In our final class activity we were once again at the Priory, this time flying UAS or unmanned aerial systems. The main objective of this activity was to give us a sense of how to go about execcuting a survey using UAS. The emphasis of this lab was on mission planning, pre-flight safety procedures. It is important to remember that although we refer to these systems as using Unmanned Aerial Vehicals or UAV, it still takes trained humans to utilize them properly.
Having an understanding of UAS is important because there are many applications of remote sensing that can be done using UAVs. They are not without issues, legally and ethically speaking, but they are almost certainly going to continue to have a large impact on geospatial information systems.
For background information on policies and regulation surrounding UAS, visit the website of the Federal Aviation Administration.
Study Area
This week we were once again at The Priory. The UAV collected flew above the grassy field by the parking lot, the parking lot itself, and some of the surrounding woodlands.
 |
| Figure 1. Map of priory location in relation to UW Eau Claire campus shown on left, map obtained from Priory Facebook page. Aerial photo of Priory shown on right, image obtained from Google earth |
 |
| Figure 2. Photo taken at the priory the day of the flights, 5/11/2015 |
When flying UAVs it is important to be aware of the weather conditions. During the day we went out, it was about 52.9 °F, very cloudy, with winds coming from the southeast at around 2-8mph. In regards to the weather the biggest concern that we had was the wind, and how it would effect the flight of the two UAVs we would fly. As Professor Joe Hupy succinctly stated,
"I'm kind of sketched out about these winds"
 |
| Figure 3. Kestral meter used to measure various weather conditions |
It is worth noting that there are various different kinds of wind, and that they can have different imacts on UAVs, for instance Professor Hupy told us at the beginning of the class that straight line winds can be relatively easy to manuever out of, but thermals can be much more difficult, and act like a riptide as the UAVs get caught in them.
Methods
Because of the inexperience that most of the class had in flying UAVs, and their respective expenses, this activity was largely a passive one, where we learned by watchng the instructor and Mike Bomber show us how to go through a pre-flight check list and then carry out a mission.
Some examples of things that were on the pre-flight checklist include:
- Weather check: used a kestral meter to take note of weather conditions
- Electrical connection on UAV check
- Frame connection check
- Prop secure check: checked to make sure all props (where the propellors connect) were good
- Battery secure check: make sure battery properly attached on UAV
- Antanae secure check: make sure antanae secure
- sensor secure check: make sure sensor secure
We then learned how to use mission planning software on a laptop to draw a flight plan, and also on a tablet device. The laptop offers more versitlity, but the tablet is nice because it can geolocate.
 |
| Figure 4. Professor Joe Hupy explaining the mission planning software on the laptop |
After we had a flight plan, and Mike and Emily returned with a battery, we went back and completed the preflight check-list.
The first UAV we got to see in action was the Iris, which had a GoPro attached as a sensor. The GoPro takes images, which are then sent to the basestation, which in this case was the laptop. Part of the pre-flight checklist addresses making sure that these are connected.
 |
| Figure 5. Iris UAV shown on the ground on the left, and in the air on the right |
The flight lasted approximately 5 minutes, and had a successful auto-landing.
The next UAV we got to see was the Matrix, which had a GEMS sensor. The Matrix is larger, and has a bigger battery, and can remain in the air for longer.
 |
| Figure 6. Matrix UAV shown on the ground on the right, and in the air on the left |
The flight lasted less than 2 minutes. This flight did not go smoothly, due to the wind. Mike Bomber used the override on the remote controller to stop the flight plan and return the UAV to the starting point. The Matrix was successfully landed, and no damage occured.
Ground control points were established so that geographic locations can be found with the imagery. When choosing ground points, it is best to choose speciffic locations that can be identified in the imagery you collect, so that it can be referenced accurately. An important thing to remember is that your data can only be as accurate as your ground control point, so using a dual frequency GPS is necessary for the best results.
 |
| Figure 7. A-setting up tesla unit, B-example of good visible location for GCP, C-taking GCP at parking line, D-taking GCP at tree, E,F-taking GCP at crack in road. |
Results
The sensors on the UAVs worked successfully, and after mosaicing images of the entire flight plan area were produced. In the figure below, I have three of these images shown. One is an RGB or true color image, the other is an NDVI or a Normalized Difference Vegetation Index (which highlights vegetation in green), and one that was labeled as overview mono fast.
 |
| Figure 8. Mosaiced images from the entire flight plan |
One part of the image that I thought would be interesting to look at was the part where you can see all of the students standing. In the RGB image you can see a blurry, but otherwise decent depiction of us. In the NDVI image, you can see that we are not in fact vegetation, and in the overview-mono-fast image, you can see another variation of us. I think that is goes to show how impressive the it is that you can even to some extent differentiate which students are which. It also goes to show why some people have concerns about this technology, and why many worry that it gives a new cause to be concerned about privacy.
 |
| Figure 9. Students in RGB true color image |
 |
| Figure 10. Students in the NDVI image |
 |
| Figue 11. Students in the Overview-Mono-Fast image |
Discussion
The two biggest problems that occured during the lab, was the short circuiting of the remote controller that caused it to start smoking, and the wind posed a potential threat of carrying away one of the UAVs. Both of the problems were handled and fixed because of our proffesor Hupy's and Mike Bomber's understanding of the system. They also show the importance of having a pre-flight checklist, as well as being aware of changing weather conditions.
I think that one interesting point that can be made when looking at the pictures of the students in the class, is how you can tell the images are taken at an angle. This can be good or bad depending on what the purpose of your study is.