Nov 10, 2020

AT 409 FIELD REPORT

 

Unmanned Aerial Systems Capstone I

 

Purdue University

 

Fall 2020

 

Week 11: 11/2-11/9

 

 

Group 2

 

Jesse Giampaolo, Nick Dayton, Cole Bramel

 

Context

 

AT 40900 is the first part of the Unmanned Aerial Systems Capstone for seniors enrolled in the UAS program at Purdue University. As a class, we are currently tracking the foliage as it changes color and eventually drops at Martell Forest from the end of September through late

 

October or early November. This data will hopefully help Aish (Forestry PhD student) and Dr.

 

Hupy (AT 409 professor) achieve species and genotype level identification of the trees at Martell. Zach and Will (UAS grad students) are also helping with this project to show that such identification can be accomplished with UAS collected data.

 

To achieve the data collection this semester, the At 409 class has been broken down into groups consisting of three students each and as a class are collaborating to collect data from a minimum of three missions per week. For this research, the more missions the better the data set will be (ideally). A mission consists of two successful flights over the two different plots of interest (NW & NE) at Martell using a DJI Matrice 600 equipped with a Zenmuse XT2 and a

 

Sony α6000 (Figure 1). The North-West plot consists of naturally occurring oak and other trees. The North-East plot consists of 10-12 year old precision planted red oak trees which have been professionally maintained (Figure 2).

Figure 1: DJI Matrice 600 equipped w/ Zenmuse XT2 & Sony α6000

Figure 2: The mission area - precision planted and regularly maintained red oak plot

 

 

Week 11 Overview

 

This week, the class met with Measure Co-founder and Chief Strategy Officer Jesse Stepler for a discussion about the measure platform. The meeting occurred over Zoom, and was recorded. After the meeting the class met in COMP 101 at the Purdue University Airport for a demonstration on replacing a broken servo in the C-Astral drone, the Bramor. According to the scheduled flights on the Outlook Calendar, there were four flights occurring this week. Professor Hupy flew on Tuesday 11/3, crew four flew on Wednesday 11/4, crew three flew on Friday 11/6, and crew two flew on Sunday 11/8. The SOP for PPK corrections was adjusted to now use the WGS84 Geoid to prevent future processing errors in Pix4D and a new workflow was released. Cole will be processing the data for crew two’s Sunday flight shortly using this new workflow.

 

Week 11 Planned Flight: 11/7/2020

 

While crew two usually flies on Thursdays, Jesse was unable to attend at this date due to being absent from the state of Indiana for two days. Cole Bramel was following quarantine protocols, and was also unable to make Thursday, so the group pushed the flight back to Saturday 11/7. This flight was then pushed back to Sunday 11/8 after Nick was unable to operate in the field after an allergy related sickness. Jesse and Nick were to be the PIC and FO in the field, while Cole was to remote login to a laboratory computer and process the data, as well as organize and catalog it.

Week 11 Conducted Flight: 11/8/2020

 

The morning of the flight, Jesse confirmed with the crew that the weather was flyable. The TAF reported low expected winds, high visibility and no significant weather events during the scheduled time of the flight (Figure 3). Jesse and Nick arrived at COMP 101 shortly after 1000, and waited for Kaleb (UTA & student) and William Weldon (GTA) to unlock the building and give us the keys to the truck. Jesse confirmed again that the METAR showed flyable weather before heading to the field (Figure 4).

Figure 3: METAR and TAF as of 0835 ET 11/8/20

Jesse and Nick arrived at Martell at 1106 EDT, and began to set up the M600. The trees over both plots showed a significant decrease in foliage since the last time crew two was at Martell to fly. The Northwest plot looked particularly bare and discarded leaves littered the forest floor (Figure 5). For the first flight over the Northwest plot, the orange batteries were used, with new settings for Measure being 85% side overlap, and 75% front overlap. The shutter was at 1/2500, and there were some wind gusts coming from the south unreported on the KLAF METAR. Takeoff for the first flight occurred at 1131 EDT and landed at 1155 EDT for 24 minutes of flight time. The batteries had started flight with 96% power reported, ended at 13% power, and were draining rather quickly inflight.

Figure 4: The Northwest plot’s foliage has decreased significantly

 

The Northeast plot was similarly bare, but since the grounds are more regularly maintained there were fewer leaves on the forest floor (Figure 6). For the second flight over the Northeast plot, the green batteries were used, with a shutter speed of 1/2500, 80% front overlap, and 80% side overlap. Takeoff was at 1202 EDT and the M600 landed at 1225 EDT. The batteries were at 95% on takeoff, and finished with 16%. There were a few planes flying well above our activities a few thousand feet away. Jesse noticed how the Measure app was lagging behind in reporting the M600’s location along the flight plan, as well as dots representing images taken were missing on the app as well (Figure 7). The latter is common and not cause for concern. However, due to a lack of confidence in Measure to accurately report battery percentage, Jesse and Nick took manual control and landed the drone before the final pass over the Northeast plot. A second flight was not performed over the Northeast plot because the remaining battery set was not one of the new ones and would not have handled an entire flight over the plot and the last few passes overlap with the Northwest plot anyway.

 

Figure 5: iPad displaying Measure App when 25% battery warning sounded

After completing the flights, Jesse and Nick returned to COMP 101 and returned the equipment and the truck. After a preliminary look over the data, everything appeared normal (Figure 8). Jesse and Nick offloaded every bit of data into the shared folder for Cole to remotely process. This time to avoid any issues with possibly deleting trigger files, no folders or files were deleted. Due to the flight being late in the week and a busy Monday schedule, processing has not yet been completed yet but will be done on Tuesday.

Figure 6: Image taken on the α6000 above the launch site