Waypoint recently sat down with Stephen Frazee, Project Engineer at WET, to learn more about the company, how fixed-wing drones are helping them save money on projects, their preference for PPK workflows and more.
Thanks so much for chatting with Waypoint today, Stephen. Before we jump in, can you give us some background on Water & Environmental Technologies?
So, we’re kind of a hybrid civil engineering and environmental consulting firm. We’re based in Montana. Our headquarters is in Butte and we have a couple other offices around the state, but we do work across the western U.S.
What type of professional drone services does Water & Environmental Technologies provide?
Most of our drone or UAV work is on the engineering side, although we also do environmental-based work such as mapping for permitting, vegetation/weed mapping and stream temperature mapping. We’re always looking for outside-of-the-box UAV applications, but the bulk of it is for surveying and engineering purposes; things like topography surveys, mine surveys, stockpile surveys, site development, landfill life expectancy and flood mapping.
When did Water & Environmental Technologies first start using fixed-wing drones?
In May 2015 we bought our first eBee. I think at this point, we’ve owned an eBee Classic, eBee RTK, and we are now flying an eBee Plus, along with a few other platforms. Some projects we fly for internal purposes, things like engineering design, and others we act more like a UAV service provider where we’re flying, collecting the data and giving that data directly to the client.
You mentioned the types of data that you put together for clients. Can you run us through some specific deliverables?
So, for example, one project we recently did was an open-pit mine site. It was 5,500 acres with ~2,400 feet of elevation change across the site. For that project, it was basically two deliverables; one was the orthomosaic photo of the whole mine site and the other was an elevation model. We ended up providing a 1-meter grid DTM for the whole site.
The photo was great for [the client], from a permitting standpoint, they use it as a background for a lot of their maps that they submit for permitting and tracking things like reclamation. The elevation data they use for tracking how things have changed at the mine from the standpoint of moving earth from point A to point B, but also, they’re looking at things like slope stability; has this chunk of the mountain moved inches or feet in the last couple years? They also use the topography to plan out future exploration efforts and more from a geology standpoint as well.
The other advantage of using fixed-wing drones is you get greater resolution. I would guess that the highest resolution that they were getting from the manned aircraft photo was maybe a six-inch resolution; whereas we can provide down to 1.5-inch or 2-inch resolutions.
Another thing that’s popular for mine sites that we didn’t really do for this project, but we do for quite a few other clients is stockpile volumes and reconciliation efforts. Typically, the client is checking the stockpile volumes we provide against what their own internal records show, whether it be from truck counts, or scales, or something like that.
With the mine project, how would a project like this have been tackled without the use of fixed-wing drones?
For that site, it’s so large that they collected this data in the past with manned aircraft, using traditional photogrammetry at a much higher altitude than what we fly. This project is a little counter-intuitive to what you typically see when comparing drone use versus traditional methods. With the manned aircraft, data collection would’ve been much faster, probably a couple hours of flight time and they’re done. You’d also get fewer photos; you probably only would’ve had – I don’t know – maybe twenty to thirty images to stitch together. Compare that to the 12,000 images that we stitched together. So, it’s a little bit against the grain in that regard but we were able to do this cheaper than what they had paid previously for a similar service from a traditional manned aircraft provider.
The other advantage of using fixed-wing drones is you get greater resolution. I would guess that the highest resolution that they were getting from the manned aircraft photo was maybe a six-inch resolution; whereas we can provide down to 1.5-inch or 2-inch resolutions. Then from what I’ve seen, the elevation model that we produced is also much higher resolution than what they had before.
So, it would have been possible with manned aircraft but at greater cost and reduced image quality, is that correct?
Right. I think somewhere there’s a break-even point where using drones is no longer cost effective versus using manned aircraft. At some point, the project is sufficiently large that it’s just going to be cheaper to fly the manned aircraft. I don’t know what that number is exactly, but for this project, it was cheaper to do it with the drone. As the FAA becomes more accepting of beyond-visual-line-of-sight [BVLOS] flights, fixed-wing drone use for larger projects will become more economical.
You mentioned that you were on your third eBee? What made you originally decide to explore and ultimately go with senseFly drones?
