By

Setup and Flight Testing - Day 1 - USC UAV Competition

The morning today was mostly dedicated to assembly and preparation of systems, our safety briefing, and our systems presentation. The preparation for the presentation was done in a rush, and we did not go over the technical details of our systems and our team’s project management as much as we would have liked.

We ran into a host of unexpected issues almost as soon as we finished assembling our plane and avionics box; as a result, it took a good part of the day to successfully receive and transmit data from the plane and demonstrate our failsafes in order to pass our flight readiness reviews and be cleared for flight tests. First, when we went to do our flight readiness review, none of our kill switches worked – because the lack of GPS was already killing the plane, but we didn’t realize this until later. Then there was a bad connection in the IMU followed by a mysterious issue with one of the power supplies within the plane cutting out. We don’t know yet whether this was a short or some other cause, because when we disassembled the avionics box to replace the power supply, it started working again and has been fine since.

Once we got to the field, we were able to perform a calibration of our IMU, a demonstration of our failsafes, and a successful manual flight with transmission of telemetry data. The flight, however, was not smooth; the plane’s ascent was choppy, and the pilot (Cameron was barely able to control it). After three failed landing attempts – one where the plane nearly exited the flight area while at only 10 m, we eventually landed the plane with no damage. We determined that the cause of this brutal flight was twofold: the deflection of the elevator and aileron control surfaces was much too too high (the deflections in these reached their maximums at only 50% input), and the COG was a little bit tail-heavy. We were easily able to fix both of these issues and should be able to fly properly tomorrow.

Work on imaging systems has also been encountering numerous issues, including problems with wi-fi connection of the imaging computer to the PandaBoard, offsets returned by ground imaging target selection algorithm too large to be reasonable, and issues with loading images and data from watched directories. As we now have basic flight functionality, we plan to focus most of our work tomorrow on imaging.

Pictures and Video Below…