AT409 Week 7 Report

Introduction

            This week from October 4th to October 10th, our group conducted two series of flights. Two M600 flights over the Martell forest northwest and northeast plots, and 6 Mavic 2 flights over the Purdue Wildlife Area (PWA) for Weldon’s research project. We conducted a two-man mission during the M600 flight and made notes on maintenance needed for the aircraft and began to switch up our roles for the PWA flights from the previous weeks.

October 6^th – PWA Search & Rescue

PIC	Ethan Hoke
SO	Kaleb Gould
Search (Manual)	John Cox (flights 1-4), Jeff Hines (flights 5-6)
Search (Loc8)	Jeff Hines (flights 1-4), John Cox (flights 5-6)
Recovery Team	Ethan Hoke & Kaleb Gould
Search Coordinator	William Weldon

Table 1: October 6^th Crew Roster

Our group along with one AT 209 student, Ethan Hoke, assisted Weldon in his ongoing research project. Once set-up at the search area, we were assigned our various tasks as seen in table 1. After we began the days flights It was quickly noted that John was able to identify the missing person very quickly for the reasons mentioned below. As table 2 shows, in some cases John was able to identify the missing person in under one minute. This was the primary reason for the switch in roles that would come later. It is worth mentioning that Jeff and John swapped rolls about halfway through the mission, Jeff previously being on Loc8 swapped to a manual photo scan and John from a manual scan to Loc8.  Previously, John had exclusively preformed manual searches by scanning photos the drone took and attempting to visually identify the missing person while Jeff was using the Loc8 software to identify the missing person’s clothing. Weldon had hypothesized that since John had been preforming the manual search of the same area several times, he started to develop a sense of familiarity with the area. Consequently, it had become much easier for him to find the missing person as he had a mental image of what the search area was supposed to look like, quickly identifying elements that had been misplaced or had changed. This obviously would not be the case in a real-world search and rescue operation. Therefore, during the last 2 flights John was reassigned to the loc8 search and Jeff was assigned to the manual search.

 

 

 

	Take Off	Landing	Data Transfer	Search Start	Discovery (Loc8)	Discovery (Manual)	Dispatch Sent	Recovered
Flight 1	9:30	9:42	2 minutes (9:46 to 9:48)	9:50	9:50	N/A	9:51	9:56
Flight 2	10:09	10:21	3 minutes (10:22 to 10:25)	10:27	10:30	10:28	10:29	10:32
Flight 3	10:49	11:01	8 minutes (11:02 to 11:10)	11:11	11:15	11:18	11:16	11:20
Flight 4	11:30	11:41	2 minutes (11:43 to 11:45)	11:46	11:53	11:49	11:50	11:56
Flight 5	12:08	12:20	3 minutes (12:22 to 12:25)	12:26	12:30	12:33	12:33	12:33
Flight 6	12:45	12:57	2 minutes (12:58 to 1:00)	1:01	1:09	N/A	1:09	11:11

Table 2:

October 6^th Important times from the mock search and rescue missions

The flights that followed were carried out efficiently, by now all of us involved know the general process of this mission well. We were able to achieve quick turnaround times between flights and whole day proceeded smoothly. This is one of the factors that allowed us to do as many flights that day as we did.

There was noticeable change in the search times after John and Jeff switched rolls. Search times became longer as Jeff did not have the advantage of familiarity with the search area that John did nor did John have the experience using the Loc8 software that Jeff possessed. There was also an instance where Loc8 gave an indication that John interpreted as a false positive that turned out being a true positive. While the search times were longer, it’s important to note that those flights give a more accurate representation of real-world operations as it is highly unlikely that volunteer search and rescue workers will have extensive familiarity of the search area nor be experienced in using the Loc8 software.

 

October 8^th – M600 Data Collection:

            John and Jeff conducted an M600 mission on Thursday October 8^th from 12:00 to 14:00. Both members arrived at COMP 101 close to 11:00 in order to prepare for dispatch. Kaleb, unable to make the flight for that day, aided by charging the batteries and ensuring all the equipment needed for the flight was present earlier in the day. As a precaution, Jeff took the SD cards to NISW 145 to ensure that each card was properly cleared and formatted while John packed the M600 and battery cases. One SD card still had flight data stored on it, so Jeff transferred the images into their proper folder within the data dump, using the time stamps as a guide to determine where the photo sets go. Once all the data was stored properly, Jeff wiped the SD cards and stored them in their sensor. At this point, all preparations were complete, so John and Jeff loaded their car with the M600 and necessary flight equipment then left for Martell. Due to a lack of space, they were unable to bring the full case, opting to take the aircraft and the entirety of the foam top since there was space. The crew departed from COMP 101 around 9:00, a bit behind of schedule.

