Lab 3: Evaluation of UAS Platforms and GPS Units for Ground Control

Introduction

On Saturday September 30 our class went out to Litchfield Mine to evaluate Unmanned Aerial System (UAS) platforms and Global Positioning System units for ground control. Litchfield mine is an aggregate mine located southwest of Eau Claire on the Chippewa River. The mine's product is sorted and stockpiled. To collect inventory of this product, the volume of the stockpiles need to be measured. One safe method to do this is through aerial photography. Data is collected with a UAS platform and analyzed with GIS. We collected data during this lab, the data will be analyzed in a later lab. Professor Hupy invited professionals from Menet Aero and Topcon Solutions to join us in the field and show us a variety of UAS and GPS platforms allowing to compare and contrast a variety of equipment. The focus of this lab is to compare the different units collecting data to be used in a later lab. 

Study Area

The study area was Litchfield Mine, an aggregate mine southwest of Eau Claire, Wisconsin on the Chippewa River, see location Map 1. They mine Quaternary Alluvium on the banks of the Chippewa river. 

Map 1: Study Area

Methods

Select Ground Control Points

Selecting ground good ground control points (GCP) is key for UAS surveys. They will ensure that all images collected by UAS equipment are accurately georeferenced. Our first task of the day was to place 16 ground control points. Our points were large foam checkered squares placed on the ground and spray-painted with their site number. These squares will be visible from the height that the UAS equipment will be flying. Our GCPs were placed throughout the study on a variety of surfaces (grass, gravel) and variety of elevations. It is best to make triangles with the GCPs, they should not be in a straight line. When placing GCPs keep detailed notes about the placement location and draw a field map.

Figure 1: Maike with unplaced GCP

Menet Aero provided AeroPoints Propeller GPS markers, Figure 2. These 2x2 foot foam pads are dubbed a "smart target". They are GCPs that include an internal GPS powered by solar. They communicate with eachother and with satellites, sending their data directly to the cloud. Their accuracy is within 2-6cm of actual position. They are lightweight, water proof, and are not damaged when driven over. Some disadvantages of using AeroPoints as a method for GCPs, is that there is that real-time data is not accessible in the field. Additionally these markers are $500 a piece, so the number of GCP in a UAS survey is limited by budget.  

Figure 2: AeroPoint Propeller GCP. The foam mat at the base of the GPS system.

Survey Ground Control Points

Once all GCPs have been placed, collect their GPS coordinates. The most accurate coordinates will be used in a later lab to georeference the UAS images. We collected GPS coordinates using a variety of equipment including the Trimble R2, Septentrio Altus NR2, Topcon HiPer, Bad Elf GPS units, and our personal cell phones. When collecting this data, it is imperative that the GPS unit is as close to the cross of the GCP as possible, as seen in Figure 3.

The Trimble R2, Septentrio Altus NR2, and Topcon HiPer were all similar pieces of equipment. Once located directly atop the GCP and leveled, geographic data was collected with an interface screen attached to the unit. See Figures 4 and 5.

Figure 3: GPS units Topcon HiPer and Bad Elf positioned as close to the cross of the GCP as possible.

Figure 4: Allison using the Topcon HiPer interface Screen.

Figure 5 shows the Topcon HiPer interface screen which shows preloaded imagery of the location as well as satellite and WiFi signal strength. Once data collection begins at each point, watch the screen for fluctuations of satellite and WiFi strength - if signal becomes poor stop and collect at that point again. After thirty seconds of data collection with optimal satellite and WiFi, click stop and save. All coordinates collected during that time are averaged and recorded in a spreadsheet. Professor Hupy will provide the class with that data after the lab.

Figure 5: Topcon HiPer interface Screen.

We also collected GPS coordinates with the handheld Bad Elf GPS units and our personal cell phones. We placed these as close to the center of the GCP cross point (Figure 3) and recorded the coordinates in our notes.

Conduct UAS Survey

Four UAS were used to conduct a survey of the study area, the DJI Phantom 3 Pro, Sensefly Ebee, M600 Pro, and C-Astral Bramor. Each platform varies in UAS type (rotary/fixed wing), takeoff and landing procedures, and specs. All followed a similar procedure however. Platform takeoff and landing locations were selected and cleared of people and obstacles, and then the flight plan was programmed on a laptop or ipad. All UAS utilized a pre-planned flight plan, as manual fight invalidated the UAS warrenty. All UAS were operated by the visiting industry professionals.

DJI Phantom 3 Pro

The DJI Phantom 3 Pro was a small rotary wing UAS with 30 minutes of available flight time. It takes 1 hour to charge the battery. The flight was planned with a programmable controller attached to an Ipad. This UAS took off and landed in the same location.

Video 1: Takeoff of the DJI Phantom 3 Pro

Figure 6: DJI Phantom 3 Pro controller Ipad Interface

Sensefly Ebee

This UAS was a fixed wing foam unit with internal compartments that housed a camera and tracking device, see Figures 7 and 8. Fixed wing platforms like this one can cover more ground on the same battery time. Its battery lasts for 59 minutes of flight and can handle up to 28mph wind speed. It is programmed to return home in the event of poor GPS signal, high wind speed, or poor connection to the controller. When photos are collected, the motor on the unit briefly turns off. The flight was planned on a laptop in the field with pre-loaded background imagery. Take off is via hand launch, see video 2. Landing is a controlled crash, as the unit flys at a shallow angle down a landing strip until it stops. A disadvantage of this unit then, is that it requires a significant amount of space to land and is not optimal in locations with tight spaces. 

Figure 7: Underside of Sensefly Ebee

Figure 8: Topside of Sensefly Ebee

Video 2: Takeoff of Sensefly Ebee

M600 Pro with GeoSnap Pro

This larger, rotary wing unit takesoff and lands in the same location. It features a reliable camera that is separate from platform. During flight planning, it accounts for photograph size and you can directly control photograph overlay in the field.

Figure 9: Closeup of the M600 Pro

Video 3: Landing of the M600 Pro

C-Astral Bramor w/Sony a6000

This is another fixed wing unit, but it is made out of hard plastic and very much resembles a small commercial air plane. It is launched with a large crafted slingshot (See Figure 10) and lands via parachute. This means that there is less control over where the unit will land. Another disadvantage is that is the parachute mechanism malfunctions the unit will crash and may be damaged beyond repair.

Figure 10: C-Astral Bramor slingshot takeoff gear

Conclusions

There are a variety of methods and technology available to collect data necessary to conduct volume-metric surveys of the Litchfield stockpiles. Each unit has its own pros and cons depending on the type of survey being conducted, accuracy and precision of data desired, and budget. In depth pros and cons of each unit is described under methodology. For example however cell phone GPS coordinates may be accurate enough for me to navigate to the mine, but is not accurate enough to pin down a GCP.

Most of the data collected during this lab is not yet available to students. However, this spreadsheet details the GCP coordinates collected via Bad Elf and personal cellphones.