Yeah, so the eBee was the first commercial fixed-wing drone that we were exposed to. A distributor in our area showed it to us and we began to envision how we might be able to use it in our business. We ultimately purchased an eBee for several reasons. We really like the simplicity and size of the system. Obviously, the eBee is not a simple system, but operating one is. The fact that the autopilot takes away most of the human error in the process was viewed as a great advantage to us — we are engineers, not pilots. We wanted something that was easy to operate and the eBee fit the bill.
We really like the simplicity and size of the system. Obviously, the eBee is not a simple system, but operating one is.
The eBee mapping drone is also lightweight and poses a minimal risk while in flight. That has its advantages and disadvantages, but to us, just from a liability standpoint, having something that weighs a pound and a half flying around in the air is a much lower risk and safer than some of the larger fixed-wing drones on the market and, most importantly, our clients view it as a minimal risk.
Mobility was another important aspect. The entire system fits in a compact case, which makes it easy to move from point A to point B without lugging a giant case or a rail-launch system. We didn’t know it at the time of purchase, but the eMotion drone software is also a huge advantage. We’ve been in the game long enough and we’ve flown a couple of other platforms to know that eMotion is by far the best flight controlling software that we’ve encountered to date.
Can you walk our readers through the typical workflow for collecting and processing your drone data?
If it’s a simple project that you could cover in one flight or less, we’ll do a quick mission plan in eMotion when preparing our cost estimate/bid to verify that we don’t foresee any issues that may cost us more time in the field.
…Having something that weighs a pound and a half flying around in the air is a much lower risk and safer than some of the larger fixed-wing drones on the market and, most importantly, our clients view it as a minimal risk.
We try to identify issues up front as we’re costing it out and maybe do high-level mission planning to generate our cost estimate or bid. Once we secure the job, we’ll go in and plan it out in more detail, do things like – assuming it’s a surveying type job where we’re going to set ground control – we’ll plan out the locations and the number of ground control points that we’re going to set. Then, we’ll tweak our mission blocks in eMotion to accommodate our ground control. After that, we head to the site and set our ground control first and get those all surveyed in with our GPS. Once we have that done, we’ll fly – at least now with the eBee Plus – we’ll fly most flights in PPK mode.
Regardless if we need the accuracy or not, we’ll usually fly PPK just as an extra security blanket. We usually just set up on the tailgate of the truck with a good vantage point of the site, launch, and let the bird do its thing, capturing the photos. Then once we land, we’ll do a quick check of the imagery to make sure, one, the quality of images is what we expected, as well as a quick check to see if there’s anything obvious, to make sure the images are focused and sharp — just high-level checks like that. If the mission is small enough and the number of images is small enough, we’ll go ahead and process the images through eMotion and then run a rapid check in Pix4D to make sure that everything looks okay.
…The eMotion drone software is also a huge advantage. We’ve been in the game long enough and we’ve flown a couple other platforms to know that eMotion is by far the best flight controlling software that we’ve encountered to date.
For the bigger projects, we usually forego that step in the field just because it takes too long, so we’ll just rely on checking the images. Then, we do all our own processing in-house with Pix4D. So, we’ll process the imagery ourselves, and then depending on the project, customise the deliverables from there. So that’s kind of a general workflow for most of our projects.
What about the mine project you guys worked on? Did that differ much to your standard process?
For the mine site, it was heavy up front on the mission planning because it was so large. It’s 5,500 acres so we had to break it into over 50 mission blocks, due to the size and the steep and varying topography. Some of those would’ve been only 10- or 15-minute missions, but because of the topography and wanting to generally fly parallel to the contours, we had to generate a lot more mission blocks than you would normally for a flat site of the same size. So, we spent more time than usual planning out our missions and figuring out optimal locations to set up and fly from to minimise field time. Maintaining line of sight was a real challenge at this site. So, we made a separate site visit just to plan out our setup locations. We ended up setting up at five different locations over the course of two days.
By using fixed-wing drones, we can survey areas that are inaccessible on foot or too risky to access. We have been able to significantly reduce our safety exposure/risk on job sites thanks to our drone use.
How many people did you have in the field for such a large project?
