Using Gazebo to visualize, validate, and simulate autonomous flight before real-world testing.
The Simulations subteam builds the virtual world where our autonomous drone learns to fly. Using Gazebo, we create realistic 3D environments that replicate competition arenas, terrain, and obstacles — allowing the entire autonomy stack to be tested safely before any real flight.
Our work dramatically reduces the risk of crashes and dramatically speeds up iteration: bugs in the control stack can be caught and fixed in simulation at near-zero cost.
Building accurate Gazebo models of the SUAS and C-UASC competition arenas — including waypoint fields, obstacle configurations, and target placement zones — so the team can rehearse full mission runs.
Creating a high-fidelity URDF / SDF model of our physical drone with accurate mass, inertia, motor thrust curves, and aerodynamic drag so simulated behavior matches real flight.
Setting up Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HIL) pipelines with ArduPilot / PX4 so the flight controller firmware runs against the simulated plant in real time.
TO BE FINISHED LATER — detail current SITL setupValidating planned flight paths and waypoint sequences in simulation before they are uploaded to the real vehicle, checking for obstacle conflicts and mission timing.
Our primary simulation stack:
Major
Year · Simulations Lead
TO BE FINISHED LATER — add team member cards
Computer Science
Focus: World modeling
Computer Engineering
Focus: SITL / HIL
Any major
Focus: Flight path testing
Simulations meets for world building, SITL testing, and integration with the autonomy stack.
TO BE FINISHED LATER
TO BE FINISHED LATER