when using the silver reference pads?
A: The silver referenced pads (or reflective targets) are not needed to achieve the accuracy reported. The accuracy results we presented are applicable to scans obtained without targets. The reflective targets are only used to have a repeatable feature on both the Elios 3 scans and reference scan so that the accuracy can be characterized, but the targets aren’t used to improve the actual accuracy of the Elios 3 scans.
Q: Is it possible to improve the accuracy, maybe with a scanner that has greater accuracy?
A: Future SW and HW upgrades are likely to provide further improvements, but we can’t commit on timelines for now. The latest Ouster sensor produces improved precision and range, and further GeoSLAM upgrades allowing for non-rigid registration of GCPs both could improve accuracy in the future.
Q: Is the accuracy error of Flyability due to the flight or the device capabilities?
A: The error is an accumulation of several factors:
- The sensor and mapping algorithms themselves.
- The drone’s flight and data capture (if the guidelines are well followed—in particular capture speed, coverage, and overlap of trajectory—the error will be reduced).
- The registration method to the reference model (e.g ICP alignment around take-off will be less accurate than using model-wide GCPs).
- The method to pick centroids of the targets used for error calculations.
Q: The flight time for the Elios 3 is 8.5 minutes but the accuracy results shown were based on three flights—so how is the accuracy based on one flight?
A: Each flight of 8.5 minutes covered the entire asset and generated a different scan, whose accuracy was then analyzed individually.
GeoSLAM and Data Processing Questions
Q: Can I create a textured mesh out of this data?
A: Not natively on GeoSLAM software, but other 3D manipulation tools can convert the point cloud into a mesh.
Q: Would it be possible to export a timestamped trajectory and point cloud to use inside TerraMatch for time based corrections?
A: Yes, a point cloud and timestamped trajectory can be exported. The data can also be reprocessed in GeoSLAM Connect.
Q: Can you combine your 3D model with picture recognition or tags?
A: Tags are being investigated at this time but we don’t have a specific timeframe for when this kind of feature might be available.
Q: After merging multiple flights within Geoslam Connect, can the file be exported to Autodesk Recap Pro? Or can Geoslam Connect mesh the point cloud?
A: The file can be exported to E57 for CAD import. Alternatively with GeoSLAM Draw, the point cloud can be meshed or exported to CAD formats.
Q: Is it possible to obtain data on displacements of production walls over time, how many pre-coordinated points should be?
A: GeoSLAM Connect does offer a Change Detection workflow to show the addition and removal of walls / objects in an environment. This can be aligned using manual alignment or using the target georeferencing workflow that will be available in the 2.3 version of GeoSLAM Connect in early 2023. If the target workflow is used, four targets are recommended as a minimum.
Q: Can I generate a colorized point cloud?
A: The point clouds you get from GeoSLAM Connect are colorized by the LiDAR intensity value, which is not real RGB color but allows you to make out some texture and shapes.
Elios 3 Operations and Best Practice Questions
Q: How can we receive training to develop skills to not only fly the Elios but to learn more about case use scenarios?
A: Flyability’s AAIM training covers specific use case scenarios. Learn more about and enroll in the AIIM training.
Q: Is there a way to use a tether system for recovery of the drone?
A: Except in silos and ore pass, we do not recommend using a tether on the drone because the risk of getting the tether tangled in the asset is very high. In our experience, the tether is usually not a good option.
Q: How does it behave in underground mines if it flies below sea level?
A: The drone behaves very well in underground mines. Regarding altitude, the drone can fly from -3000m (-9,842 feet) to 5000m (16,404 feet). However, in high altitude you need to take special precautions, as the air density is very low. See this case study for more information on flying at high altitudes.
Q: We want to map culverts, some of which are quite long. What is the max distance with the Range Extender? How far can the Elios 3 fly from the pilot?
A: It will depend on the size of the culvert and its shape. In straight lines, the drone can fly up to 500 meters (1,640 feet). However, symmetrical environments present challenges for the LiDAR technology. In a symmetrical sewer, there is a high probability that the GeoSLAM will not work very well.
Q: Is it possible to integrate another type of sensor into the Elios 3? Which protocol can I use?
A: We don’t offer this option at the moment, but the Elios 3 does have a second modular payload (in addition to the LiDAR payload) and we are working hard to offer new sensor integrations.
Q: Can I capture images with Flyability's drones? What are the images' resolution?
A: Yes, Elios 3 is a professional visual inspection tool, with powerful lighting and the ability to capture 12Mpx images.
Q: Can the drone be bought without the LiDAR payload?
A: Not at the moment. However, the LiDAR payload can be removed by the user for missions where the LiDAR isn’t needed or flight time must be maximized.
Q: What is the distance limit of the LiDAR?
A: The distance limit of the LiDAR is 30 meters (99 feet). But for high accuracy GeoSLAM mapping we recommend you fly at a max of 10 meters (33 feet) from the area/object you want to map.
Q: What is the file extension of the LiDAR?
A: The file extension is .bag. But after processing you can export the point cloud to i .las .laz .ply .txt and .p57.
Q: How many points per second does the LiDAR sensor create?
A: 300,000 p/s
Q: Is the lidar single or multiple return?
A: The OS0 is multi-echo and returns the strongest echo.
LIDAR DATA COLLECTION QUESTIONS
Q: How do you control the Elios 3’s flight speed during data collection to keep it constant?
A: There are different flight modes. If you stay in Assist mode (ATTI), the drone will not fly faster than 1.5 m/s (5 f/s), which is a good speed for capturing LiDAR data.
Q: About pipe inspections (very similar features in the travel), how can I prevent a wrong SLAM result? Can you give me some additional recommendations for best practices?
A: Pipes being very symmetrical, it can indeed be challenging to obtain a successful 3D model in a pipe using GeoSLAM, which only uses LIDAR, and thus requires some geometrical features to lock-on. Therefore, if some geometrical features are in the pipe (e.g. bumps, 20cm+ objects, or corners), keep them as long as possible in the field of view of the LiDAR. Since the field of view is wider on the sides of Elios 3, it may help to fly sideways.
Q: Would it be possible to control the scanner field of view used to create a point cloud in GeoSLAM Connect now or in the future? For example, if you only want to use the button 120 degrees from the scanner?
A: This has not been investigated yet.