Drone Platform Considerations
Author: Lewis Graham
Published On: February 03, 2021
One important consideration when you decide to enter the drone LIDAR mapping business is "what drone should I buy/use?"
My advice is to consider your selection in the following order:
Products are the data you need to produce. Will you be doing wire collection or topographic mapping? Remember, if you try to do everything, you'll probably end up not doing anything particularly well. Try to focus on a primary product area.
After you know your product thrust, look for the sensor that will collect the source data you need to generate the products.
For example, if your primary application is bare earth volumetrics where you have no overhead or vegetation to contend with, why even use LIDAR? Just select a photogrammetry system. Of course, bare earth volumetrics might not pay the bills, so you probably also want to be able to collect topo data over vegetated areas. This will require LIDAR. How much vegetation? Are you working in areas with nothing but sagebrush or are you contending with the verdant kudzu jungles of the southeast? I'll write a separate note on LIDAR/3D Imaging Sensor categorization, but I generally consider these as falling into one of the below classes:
These sensors typically have network accuracies in the 10 cm (RMSE) or worse range. An application would be collecting assets such as what is attached to a utility pole, pole locations, attachment points and so forth.
"Utility" grade sensor
This will have network accuracy in the 5 cm range as well as "decent" vegetation penetration. It is an "everyday" sensor that is sufficiently inexpensive to consider multiple systems.
"High Density, Utility" grade sensor
These are used where you need to see through dense vegetation or perform wire detection but are satisfied with network accuracy and precision in the 5 cm range. Certainly dense vegetation topo (with 1 foot contour extraction requirements) and wire extraction fit this model.
"Survey" grade sensor
These sensors typically have network accuracy better than 3 cm, RMSE and low "noise" (i.e. high precision). You would use these in work that warrants this level of accuracy/precision such as hard surface highway work. Right now the only real choice is RIEGL with Optech soon to join the club with their new CL-360 scanner.
Ok, so now to the essence of this note. You know your products. This has driven your selection of sensor and now it is time to consider the drone itself.
Before I delve in to the details, think about your operational scenarios. Do you need to accommodate more than one sensor on the same drone (e.g. fly a True View 3DIS for some missions and an inspection camera for others) or are you looking at dedicated use? Do you need to transport the drone on a commercial aircraft to reach some of your target mapping sites? Is a one man carry packing case needed or will you always have at least two persons on site? Will you be doing work for a US government agency that allows only drones from an approved list? You get the idea - try to think of all of the operational aspects of where you will collect project data.
There are generally three types of drone platforms that can be used for aerial survey; rotary wing, vertical takeoff/landing (VTOL) and fixed wing. Here I am only discussing rotary wing since this is the dominate platform choice for LIDAR operations (since relatively high stall speed introduces a lot of mission planning complexities when considering VTOL or fixed wing solutions).
Consider Your Restrictions
The first consideration is restrictions placed on your decision by the sensor manufacturer. Some systems immediately dictate the drone. For example, if you are buying a DJI L1 for inspection work, you will have to couple it to a DJI M300 since this is the only drone that supports the L1.
I have to say, except for small photogrammetry solutions such as the Phantom 4 RTK, I am not a fan at all of integrated (monolithic) sensor/drome platforms (see Figure 1). The big issue here is that if you have a platform problem (bad Electronic Speed Controller, broken frame, etc.), you are out of business until the drone is repaired. Can you imagine the frustration of having a high end RIEGL scanner unusable because the drone is down?
The other problem with the higher end monolithic systems is transportation. They tend to be in very large cases that require two person lift. Maybe you can afford to put two persons on every job because that's what it takes to move the drone, but this is not a restriction I would want to place on my crew deployment!
Figure 1: Monolithic System
Consider the Requirements
Assuming you have some latitude in selecting the drone, you will have to consider required parameters to accommodate the sensor. Examples include:
- Customer Restrictions - some customers may have a preapproved list of drones, not allowing you to use anything not on the "white list." If the only drones on the list are not compatible with your sensor selection, ask your sensor manufacturer to assist you in working with the customer.
- Mass ("weight") of the sensor - the drone has to have sufficient lift capacity to carry the sensor!
- Flight duration when burdened with the sensor
- Field of View (FOV) considerations - for example, if your sensor has a 120° cross-track FOV, you will want to ensure that drone components such as landing gear are not in the active FOV
- Does your sensor require an external power source? If so, the drone will need to either carry that external source or be able to supply the correct auxiliary power.
Once you have the narrowed down your choices based on sensor, it is time to look at the remaining criteria.
- Ease of service - if the drone has to be factory serviced, is there a service depot at least in your country (I have heard horror stories of drones needing to be shipped from the USA back to Australia just for maintenance!)
- Are batteries readily available and for a reasonable price?
- Do I need to carry the system on a commercial airplane for transport to the survey site? Here look not only at the size/mass of the drone case but also the battery situation. For example, DJI uses 6 individual batteries in the DJI M600 Pro. Each battery is under the capacity limit set by FAA so you can carry multiple sets as carry-on luggage.
- Does it have a reasonable solution for a flight management system (FMS)? A custom computer housed in a pelican case looks very cool but is a real pain to manage on site. An iPad or Android tablet is much more convenient, especially if your pilot in command (PIC) is the only person who will be involved in the mission (since the PIC needs to be able to maintain line of sight with the drone while operating the telemetry station).
Depending on how handy your crew is, you will want to consider the availability of spare parts and what is end-user replaceable. For example, you can purchase spare arms for the DJI M600 Pro for a few hundred dollars each and easily replace these should you have an "incident" (and trust me, you will have incidents!)
While the above seems rather involved (and it is!), rest assured that when you discuss your sensor needs with us, we take a holistic approach and cover all of these "peripheral" details. If you have specific questions about platform choice, please do not hesitate to shoot us a question.