Unmanned Aerial Systems Mission Planning

Introduction:
This activity is a short course in Unmanned Aerial Systems or UAS. We will discuss the basics of UAS and the different platforms that are commonly used. This activity will take us from beginning to end, from mission planning to the actual flight to image processing and interpretation. We will not go into the detail that we have went through in the actual UAS course here at the University but if you would like to know more about UAS, you can check out my blog here: http://nikandersonuas.blogspot.com/

Study Area:
Our study area for the image gathering portion of the activity is on the point bar across the Chippewa River from campus (Figure 1). There is a big 30x30ft "24" that some kids made in the sand with rocks and other sediment. I do not know what this resembles but I do know that it made for a quick study area that we could capture imagery of to then process later on.

Figure 1: Map showing the study area in relation to the City of Eau Claire. NOTE: This imager was taken during high water and does not reflect the water level at the time of this activity (It was much lower)

Methods:
The first thing that we did during this lab was getting introduced to UAS and the different platforms that we work with. Professor Hupy took us through some of the advantages and disadvantages of both multicopters (Figure 2) and fixed wing aircraft (Figure 3). For example here of some advantages and disadvantages of each:
Multicopter Advantage: Can fly in a variety of locations with a small turn radius and no runway.
Multicopter Disadvantage: Is not suitable for covering a very large area and has a shorter flight time.
Fixed Wing Advantage: Can cover large study areas and has a longer flight time.
Fixed Wing Disadvantage: Needs a runway or larger area to take off and has a large turn radius.


Figure 2: A Hexacopter

Figure 3: A fixed wing aircraft

This is a great transition into mission planning because in order to effectively plan a mission, you must know some of the advantages and disadvantages that I just went over. For instance, our mission was to collect imagery over some rock formation on the sand bar. Since the rock formation was going to be relatively small, the multicopter option gets a point. Since we were going to need to be focused on the rock formation and collect many different images, multicopter gets another point. Finally, since it would have been difficult for a fixed to land (since all there was were water and sand) fixed wings lose a point. All in all, we decided multicopter to be the option that best suited our study. Once we had that figured out we could go into the mission planning software called Mission Planner and start mapping out our study. This software is very easy to use and we simply tell the UAS where we want it to fly, how high the altitude should be, and other parameters such as how fast the UAS should be flying. The interface looks something like Figure 5 and simply requires us to draw a box around our area of interest. We can go further and tweak the parameters because the software will also tell us important details such as how long the mission is going to take and how many pictures will be taken at this speed. These are all very important when planning out a mission because you want to make sure that your batteries can last long enough for the mission to complete and you don't want not enough or too much data. Figure 6 below is a good representation of how we can change parameters on the planning software. In this example I switched the angle that the UAS would fly at from 90 to 0 and by doing so I told the UAS to make longer swaths and therefore decreased the amount of total time it would take the UAS to make it turns. This can significantly cut down on flight time.

Figure 5: Mission Planning software showing the proposed flight path of the UAS

Figure 6: Example of how changing the flight angle (bottom right) can decrease flight time

Once the Mission is planned, we can relocate to the study for the launch of the UAS. Now usually we may have a multicopter checklist that we can go through to make sure everything is in order so that we will have a safe flight for both the UAS and more importantly, bystanders. This multicopter that we were using is a relatively small UAS platform called the Phantom (Figure 7).  The Phantom is a quadcopter which means is has four rotors on it. It is also self correcting so while the flight is in progress, if we switch to manual (as we did for this study) we can easily fly the UAS and it will correct itself. This basically means that when one rotor starts to drop or loses RPMs, the other rotor will make up for it and gain RPMs. The pilot for this activity was our Professor and he flew the Phantom over the "24" and took pictures manually (Figure 8).

Figure 7: Photo showing the Phantom while our Professor explains some of the specs of this UAS

Figure 8: Image displaying the flight in progress

Once the flight was over and we collected all of our data, we head back to the lab for a brief introduction of how to process these images. The software that we use for this is called Pix4D and if there is one thing you need to know about image processing with this software is that it will take a very very long time. Our professor gave us the know how of which buttons to push and after that It was up to us to process the images ourselves. When I processed my images it took me around two hours to do in the "Gucchi" lab. The end result of the processing from Pix4D (Figure 9) is a few folders that they automatically create in your own student folder. The two main resulting products that we got were one DSM (Digitial Surface Model) (Figure 10) and a Mosaic (Figure 11). The DSM can give us a good idea of the elevation of our surface while the mosaic compiles all of the photos together into a high composite image of our feature "24". I had an issue getting a DSM of the entire study area, either this was an export issue or this is just how it is and I myself have to mosaic those DSMs together. On the other hand, the mosaic that I got was in excellent condition and had a very high resolution.

Figure 9: The end result of the Pix4D processing

Figure 10: The DSM result of the processing in tile format. You can see that the left of the photo is green which stands for lower elevation, which makes sense because it is closer to the river.

