Multirotor, fixed-wing and VTOL flight models with the physics written out in full — plus 42 runnable simulations and a gym for teaching a drone to fly itself.
Seventy-eight seconds: nine algorithm families in depth, then every one of the 42 simulations in the library. Every frame is simulated by scripts/make_promo.py at render time — the same models this library ships, integrated live. No stock footage, and nothing that can drift out of sync with the code.
pip install flybots
flybots doctor # verify the install, run the physics self-checks
flybots list # browse 42 simulations
flybots run pid_hover # render one to a GIF
flybots train hover # teach a quadrotor to hold positionOr drive it from Python:
from uav_sim.vehicles.fixed_wing import create_fixed_wing, FixedWingPreset
from uav_sim.control.fixed_wing_autopilot import FixedWingAutopilot, AutopilotCommand
aircraft = create_fixed_wing(FixedWingPreset.SKYWALKER_X8)
aircraft.reset_trimmed(altitude=120.0) # solve for equilibrium flight
pilot = FixedWingAutopilot(aircraft.fw_params) # gains derived from the airframe
command = AutopilotCommand(altitude=160.0, airspeed=20.0, course=1.0)
for _ in range(12_000):
aircraft.step(pilot.compute(aircraft.state, command, 0.01), 0.01)
print(aircraft.state[2]) # 160.0No solver is wrapped. No dynamics library is imported. The Newton-Euler equations, the aerodynamic coefficient build-up, the Kalman recursions and the sampling-based planners are all written out, in NumPy, next to the citation they came from.
That is the point: the code is meant to be read. If a line needs a comment to explain what it does, something is named wrong; if it needs one to explain why, it has one.
A worked example
The fixed-wing model is the best illustration. It carries thirty-odd stability derivatives, and every one of them changes the aircraft's behaviour — there is a test that fails if you zero any of them.