A Completely Autonomous sUAS Drone
Troubleshooting of power distribution PCB after analyzing data logs from flight computer
Debugging of auto pilot code using wireless communication protocol in the field
Drone being assembled and wired using a custom made wiring harness
Structural failure of upper center plate after a harsh landing
Final assembled drone in background with the small scale proof of concept in the foreground
Project Overview
Design, build and test an autonomous drone to compete at the annual UAS Showcase sponsored by Raytheon Intelligence and Space
The drone will have to autonomously complete three challenges
Lift off from the zero-yard line on a football field, detect the 30-yard line and land as close as possible to it
Lift off from the 30-yard line and find a unique logo anywhere on the football field, navigate towards it and land directly on the logo
Navigate through an obstacle course in as little time as possible
Additionally, I was the hardware design lead for the team (a total of ~14 people) so my responsibilities included:
Design of the overall powertrain of the drone
Becoming FAA Part 107 certified
Development of the fully parametric CAD model for the 3D-printed drone frame
FEA simulation of the mechanical loading on the drone during landing
Flight controller programming and tuning for stability and maneuverability
Image recognition hardware selection
GPS/RTK hardware selection, setup and testing
Managing the team of other engineers and ensuring deadlines were met
Combination of independent development work and cross-functional collaboration with Raytheon
Challenges:
Structural optimization of the drone frame to survive testing
Stability of drone in high wind environments
Converting the vision data into a form that allowed for precision landing
Tilt mechanism to move camera from downward facing to forward facing
Obstacle course maze solving algorithm implementation
Long range communication
RTK GPS setup
Team management at times
Results:
The design won 2nd place overall at the UAS showcase
The drone completed all 3 challenges without any human intervention