During senseFly’s recent Drone Surveying 101 webinar, we dedicated more than 20 minutes to an open Q&A session on using survey drone technology, with the answers supplied largely by our two client presenters: Marc Cañas of Jacobs and Aidan O’Connor of ASM Ireland. This article details their responses in full, spanning topics such as: image overlaps, insurance, resistance to UAV technology, drone vs. LIDAR point clouds and more.
Here, in part 1 of this exclusive two-part feature, our presenters answer the following questions (click a question below to go straight to its answer):
- What’s the best overlap for stitching drone pictures?
- What about insurance for drones?
- Did you encounter any resistance within the profession to using survey drones?
- Can you put UAV use into perspective versus LIDAR aerial surveying?
- Aidan, why did you use GCPs when the eBee RTK doesn’t require these?
Marc Cañas, Jacobs: That depends on what your output is—if you’re trying to get a high resolution digital ortho model or if you’re just looking at imaging.
What we’ve found is that if it’s just imaging and you’re not really concerned too much with the underlying data—you just want a great orthomosaic—then you could really go down to a 60%-40% lateral on that, and that would be just fine.
60% / 40% … works actually quite well for orthomosaics
But if you’re looking at getting some higher accuracies in your digital surface model and digital terrain model—depending on the terrain and depending on if it’s urban or hilly, or whatever it might be—you may need to increase that upwards of 75%.
In general, 60/40 for easy terrain, which is actually the default that the eBee defaults to, works actually quite well for orthomosaics.
Brock Ryder, senseFly: That’s an interesting question. We actually have solutions in the U.S. now, in North America, for insurance [more]. The market is definitely maturing out there. We have many discussions with insurance companies and I think it was an initial step by some to say, “OK there is a market here, we want to provide this”.
In the early days, I know with Aidan, it was tough to get companies to insure you for this type of work, but now we’re certainly seeing a drive and an interest from many insurance companies and I would expect that to expand worldwide quite quickly. As I said, we have a plan in action for North America now, for the U.S., and we are looking at other options worldwide as well.
Aidan O’Connor, ASM Ireland: I’d just like to add that I think the U.S. is kind of where Europe was maybe four or five years ago. We found it very hard to get insurance in the early days and it took weeks on the phone trying to get anybody to insure us. Now it’s the case that people are filling out the market, falling over each other trying to get your business. It has just exploded in the last two years, absolutely exploded. It just depends on where you are really.
Marc: If I could just jump in really quick, this is an important question, especially when you’re dealing with agencies in the U.S. specifically. Agencies, or owners, or even large companies such as Jacobs, they will insist that you have insurance. There are many insurances available, even from last year to this year I’ve seen tremendous availability of folks out there insuring. We attended the Commercial UAV Expo this year in Vegas and I saw five to seven companies offering insurance where last year we only saw one. So, I think the trend is changing.
Aidan: Resistance is not really the right word. They effectively do different jobs and generally traditional surveyors are more keen on surveying smaller areas to a higher degree of accuracy, millimeter accuracy. What we’re generally doing is surveying larger areas to kind of maybe a 50 to 60 mil accuracy. So they effectively complement each other. For a lot of projects we go in as the first point of call.
What we’re generally doing is surveying larger areas to kind of maybe a 50 to 60 mil accuracy. So they effectively complement each other
Just recently, a project we worked on three years ago as an initial design has gone back out to traditional surveyors as they want more accuracy on stuff like bridges. So really what it’s done is it’s brought the use of manned aircraft, which were really only for bigger projects, it’s brought that use down to the smaller project. The traditional surveyors are still in work and the majority of projects we work on, we work on together.
Brock: Can I jump in there as well, just for a moment? It’s quite interesting and I think it’s an exciting time. We had the pioneers of this drone industry—people like Aidan and like Marc and many other users—who really went out there and proved this, proved the accuracy, proved the data and so forth. Where we are at the moment is we have over 320,000 flights that have been completed with the eBee platform. 12 million hectares, roughly, of coverage that we know of at the moment. So we’re now at that next phase of that expanded growth. I think it’s accepted. We’re pretty well comfortable with the acceptance of this technology out there and we’re now onto that next stage.
Aidan: I’d like to add, we’re actually doing a lot of the jobs that surveyors generally didn’t want to do, in so far as difficult terrain; terrain it was probably going to take you weeks to try and get a handle on, and generally not the nicest environment to work on. So generally we’re kind of doing that type of work.
… we’re actually doing a lot of the jobs that surveyors generally didn’t want to do
Just yesterday, I got thanked by a surveyor that’s been working in the industry for 25 years—it’s the work he used to hate doing, the big difficult jobs, rocks and landfill sites, areas where the environment isn’t particularly hospitable. We’re in and we’re out in a day. These guys don’t have to crawl over landfills and in safety areas, and quarries, we’re taking that sort of risk out of it.
Explore a point cloud from one of Aidan’s quarry flights:
Marc: From my perspective, I’ve evaluated both LIDAR and conventional photogrammetry. I think it’s really simply put: aerial mapping from photogrammetry, or from a UAV, is more akin to your conventional aerial mapping and photogrammetry than it is to LIDAR. Aerial LIDAR is a completely different platform, where it does have more penetration down to the ground level. It is really a different technology and it’s a different tool. I think they both have their applications and their uses, but they’re definitely a different application.
Aidan: I’d just like to add that the big difference really with UAV technology, used in photogrammetry, is that what you’re really getting is a surface model, so you do inevitably pick up stuff like buildings or heavy vegetation. LIDAR has the ability to penetrate through these, whereas we generally have to take these out manually.
… the big difference really with UAV technology, used in photogrammetry, is that what you’re really getting is a surface model
What I’ve found generally is that our accuracy on surfaces—like roads or even on quarries, surfaces that are freely open—are generally much higher and the coverage, again, it’s specific. Generally, LIDAR data, from what I’ve found, particularly if it’s in Ireland, it can be quite old if it’s off the shelf and it’s quite expensive to get LIDAR data just for one particular project. So it’s a gap we’re filling. We generally don’t have much of a problem filtering out buildings and trees and stuff like that you know. It’s quite an easy process. It’s just one thing to bear in mind that they are two entirely different technologies in effect.
Aidan: Well the main reason we used GCPs is because we’re an early adopter of the technology; the eBee that we have doesn’t have the capability to use RTK. Nowadays it’s something we’re actively looking at, especially the eBee Plus, it does kind of excite us, especially the PPK element. It’s probably something we are going to upgrade to and we do see the benefit of it.
The technology is evolving quickly but the RTK is special and it is a game changer again
The eBee that we operate at the minute is actually the second UAV we’ve had from senseFly. We had the swinglet, which was five years old or more, that was the first generation. The technology is evolving quickly but the RTK is special and it is a game changer again. It’s something we will probably upgrade to next year I would imagine.
Look out for part 2 of this article coming to Waypoint soon (subscribe for free above). This future post will cover the following topics: absolute accuracy, BVLOS benefits, how to go from DSM to DTM, determining flight height, dealing with the data, margins of error and more.
Watch senseFly’s full Drone Surveying 101 webinar on demand: