Drone, Photogrammetry Map Hurricane Matthew Damage
In the following article, see how proactively teaching about the value of his deliverable opened the way for geospatial analyst Kwasi Perry to take his business to the next level.
Geospatial analyst Kwasi Perry is founder and lead consultant at UAV Survey Inc., based in Houston. Perry puts a strong emphasis on educating clients. It has made all of the difference in supporting the success of his start-up. His strategy in a nutshell is to research a particular company, gain knowledge on how unmanned aerial vehicles (UAVs) —also known as drones — can help them, schedule an initial consultation with the company to present his findings, and then hopefully make a deal.
“By the time the presentation has ended, we’ll usually have a client. So we have about an 85 to 90 percent [rate] of turning that into business after that initial presentation.”
Perry has a bachelor’s degree in geography and geospatial science from North Carolina Central University and studied geoscience as a graduate student at Texas A&M University. He is a former federal geospatial intelligence analyst with deployment experience in both Iraq and Afghanistan, and launched his UAV company in September 2015. UAV Survey Inc. flies drones and collects aerial data for companies looking to contract out. It also offers procurement, training, consultations and certification assistance for companies interested in developing in-house UAV programs. Perry says his business is open to helping with pretty much any drone-related needs, but that it leans mostly geospatial because of his geospatial background.
Every job he has taken on thus far has involved drone-based photogrammetry and, thanks to his proactive approach of teaching about the value of his deliverable, he has attracted big-deal clients, mostly from government and academia. One example is a job he recently completed for the State of Florida.
Post-Hurricane Damage Assessment
The assignment was to collect aerial data by drone over parts of Florida immediately after Hurricane Matthew, which touched down in fall 2016. The acquired data would provide a quantifiable damage assessment that would illustrate the effects of Hurricane Matthew on infrastructure. The client was the State of Florida’s emergency response team. They contracted civil and damage assessment engineers from the University of Florida, which in turn contracted UAV Survey Inc. to handle the drone operation.
The university reached out to Perry because he had worked with them before. He says they weren’t exactly sure what type of data he could provide going in, but they knew he could offer pictures and video at the very least. That’s when he told them, “While I could do pictures and video, and I will put that as part of your deliverables, I think we should go with a photogrammetry product, which is a 3D mesh, because it is detailed enough to see roof damage, where you can actually quantitatively assess roof damage size wise and everything.”
He was contacted about the project about two days before the hurricane hit Florida, so he flew to Gainesville, Fla., from Texas right away and left Gainesville once Hurricane Matthew had moved north of his assigned spots. The areas he was responsible for were Flagler Beach on the Atlantic coast and an area just south of Marineland, about 20 miles north of Flagler Beach. He says they arrived at Flagler Beach, the first stop, as soon as they possibly could while having daylight and safe weather conditions, before the winds moved things like shingles and tiles around too much.
Perry describes the setting as a typical coastal area, picturesque with homes not far off the coastline. The houses generally appeared unaffected by serious destruction, he says. Most of the residential damage was to roof shingles and tiles. However, the nearby A1A highway wasn’t so lucky.
“It totally destroyed the A1A highway. It was really bad from a civil infrastructure standpoint. It just totally eroded away and broke up the highway, which was recorded in absolute detail with our 3D model,” Perry says.
The Right Tools
Perry captured photogrammetric data and video of 25 acres of Flagler Beach and about 40 acres of the area near Marineland using a DJI Inspire 1 drone and an X3 12-megapixel camera with 4K video capability. For 25 acres, Perry says it took just 15 minutes to collect the needed data with the drone, then 13 hours to process it using Pix4Dmapper, all on his own. “So when you look at start to finish, it takes one day to get that product done and delivered. … I have an extremely powerful laptop. It’s 64 gigs of RAM, i7, has a desktop-grade graphics card in there. So it’s just as powerful as really strong desktops, but it’s in a laptop. So it allows me to process onsite if need be.”
In the end, he delivered a 3D point cloud model, digital terrain model, and pre- and post-damage orthomosaics. While he did not have enough notice to capture pre-damage data himself, Perry was able to overlay his post-hurricane orthomosaic on pre-hurricane data from Google Earth. Because he had collected at 2-3-centimeter accuracy, he says it overlaid perfectly. It showed roof damage, scattered shingles, tiles, and aluminum and vinyl siding. “It is amazing because it gives an idea of just how bad things were afterward,” Perry says.
A vital aspect of getting the State of Florida a useful dataset from which they could effectively evaluate the impact of Hurricane Matthew was choosing the right tools.
From a UAV standpoint, he says you can spend as much money as you want in the drone world. He selected the DJI Inspire 1 because he views it as the perfect medium between capability and price for companies like his that are just starting off. An important question he asks himself is, “Can it collect the data like I need it to collect?” If the answer is yes, he says there’s no need to pay for more bells and whistles. Perry also likes the portability of the drone. He says the smaller, more packable and accessible UAV seems to be the wave of the future and that he definitely sees the technology getting better there.
“It gives you enough to create a proof of concept and do the type of jobs to get investor confidence so they can say, ‘If this is what he’s doing with this, if we put in some money we can get the other platforms as needed and really go from there.’”
For data processing, Perry says he chooses Pix4Dmapper for ease of use. He has an unofficial relationship with Pix4D because he uses their product so extensively. He likes that the software can be as simple or as detailed as desired by changing the settings. “It’s a very powerful and flexible platform. Somebody new can use it right away and somebody experienced can get even more out of it.”
With the project being as short-notice as it was, Perry says he is very happy with the turnout. The challenges were not so much technical as they were administrative. “The unique part of it was the actual event that happened, but everything else geospatially, even though I didn’t actually do that [kind of project] before, I understood it in my head before we even went out there.”
The approach to accessing the site was new for Perry because he arrived before residents were allowed to return. He was given a placard and tag for his car that read “State of Florida Emergency Response Team,” which gave him authorization to get past National Guard and police roadblocks.
Speaking of law enforcement, he says the legal aspect of operating a drone and gaining acceptance from authorities is probably the biggest challenge he runs into. He expects this to continue into the foreseeable future. What he worried most about on the trip wasn’t data acquisition or processing; it was, “Will law enforcement allow us to do this?” He says law enforcement isn’t always knowledgeable of the rules surrounding drone operation and the permissions Part 107 and Section 333 exemptions give. “They can make the final decision and say, ‘This paperwork may look good, but I don’t want you flying,’” Perry says.
In a situation like this, he says the smart thing to do is listen and come back to fly another day. It isn’t worth putting the business in jeopardy. This is an important reason to have a thorough understanding of what can be done under whichever Federal Aviation Administration (FAA) certification or exemption one has, he points out. “At the very least, you may be able to talk to them. It’s really good for you to know what you can do, why you can do it and be able to explain it in a manner that’s not inflammatory. … Sometimes it’s just a matter of education to law enforcement and they’re like, ‘Oh, OK. I see what you mean.’ So you really have to be an ambassador for the UAV community.”
This level of professionalism and eagerness to explain, which Perry also applies to clientele, has helped him reach a huge milestone. In his initial business plan, his ultimate aim was to go global. Now, just over a year after launching UAV Survey Inc., he has taken on his first international project. Not long after his Florida project, he was contacted by the University of Notre Dame about collecting and processing data for a site affected by Hurricane Matthew in Haiti.
“It ended up working extremely well where they were excited and shared this with their colleagues at other universities. So, coincidentally, that’s how the thing in Haiti started. It was driven by the work that I did and the product I delivered to the University of Florida, and then from there the State of Florida.”