drone-pathgen

overview

drone-pathgen is a lightweight python wrapper around gcopter – a trajectory optimizer for multicopters. it lets you generate smooth, collision-free drone paths directly from python swiftly and painlessly.

applications

• autonomous delivery
• search & rescue navigation
• infrastructure inspection
• simulation & research prototypes
• indoor warehouse flight planning

features

• smart path planning that avoids obstacles and respects real drone physics
• voxel-map support for arbitrary 3d environments
• smooth, jerk-limited trajectories within safe-flight-corridors (sfc)
• built-in optimization for time, energy and feasibility
• numpy-friendly api – results come back as plain ndarrays
• optional open3d visualization extras [see screenshot below]

installation

prerequisites – make sure ompl, eigen3 and boost are available on your system.

# macos [homebrew]
brew install ompl eigen boost

# ubuntu / debian
sudo apt-get update
sudo apt-get install libompl-dev libeigen3-dev libboost-all-dev

install the library:

core only

uv add git+https://github.com/u-k-g/drone-pathgen.git
# or with pip 
pip install git+https://github.com/u-k-g/drone-pathgen.git

with visualization support [adds open3d]

uv add "git+https://github.com/u-k-g/drone-pathgen.git[viz]"
# or with pip
pip install "git+https://github.com/u-k-g/drone-pathgen.git[viz]"

supported python versions [3.8 → 3.12]

quick start

import numpy as np
import gcopter_cpp as gc

# create api and configure environment
api = gc.GCopterAPI()
api.configure_map(
    map_size=np.array([20, 20, 10], dtype=np.int32),
    origin=np.array([-5.0, -5.0, 0.0]),
    voxel_scale=0.5,
    obstacle_points=[np.array([0.0, 0.0, 2.0])],
    dilation_radius=2
)

# set trajectory endpoints
start = np.array([[-4.0], [-4.0], [1.0]])
goal = np.array([[4.0], [4.0], [3.0]])
api.set_endpoints(start, goal)

# plan trajectory
success = api.run_inference(
    planning_timeout=5.0,
    time_weight=50.0,
    segment_length=2.0,
    smoothing_epsilon=1e-3,
    integral_resolution=8,
    magnitude_bounds=np.array([5., 10., np.pi/3, 5., 15.]),
    penalty_weights=np.array([1, 1, 1, 1, 1]),
    physical_params=np.array([1., 9.81, 0., 0., 0., 0.01])
)

# get results
if success:
    stats = gc.TrajectoryStatistics()
    api.get_statistics(stats)
    print(f"✅ planned {stats.total_duration:.2f}s trajectory")

documentation

comprehensive documentation available in docs/

• installation guide - system setup and troubleshooting
• user guide - getting started and core concepts
• api reference - complete method documentation
• examples - 9 practical code examples and use cases

working examples in examples/

• examples/basic_pathgen.py - core trajectory planning workflow
• examples/visualization.py - 3d visualization with open3d

license

released under the mit license – see license file for details.