Figure 11: The resulting mosaic of the processing

The last part of the activity that we had to do was to put in one hour on the Real Flight Flight Simulator. We were tasked to do one half hour with a fixed wing and one half hour with a mulitcopter. I have already put in many hours on this simulator for the UAS class so I should be able to quickly explain how these platforms feel when flying. The fixed wing in my opinion is the easiest and most fun to fly. You can fly these in first person view or third person, with third person being the most realistic. Usually you have a three or four channel plane when flying fixed wings. The three channel planes consist of a rudder, elevator, and throttle while the four channel planes consist of a rudder, elevator, throttle, and ailerons. The rudder controls your Yaw, while the elevator controls your Pitch, and the ailerons control your Roll (Figure 12).  Obviously your throttle controls how much power you are giving the rotors and how fast you want your plane to fly. One thing to note with the fixed wings is that they are quite touchy and it doesn't take much to maneuver them so when you are flying, you want to take very small controlled movements.

Figure 12: Diagram illustrating roll, pitch, and yaw on a fixed wing

The multicopter is a little harder to fly because it doesn't always fly in the same direction. Planes are always moving forward while multicopters can move up and down, forward and back, and side to side. It is also hard to distinguish where the front of the UAS is because they are typically symmetrical in shape. These are also touchy but the hardest hurdle to overcome is know which way the multicopter is orientating itself and from there, pushing the right sticks to make it fly where you want it to go. One advantage these have is that you can fly them from multiple camera angles. They can fly in first person, third person, chase, or through the "bombay" doors.

Discussion:
If you told me I could only have one platform and I had to do as many geospatial tasks as possible with it, I would pick the mulitcopter. I would do this because you can control them so much easier and you do not need a lot of room for them to operate. With the Mission Planning software that we have, we are able to plan out the UAS's flight path ahead of time and not worry about pilot error when in flight. Yes, we would always have a pilot on hand so they could take over in case something went wrong, but for the most part, you wouldn't need to. The multicopter is able to take a variety of sensors that you could use for a lot of different studies and monitoring. With all this being said, I do recognize that every situation calls for its own platform and I think the fixed wings are great when you have a large study area and you need to be flying over 10mph so that you can get it done quickly.

Scenario Discussion:
We each had to pick a scenario that our professor has come across through consulting work and give our own consulting report. The scenario I picked was:
A pineapple plantation has about 8000 acres, and they want you to give them an idea of where they have vegetation that is not healthy, as well as help them out with when might be a good time to harvest.

Mr. Dole,

I have reviewed your project goals and have came up with a gameplan on how we can get this project rolling. You noted in our last email that you wanted to know where your crops were not healthy. The solution to this problem is using our fixed wing UAS and attaching a NDVI monitor to it. NDVI stands for Normalized Difference Vegetation Index and it gives us a very good representation of how healthy your crops are because the sensor looks at photosynthetic activity. Basically the sensor shoots visible light down on the vegetation in the red spectrum. Healthy vegetation reflects more red light while vegetation that is not healthy absorbs the red spectrum. The end result will be a collection of maps that display healthy vegetation as green and unhealthy vegetation as red. This will be simple to interpret but I can always go through the results with you anyways. If you would like to learn more about NDVI then you can click on this link: http://earthobservatory.nasa.gov/Features/MeasuringVegetation/measuring_vegetation_2.php

The other issue that you wanted to address is when you should harvest. Pineapple as you know should be harvested when they are a rich yellow. So what I can do for you is to develop a sensor that is going to shoot visible light down at 5.8 micrometer wavelength which will pick out yellow the best. What will happen is the pineapples that are a rich yellow will give us a high value because they will reflect more of the light. The result will be a series of maps that show you where the highest value pineapples are that are ready to harvest. This will save you time and money because you will be harvesting at the most optimal time.

Both of these sensors will be flown with our fixed wing aircraft. Due to the large study area I need a platform that will cover ground quick but also maintain a high quality image. I will need to take off and land multiple times so that I can switch out batteries and start downloading some of the images off the sensor since this is going to be a data intensive project. After all of the images are taken I will process them and give you a final report on your 8000 acre plantation.

Let me know what you think and when I can start,

Nik Anderson



Conclusion:
This activity hit all the basics of UAS, from the different platforms to the different applications and analysis. I believe that UAS is going to be a difference maker in many realms. They already have studied so much and helped so many people, I just hope that the UAS community will stray away from any negative publicity because that will only further discourage people from allowing them to operate. Furthermore, I think that knowing all the different steps of how to operate UAS will greatly improve any chances that I may have to work with them some day. It is not enough to just learn how to fly them, because most scenarios are going to call for auto pilot situations anyways. Knowing the different platforms and the planning software will make any missions that I may fly far more efficient.


https://howthingsfly.si.edu/flight-dynamics/roll-pitch-and-yaw
http://www.multiplazaonline.com/Honduras/San_Pedro/logos_Sn_Pedro/H1_SuperJugos.htm
http://diydrones.com/profiles/blogs/full-autonomous-cross-country-soaring-flights-with-the