Explore

End of Semester Review

Given the name of this course, I expected to learn about UAS aircraft maintenance and the like. Having talked to Professor Rose in the past as well as Lercel in the previous semesters, I had a fairly decent idea of what to expect. I've been told before that 209 is about learning how to build aircraft and is a more hands-on class than 109 was, so building the Believers was relatively expected.

I didn't expect to be working with any sort of proprietary software, however, and that was probably the biggest surprise or difference from my expectations, and that probably enhanced my experience in this class, seeing as the two ways the industry seems to be leaning are fully prebuilt closed systems like DJI, Skydio, Autel, or other system integrators, if you will, and then the others, which are like Windracers, Freefly Systems, and Wingtra, which use primarily off-the-shelf components but run custom firmware.

Additionally, I didn't expect the labs to be so flexible—it seems like there were a lot of changes made on the fly to reflect where we were at as opposed to a more rigid curriculum.

Overview of the semester:

Some of the significant things we did in the early labs were familiarizing ourselves with the Believer airframe, creating an example build process, cataloging all of the parts and learning what they do, and creating an example build guide based on what we had read on a blog. Our next significant lab for the Believer was the DATX setup and modification, where we modified the existing controller's hardware and software to Windracers’ standards—we removed the extra switches and re-flashed the controller, as well as binding and soldering a new receiver. The modifications we did to the controller were to reduce complexity and remove unneeded hardware switches. Next, we set up the RFD900X, got it to communicate with the computer, and linked it to the ground and air modules. We tested and verified that those two modules could communicate with each other, then installed firmware to the flight controller, as the Believers use closed-source software. This was a particularly interesting process, as we installed it through Mission Planner but were unable to use any of the regular Mission Planner tools aside from flashing. At this point, we’re getting to pretty recent labs—and this is where the build sort of started—with the Believer avionics and parameters and build guide labs. These labs focused on connecting the cube flight controller to the ground control station as well as connecting the transmitter and receiver to the flight controller. We used the AIDA PC to connect to and configure the Cube Pilot flight controller with a cable and then used the telemetry modules we configured before to establish a wireless link between the ground control station and flight controller. This went relatively smoothly—and we were able to modify the serial X parameters. We also configured the airspeed sensor and GPS in this lab. In the next lab we soldered a header pin to the receivers, updated the transmitter to use the backpack module, and altered the flight controller's software, as it wasn't set up properly. This was our first instance of real troubleshooting, with the version and parameter files having some sort of discrepancy that caused an error. The new file we were given was also problematic, but we ultimately resolved those issues by using a different path to upload the file. This new file seemed to work fine once uploaded, but the GPS, which had been working previously, didn't now. We fixed the GPS issue by flashing a new firmware, and with the different firmware, the GPS parameters were changed—then we reinstalled the Windracers firmware. This brings us to the most recent and final labs—the motor and controls lab is where we finalize calibration and setup of the DATX and flight controller system. We were able to validate our mapping and control parameters in the real world—attaching servos and a brushless motor and ESC to our flight controller. The final lab of this semester was assembling the Believer airframe and beginning to solder the components together—although we didn't complete the assembly of the Believer, we are well on our way to having a finished aircraft in a couple more lab sessions.

Building from scratch

Obviously there's a lot more setup and tuning for a system like this as opposed to the preconfigured plugin-and-play systems like DJI—although this did give us a lot more flexibility as far as payload, the components that go into this aircraft, and the software that we would want to use—currently we're using the WinRacers software, but nothing is stopping us from using a mission planner or ArduPilot firmware on this flight controller. I feel that this is probably more representative of a UAS engineering workflow than the DJI or other similar systems would be, as those are great platforms for things like surveying or image capture—but they are relatively rigid in their applications or are limited in their form factor—so learning how to use a system like this is a lot more valuable and versatile for engineering design and experimental applications. There's a lot more troubleshooting with systems like these, and things don't just “work” as they do in the more plug-and-play systems.

Skill growth and technical learning

I'm not sure if I developed any completely new skills, but I certainly strengthened existing ones—especially soldering. It was interesting to go back to hardware assembly and kind of build out the nervous system, if you will, of the aircraft on a table and then put the shell together and install it. I think the primary skill that I developed in this lab was the ability to troubleshoot—particularly in this closed-source, relatively low-documentation firmware system. Learning how to use the diagnostic tools and firmware management things available was probably the most valuable skill I built upon. Of course, as with any lab, working in a group was both the greatest strength and greatest weakness of a project like this—everyone has different amounts of knowledge, different strengths, and different weaknesses, but ultimately I think our group worked well together to overcome the issues with the ever-changing resources available to us, the semi-complete versions of firmware we were given, and the many obstacles that seemed to happen for the first time in our lab session. This is probably why our group worked so well together, as we were able to “divide and conquer” and shuffle members around to learn different things when we had the time as well as put the members who are strongest at search and tasks where they needed to be during crunch time.

What surprised me

I think the curriculum and actual work in labs was easier than I expected, though I think the lecture and quiz content was more in-depth or more challenging than expected. But the thing that was most educational to me was probably working in a group and seeing what most people seem to struggle with and what seemed to come naturally—it really showed me the way that most people seem to approach problems, and that really helps me to better understand how to improve my teamwork and how I can contribute best to my group in a field that I'm fairly versed in. I don't think I quite had any moments where things finally “clicked,” but when we finally started assembling the airframe, it felt like everything that we had done up to this point was finally coming together.

What I'm most proud of

I think a skill that I'm proud of is my soldering—Rose noted that my solder joints were among the cleanest that he had seen from a student, and because I was also also able to teach that skill, it makes me more proud of it. I'm not sure about a problem I helped solve—most of the problem-solving was done as a team or was aided with new information or new resources—but I think some of the little tweaks I provided in assembling the airframe were probably problems that would have needed to be solved in the future. I think the thing that I contributed to my group the most was probably taking the time to teach some of them how to solder or how to brush up on certain skills that I think they would have liked to spend more time on in the skill-building labs—like teaching Hunter how to improve his soldering.

Overall Takeaways

This course definitely improved my confidence with UAS hardware, particularly that working with it, as much as it can be challenging and frustrating, is still more or less as I remember it—and that even for smaller systems like this, there are still resources out there that can provide you help and troubleshooting steps. I'm definitely taking all of my reinforced skills and particularly the lessons I learned in teamwork and group work into every future course, internship, and project—at least in aviation—but more specifically I will probably be taking this experience with closed-source software more to heart, as I may have the opportunity to work on or work with a Freefly Systems drone, which seems to be somewhat similar in approach to the way that Racers is treating the firmware on the Believers.

As for an overview of the semester, I think that this was definitely a much more engaging class than most of the UAS courses. As much as I enjoy flight labs and learning of the ways of the industry, nothing really makes me learn the way that hands-on things do, particularly building things, so a lab like this is really exciting and engaging to me. Even though I have somewhat of a background in building aircraft and tinkering with electronics in general, this is different since it was being built as a group, it's being built to a certain specification, and we have more of a guide to follow.

Want to print your doc?
This is not the way.

Try clicking the ··· in the right corner or using a keyboard shortcut (

CtrlP

) instead.