Well, we had me and one other pilot in the field and we split it into two days. We were each equipped with an eBee Plus. We flew a little over half of the site on day one and flew the rest on day two. This site is notoriously windy and for both days we experienced consistent 15-20 knot winds with gusts upwards of 30 knots. Needless to say, we chewed through some batteries. We each had eight batteries and we started charging spent batteries as soon as they cooled down following a flight.
Let’s talk accuracy for a moment. Obviously, you save a lot of time using a UAV, especially when mapping a large area and a PPK workflow. In your experience, are fixed-wing drones like the eBee accurate enough instruments?
Considering how you’re collecting the data and the ease and speed with which you can collect it, the accuracy is astounding. And when you factor in the accuracy you get with PPK, and it depends on the nature of the flight and what kind of ground cover, but in good bare ground or bare earth conditions, we’ve found that the vertical accuracy with PPK can be around fifteen hundredths of a foot – with or without ground control.
Regardless if we need the accuracy or not, we’ll usually fly PPK just as an extra security blanket.
Those accuracy numbers are typical for smaller projects. With the size of the mine project, our expectations weren’t as high because our ground control network wasn’t as dense as we would typically use for smaller projects. To save on time/cost, we reduced the number of ground control points we used. In the end, we had about 50 ground control points, some of which we used as checkpoints. The vertical RMS error for our checkpoints was 0.15 feet. That’s impressive for any project, but considering the size of this project, that level of accuracy is quite impressive.
Being engineers and surveyors, we’re strict on our QA/QC procedures, so we rarely ever fly a project without ground control. In our minds, that’s just standard practice and that’s how we can guarantee our clients the accuracy that we promised. But yeah, we’ve made flights where we’ve done PPK and I’ll just use the ground control points as checkpoints and the accuracy will be a tenth of a foot or greater.
Considering how you’re collecting the data and the ease and speed with which you can collect it, the accuracy is astounding.
Why not just use LIDAR? Is there an obvious reason you chose to not go that route?
Yeah, so for this mine project, there were some areas that were heavily vegetated or at least forested with significant canopy cover. Still, a vast majority of this site was bare ground. Photogrammetry shines in bare ground or very low or minimal vegetation scenarios. LiDAR is typically a lot more expensive than photogrammetry, whether you’re talking about drones or traditional manned aircraft. I would say it depends on the project, but in this scenario, the vegetation was limited enough that we didn’t need [LIDAR] because the added accuracy that we would’ve gained from LidarR in the vegetated areas was not significant enough to warrant the added cost of going to LIDAR. I would say that cost was probably the biggest driver there.
If you had to sum it up, what would you identify as the main benefits of using professional drones for a project like the open mine pit or in general?
For us, it’s opened a whole new line of business, a whole new service that we can offer to clients. Beyond that, the quality of the data that you can get in such a short amount of time and with relatively low cost, that is probably the biggest advantage for us. Having an up-to-date and very high-resolution aerial image of our sites is also invaluable, from an engineering standpoint, and then you factor in the resolution of the topography data that we can generate, that’s a huge advantage as well. It just allows us to be more accurate and precise in our engineering designs and to deliver a more accurate and higher quality product for our clients.
LIDAR is typically a lot more expensive than photogrammetry, whether you’re talking about drones or traditional manned aircraft. I would say it depends on the project, but in this scenario, the vegetation was limited enough that we didn’t need [LIDAR] because the added accuracy that we would’ve gained from LIDAR in the vegetated areas was not significant enough to warrant the added cost
Safety is another important aspect. By using fixed-wing drones, we can survey areas that are inaccessible on foot or too risky to access. We have been able to significantly reduce our safety exposure/risk on job sites thanks to our drone use.
The quality of the data that you can get in such a short amount of time and with relatively low cost, that is probably the biggest advantage for us.
Are you able to define what the ROI has been on your drone purchases?
What I can tell you is that we’re doing well enough that we can easily justify upgrading to the latest, greatest model frequently. A downside in this industry is that the technology evolves so rapidly. But we’ve been doing enough work and have been able to recover enough cost from it to justify numerous purchases.
Thanks for chatting with us today, Stephen.
You’re very welcome.