Besides my robovac, I confess to knowing next to nothing about LiDAR. But don’t let that deter you from reading on! I promise I’ve brought in the big guns.
While I’ve been working in the Earth observation industry for 25 years, my area of expertise is multispectral and hyperspectral data analysis. So when someone asks for my advice on purchasing or processing LiDAR, I share my own reservations, and then defer to the experts. With that in mind, I’ve asked Reece Fraser from Neilly Group Engineering to contribute his thoughts.
Banner image from the National Ecological Observatory Network – NEON
Contents
What is LiDAR?
An acronym for Light Detection and Ranging, LiDAR is an active form of remote sensing. This means that the system sends out a pulse of energy (typically green or near infra-red light), and measures its return. Robotic vacuum cleaners use LiDAR to detect obstacles, avoid furniture, and create efficient cleaning paths. Self-driving cars use LiDAR in a similar manner on the streets. We even use this technology in high performance sports for tracking player movement as well as speed and trajectory of balls.
In the world of drone mapping, LiDAR is commonly used to create highly accurate 3D maps of the ground surface, even in areas covered by dense vegetation; to help archaeologists detect ancient structures hidden beneath forests, jungles, or desert sands; and to measure the height and biomass of vegetation. Bathymetric LiDAR can even map the seafloor!
Do I need experience to get started?
Reece Fraser is an environmental consultant with considerable expertise in hydraulics, UAV photogrammetry, GNSS survey, LiDAR, and geomorphology. He offers the following suggestions for the things you need to know before delving into the cost of LiDAR:
- How to setup an RTK connection from a GNSS base station to a receiver and/or UAV;
- How to do this in areas where there’s no mobile reception (i.e. no AUSCORS or NTRIP caster to provide base corrections);
- How to PPK data (photogrammetry or GNSS survey);
- How to convert between Ellipsoidal and various local geoids (e.g. in Australia – AUSGEOID09, AUSGEOID2020 etc) and what that means; and
- How to handle large point-cloud datasets – ground, vegetation, and building classification, manual adjustment of point cloud data (i.e. to line up with other datasets – horizontally and vertically), how to turn a point cloud into a DTM and DSM.
That’s a lot of acronyms! The key take away is that if you want to get the most out of your hardware investment, you need to also invest in your education. And be ready for a steep learning curve!
Is there special LiDAR software?
If you’re looking to purchase one of the DJI LiDAR equipped drones, you’ve probably come across DJI Terra. Reece stresses that that DJI Terra is NOT sufficient to process the data. DJI Terra does not do strip-alignment well.
Reece has been using TerraMatchUAV and TerraScanUAV for strip alignment. After extensive research, he suggests TerraSolid and LP360 are the only software to offer strip alignment well. RockCloud also does it – but using RockCloud has a minimum price-per-km2 that you have to pay to process the data – so as well as the capital costs of purchasing the UAV + LIDAR, you’re always paying to process the data. Both TerraScanUAV + TerraMatchUAV (as a bundle) and LP360 are not cheap – so check out your local distributor and factor that on top of the price to buy your LiDAR platform.
He notes that you’ll find yourself in a bit of a pickle if you go out and obtain a L2 and M350 and don’t factor in the requirement for strip alignment. However, DJI Terra (the software DJI provide, that you need to get the point-cloud out of the L2 or L1) is continuously improving and hopefully one day will do the strip-alignment right consistently. With recent firmware updates, he’s finding that he doesn’t need to do strip alignment in Terrasolid so frequently.
Can I map large areas with LiDAR?
Once you actually have the kit, it’s very tempting to scan large areas. Reece has three sets of batteries for the M350 (i.e. six individual batteries or three pairs) and that will allow about 100ha, on a good day. They have fitted-out a car to have a triple-battery-system which is capable of running the M350 battery charger so that they can continuously charge in the field while flying – further financial and time cost to implement. They have found that 3x battery pairs and this kit is sufficient most of the time; however, have been commissioned for some large projects (20km2) and deplete the triple-battery system in the car, even when the car’s running… The batteries are >$1K AUD each, and remember to double the price for a pair – not cheap.
Of course the area you can cover will depend on your flight altitude (read more about that here), and the type of drone and sensor you purchase. Here’s a great paper evaluating three different LiDAR systems to check out.
Do I need a license?
In some parts of the United States, you need to be a licensed surveyor to provide mapping services, so it’s a good idea to check your local regulations.
In many countries it is also possible to fly a drone that weighs less than 2kg without a license. However, all mapping quality LiDAR systems fly on drones that exceed this limit. In Australia for example, this means that you will need a remote pilot license (RePL), certified to 25kg, and must operate under a remote operator certificate (ReOC). Make sure that you take into account the cost of these certifications and ongoing compliance requirements when considering your budget for starting your LiDAR capability.
Final word
Every now and then I second guess myself when someone asks if I would like to get involved in a LiDAR project. But Reece reminds me of the complexity of building this capability and I’m happy with my decision to stick with multispectral and hyperspectral data! It’s a huge investment of time and money to do it properly. So if you are lured by the shiny drone, I encourage you to consider the full package before making your final decision. Perhaps even work with a local consultant like Reece to learn the ropes and better evaluate if you’re ready for the investment.
And if your goal is to create digital surface and digital terrain models, photogrammetry may be perfectly sufficient for your needs. Way cheaper too!
If you enjoyed this blog, you might also like:
