Early Excavation Progress Pics and Videos of Upcoming Industrial Park by Drone - Lawrenceville, GA

This is a compilation of some aerial drone progress pictures and videos from a construction site in Lawrenceville, GA. This is my first attempt at progress pics – and it is much harder than I thought. The video contains the work from 8 visits to the site. I plan on returning to continue to take pics as the site progresses and eventually finishes.

It is very enjoyable capturing and showing off the excavation work and the crew bringing the project to life. These guys work hard in the Georgia heat.

Let's go into some behind the scenes details about the many issues I experienced while setting up these progress pictures:

The main issue I had with taking progress pics was putting my drone and the gimbal in the exact same position each visit. This is easier said than done. I have multiple axes in the sky to align with each time. I have to align the drone’s position on the X, Y, and X axis. As well as align the gimbal angle and white balance. Of course the camera settings need to be solid for each pic, too. Exposure does vary quite a bit between takes due to changes in weather. A sunny shot is quite a bit different from an overcast shot. Your primary source of light (the sun) is no longer providing direct light when it is cloudy, which provides a nice level of contrast. When it is overcast, the clouds scatter the light in every direction. This diffused light kills shadows and contrast.

Now, I messed up some of these things early on. Aligning the drone perfectly in a 360 sphere in the exact same spot is nearly impossible. I’m sure this would be easy with a RTK-enabled drone and a GPS base station, but I do not have this equipment. I am using consumer-level GPS in my Mavic 3 Pro. This means I can’t simply put in the same GPS coordinates on each flight to get the same exact position. If I track and continuously return to the exact GPS coordinates -- it won’t work, because consumer-grade GPS is not exact. It can vary by about 10 or 15 feet.

So, I do the best that I can. I also enjoy figuring things out on my own.

What is easy to align is my altitude - I simply record my altitude at each location, and make sure I return to that exact altitude every time I come back. However, one issue I ran into is that if I take off from a different area, my altitude readings will be different, because the drone is tracking my altitude from my take off location only. Most of the pics in this series were shot at 150 ft high. The two 90 degree shots looking straight down (nadir) were shot at 390 ft. That covers the Y axis. Variations in this axis, as stated before, can be reduced by taking off from the exact same spot each time.

What about the X axis? Horizontally? This is actually the easiest axis to align. I simply find permanent things on the ground that do not move (a light pole, a manhole cover, etc) to align myself with. I get right over whatever object I’ve chosen, switch my camera to the nadir angle, and align my crosshairs with the object perfectly. Variations in this axis are introduced with wind, or maybe something as simple as very slightly bumping the movement stick.

That’s X and Y. What about Z? This is all about where my drone is facing. This one is the toughest. My drone has to be facing in the exact same direction each time. Even a 1 degree difference is noticeable. If I am facing 220 degrees, I have to make sure I return to this exact spot every time. However, the drone does not easily give out this information. There is not an option to display this figure on my screen, but the drone is keeping track of where I am on this Z axis. To see this info, I have to pull up the Waypoint mode and create a Waypoint. The direction degree is recorded and displayed in this mode. However, I do not like using Waypoints because of the lack of precision in consumer-level GPS - which is what this mode uses. The Waypoint mode can set up an X and Y position that you can keep returning too, but it will be different every time. This is why these have to be done manually, by eye (aligning with a light pole, for example).

After X, Y, and Z, we now have the gimbal angle. This one is moderately easy, because the controller gives me my gimbal angle in degrees. It is simply a manner of lining that up and holding it.

One issue I ran into was not in relation to the X, Y, or Z axis: white balance. For the photos, I used a mix of auto while balance, and custom white balance numbers. This was not the right thing to do. I should have picked a white balance at the start and stuck with it, or kept close to it. White balance demands do vary by conditions. So, there is a little room for play here. The white balance needs to be accurate to the scene, in the end. But since most of my shots were shot in RAW, I had a little more room to play around with white balance, exposure and color in post.

Overall, I am enjoying this, besides the multitude of issues I am running into.

I greatly look forward to continuing this project.

#drone