Northwest Plot:

Upon arriving, John and Jeff promptly unloaded the equipment and took some time to discuss preliminary questions such as crew roles and which plot to fly first. John offered to be PIC and Jeff remained in a familiar role as SO. Since the crew was down a team member, they figured keeping roles as close to standard as possible would make for a smoother, more familiar experience. With questions answered, the flight crew followed the pre-flight checklist, confident in their movements and actions, and began the mission with no issues. Jeff transitioned from an SO to a VO at the start of the mission, making sue to mark down takeoff (12:40) and landing (1:01) times for the metadata and maintaining consistent line of sight. John kept a close eye on the tablet, looking for any deviation in altitude, speed, or photos taken. At the start of each leg, the PIC called out the start and announced when halfway through the mission, three-fourths through, and when on the last leg. No abnormalities in the aircraft or the flight area were observed during the flight. Once the aircraft had finished and returned to home, John brought down the aircraft, calling out the altitude every 100 feet descended and took manual control around 50 feet. The M600 set down gently and promptly shut down. Jeff ended the PPK session and the team prepared for the second flight

Photo 1:

Photo of the flight plan for the Northwest plot of Martell

Northeast Plot:

After finishing the first flight, John and Jeff replaced the batteries from the blue set to the brown set. While John set-up the next mission, Jeff double checked all the sensors to ensure nothing had come lose or changed during the flight. The only noted difference was the shutter speed had reverted from 1/3200 to 1/4000. Since the aircraft powered down, the camera shut down as well, hence the change. The shutter speed was adjusted back to 1/3200. Once John had prepared the mission, the crew referenced the checklist to ensure that nothing was skipped in preparation. Once verified, John took off at 13:14 and climbed to 500 feet. Once the mission began, both crew members devoted their full attention to their tasks, Jeff - acting as a VO and calling out any abnormalities & John – monitoring mission progress and flight characteristics. The mission proceeded with no interruptions until the last leg of the flight. John received a battery error on the tablet. This error prevented the tablet from updating the speed, altitude, or photos collected. John immediately called out the disturbance and Jeff watched the aircraft looking for any deviations or unusual behavior. Before long, the mission ended, and the aircraft returned to home. John was still able to initiate an automatic descent, though around 150 feet he took over manual control of the aircraft and set it down manually. The aircraft laned at 13:38. Both members were unfamiliar with the error, so they proceeded with immense caution during the last leg and descent. The crew consulted Weldon and Kaleb, two trusted authorities on the Purdue UAS fleet, and hypothesize the error occurred due to the batteries having gone through too many cycles. A bit shaken, both crew members tore down the aircraft and flight area, stored the equipment, and started back toward Niswonger.

Photo 2:

Photo of the flight plan for the Northeast plot of Martell

            Post Flight:

John and Jeff returned to Niswonger at 2:20, a bit past their expected time of two hours. After arriving, they split up the tasks for efficiency. Jeff gathered all the SD cards to begin the data transferring and analysis while John stored the aircraft and put the batteries on charge. The door to COMP 101 was locked which set back the plan, but soon they were able to flag down Zach and get into COMP 101. Jeff began the data transfer at this point. John finished his duties before long, placing the blue batteries on charge. Both sets would have gone on the charge, however one set had not charged while crew 1 flew their mission as the charging station was unplugged. Jeff had finished his data transfer and was about to post process the data but ran into issues finding EZSurv on multiple computers. After logging into and out of multiple computers, he eventually found EZSurv on one of the computers and began to post process the A6000 photos. In total, entire data handling process took about an hour and a half. Before leaving, Jeff returned to COMP 101 to return the SD cards and store the charged batteries in the battery case. Before leaving though, he put the brown batteries on the charger and returned the battery case to its designated spot.