Drone Lidar Components Present Data Quality Challenges
The use of drone LiDAR requires significant training and expertise especially when using inexpensive systems.
Riegl LiDAR flying over Treasure Island, California. Provided by Jamie Young- PrecisionHawk
(Editor Note – This is a cautionary tale told in the author’s words about the perils of self-proclaimed drone lidar experts.)
Training is Needed
There are a lot of people with drones who are collecting aerial imagery and becoming self-proclaimed mapping professionals in the process, but in fact they have little, or no experience. These individuals then make the next step to drone LiDAR which is much less forgiving then imagery because the systems are much more sophisticated and require a deeper understanding of the technology.
This is a similar phenomenon to what occurred when manned LiDAR became prevalent in the mapping profession. Currently, it seems that everyone is buying a drone LiDAR system which is presenting consumers of LiDAR data with a number of difficult problems. Only making matters worse is the fact that many of these LiDAR sensors are made with less expensive components which also affects the quality of the data.
Recently, I had a discussion with a combatant, LiDAR “Genius.” Turns out he had little experience with the technology yet insisted that he knew much more about the subject. He was downright militant with me explaining that he was going to take a M600 DJI drone with its GPS and IMU components, attach a Velodyne laser and obtain centimeter accuracy. I did not agree.
The takeaway from this conversation is that the lack of understanding of LiDAR technology in the drone world is becoming alarming. LiDAR has been mainstream in the professional mapping profession for over 15 years having become commercially viable as a mapping technology roughly 22 years ago when the positional systems needed to geo-reference the laser were integrated with the required accuracy.
LiDAR representation by Echo using a Riegl MiniVUX -1UAV. Provided by Jill Wrenn -PrecisionHawk
Positional Accuracy Challenges
The positional system is the GPS and IMU components of the LiDAR system and they are greatest source of positional error in the LiDAR data. This is more pronounced with drone LiDAR systems because the IMUs are much less expensive which causes greater error (drift) as it relates to the accuracy of the data. Luckily, this can be somewhat offset by the flying altitude of these sensors. It should be cautioned that this doesn’t mean that careful planning is not required when using these systems.
Positional systems require smooth and continuous transition in their movement, so the multi-rotor, copter systems are not conducive to this type of movement. It is imperative to get the best positional system for these sensors as a result of this. A DJI M600 navigation and location system is less than adequate for cm positional and locational data. I have never known high accuracy map data to come from systems that can only provide a location within 4 meters and attitude information within a few degrees, much less if the attitude information is even recorded properly.
LiDAR surface model and points represented by echo of mounds in a LiDAR data set. Provided by Jill Wrenn – PrecisionHawk
To further complicate the situation, the laser technology used for most drone LiDAR are marginal at best and yield less than favorable results in most cases. Some of the issues that have been observed are the complete lack of understanding of how to operate a LiDAR sensor as it relates to alignment and initialization of the sensor. In most cases these sensor issues need to be more rigid with these less expensive positional and locational systems.
Drone Lidar from Riegl Performs
The lasers, with the exception of the Riegl drone LiDAR sensors, in most cases cannot achieve the desired results. The calibration of these systems is not completely understood as it relates to what the mapping professional expects. I have yet to see an intensity image even close to what is expected with the exception of the Riegl Drone LiDAR.
Reflectance data from a Riegl MiniVUX-1UAV. Provided by Jamie Young – PrecisionHawk
It should be understood that all LiDAR is not created equal and just because someone has a LiDAR it doesn’t mean they can collect accurate LiDAR data. There are drone operators that collect LiDAR and there are LiDAR operators that use drones. It shouldn’t matter what platform LiDAR is operated from if the company understands the technology. Professionals will know how to get accurate LiDAR data from any platform.
In closing, it is imperative that you research drone LiDAR companies before choosing which company collects your LiDAR. With the right company drone LiDAR can provide a wealth of information at a cost effective price.
James Wilder Young
PrecisionHawk Director Data Services – LiDAR Services
Note – If you liked this post click here to stay informed of all of the 3D laser scanning, UAS, autonomous vehicle and Lidar News. If you have an informative 3D video that you would like us to promote, please forward to editor@lidarnews.com.
Good top-level article on some of the pitfalls. As we discussed at a recent meeting of the Maryland Society of Surveyors, the real need is to understand 1) how accurate do you NEED your survey to be (requirements based), and 2) what accuracy is the TOTAL system you are using capable of (capability based). If you are using GCPs in conjunction with your airborne system you are adding significant accuracy. If you are using a $60K LiDAR vice a puck you are adding accuracy. As the article distinctly points out, you need to know the sources of errors and inaccuracies and account for them. Match your capability to your requirements. Nice job.
The quality of the sensor, the IMU, GPS and software used to process all play a major role in the quality of data achieved, not to mention how environmental conditions or elevation. Most of the higher end solutions will rely on Applanix for the IMU, which is the standard for most mobile mapping solutions, and I higher end sensor like Riegl or FARO. For example, the Stormbee drone uses Applanix and FARO on their customized customized drone with GNSS. I know GeoSlam is also claiming to deliver mobile lidar with cm accuracy but this must be tested. They do have the best SLAM in the business, my opinion only, but it will depend on the configuration and relative accuracy uncertainty of each of the components. A good practice is always to take reference measurements for validation, understand manufacture tolerances of the IMU, GPS, Lidar and possible imaging used. To be honest I am not very confident in the low cost solutions on the market today. You get what you pay for.
I find this article on the smug an arrogant side. Statement such as ” everyone is buying a drone LiDAR system” is out of touch with reality. And blanket opinions that only the most expensive equipment will suffice also ignore that state of the industry. I find it disconcerting and troublesome that he would make a blanket statement only Riegl LIDAR can provide good data.
Sorry, but this is a very one sided, opinionated and misleading article.