Skip to content
Projects

VANdexj/OmniRisk has 65 GitHub stars and 9 forks. GitHot measured 0 stars in 24 hours, +2 in 7 days, and +7 in 15 days. Last checked 2026-11-19 17:48 UTC. Last push 2026-11-16.

How these counts are measured

Measured star growth

in 24h
0
in 7 days
+2
in 15 days
+7
in 1 month
Collecting history

Net star changes over overlapping periods, not daily totals. Missing measurements are left blank. Stars measure attention, not software quality.

About

OmniRisk: omnidirectional trajectory-risk learning for agile quadrotor dynamic avoidance. One forward pass scores every maneuver onboard; real-world evasion at up to 15 m/s.

Machine Learning

README.md

Original
OmniRisk: omnidirectional trajectory-risk learning

Omnidirectional Trajectory-Risk Learning for Agile Quadrotor Dynamic Avoidance

Yifan He†1 · Yang Liu†1,2 · Wenhao Zhao†1 · Hai Lin1 · Deping Zhang1 · Mingze Ma1 · Xin Zhou1
Fei Gao1,2 · Huan Yu*1,2 · Zipeng Dai*1 · Ziming Ding*1
1Differential Robotics · 2Zhejiang University   †equal contribution · *corresponding

Video arXiv Paper · coming soon BibTeX

arXiv Paper License ROS Jetson

OmniRisk real-world flight: the quadrotor consecutively dodges thrown balls

OmniRisk moves trajectory-risk evaluation offline. Onboard, one forward pass scores every candidate
maneuver around the drone, so planning cost does not grow with the number of obstacles.


360° omnidirectional perception & action · 15 m/s relative encounter speed · 36 → 1 maneuvers scored in one forward pass · 30 + 4 collision-free trials · 0 real-world fine-tuning

Demos

Repeated trials

Single-ball tests: the quadrotor dodges a thrown ball in each of 30 trials; a counter shows balls launched
Single-ball tests
30 trials, all collision-free
Multi-ball tests: the quadrotor dodges several balls thrown in succession; a counter shows balls launched
Multi-ball tests
4 trials, 12 balls in total, all collision-free

Open vs. constrained space

Top view, slowed to 0.2x: in open space the quadrotor dodges a thrown ball
Open space
top view · ×0.2 · nothing nearby, free to dodge in any direction
Top view, slowed to 0.2x: with obstacles on both sides the quadrotor escapes along the free corridor
Constrained space
top view · ×0.2 · obstacles on both sides, escape stays in the free corridor

Dynamic + static obstacles

A ball is thrown across the quadrotor's path as it flies between two rock-shaped obstacles; it dodges without touching the rocks
Crossing ball among rocks
dodges a ball thrown across its path while clearing rock-shaped obstacles

Different thrown objects

The quadrotor dodges a thrown ball
Thrown ball
motion trails show the ball's arc and the escape path
The quadrotor dodges a thrown tennis ball
Tennis ball
small and fast, only a few LiDAR returns
The quadrotor dodges a thrown dumbbell
Dumbbell
irregular rigid body, tumbling in flight
The quadrotor dodges a thrown plush toy
Plush toy
soft, deforms mid-air

Results

84% vs 8% success with 6 obstacles at 10 m/s (vs FAPP) · 20 ms network inference on Jetson Orin NX · 90.7% real-world success indoor, n = 150 · 86.7% real-world success outdoor, n = 150
Success rate vs. obstacle speed, 6 obstacles, n = 50. OmniRisk 96 / 90 / 84% at 2 / 6 / 10 m/s; FAPP 74 / 40 / 8%; SIMP 66 / 0 / 0% Input ablation, success %, n = 50. Head-on: depth only 62, depth + mask 88, full tensor 96. Crossing: 22, 30, 92. Rear: 26, 34, 94

Planning latency on Jetson Orin NX (6 obstacles)

Method P95 (ms) Deadline miss
SIMP 2110.9 96.3%
FAPP 17.6 0.0%
OmniRisk 45.6 3.0%

Zero-shot real-world success (n = 50 per cell)

Object Indoor Outdoor
Tennis ball 94.0% 90.0%
Foam basketball 86.0% 82.0%
Plush toy 92.0% 88.0%
All types 90.7% 86.7%

[!NOTE] FAPP plans faster (~17 ms) but refines a single trajectory; with 6 obstacles at 10 m/s its success drops to 8% vs. 84% for OmniRisk.

Hardware

OmniRisk quadrotor platform
Component Model
LiDAR Livox Mid-360
Onboard computer NVIDIA Jetson Orin NX
Flight controller NxtPX4v2
Odometry FAST-LIO2
Moving-object detection M-detector

Roadmap

  • Training code, simulator & dataset generator
  • ROS planner & real-world deployment guide
  • Pretrained models
  • Simulation demos
  • Project page

Quick start

Tested on Ubuntu 20.04 + ROS Noetic + CUDA (x86) and Jetson Orin NX (JetPack 5.1.x).

Repository structure
OmniRisk/     learning-based planner: network, training and the ROS planning node
Simulator/    CUDA LiDAR / depth simulator and dataset generator
Controller/   SO(3) quadrotor dynamics and controller for simulation
ros_ws/       onboard stack: FAST-LIO, M-detector, px4ctrl, EKF, Livox driver
docker/       simulation (x86) and deployment (Jetson) images
scripts/      one-shot install and build
Installation
bash scripts/build.sh               # install dependencies + build
bash scripts/build.sh --build-only  # rebuild only
bash scripts/build.sh --no-sim      # skip Controller / Simulator
Pretrained model
mkdir -p OmniRisk/saved/run_1
wget -O OmniRisk/saved/run_1/epoch60.pth \
  https://github.com/VANdexj/OmniRisk/releases/download/v1.0/omnirisk_epoch60.pth

Then run with --trial 1 --epoch 60 (trt_transfer.py uses this checkpoint by default).

Simulation
# T1 Dynamics
source Controller/devel/setup.bash
roslaunch so3_quadrotor_simulator simulator_attitude_control.launch

# T2 Environment + LiDAR
source Simulator/devel/setup.bash
rosrun sensor_simulator sensor_simulator_cuda

# T3 M-detector
source ros_ws/devel/setup.bash
roslaunch m_detector detector_mid360.launch points_topic:=/lidar_points points_in_world_frame:=false odom_topic:=/sim/odom

# T4 OmniRisk (add --flight_mode nav, then set the goal with RViz 2D Nav Goal)
source ros_ws/devel/setup.bash && source omnirisk_env/bin/activate
cd OmniRisk && python omnirisk_node.py --trial <n> --epoch <k> --odom_topic /sim/odom --lidar_topic /lidar_points

# T5 RViz
source ros_ws/devel/setup.bash
rviz -d OmniRisk/omnirisk.rviz

# T6 Joystick goal
source ros_ws/devel/setup.bash
cd OmniRisk && python3 rc_goal_teleop.py _odom_topic:=/sim/odom
Training
cd Simulator && source devel/setup.bash
rosrun sensor_simulator dataset_generator   # → dataset/

cd ../OmniRisk && source ../omnirisk_env/bin/activate
python train.py                             # → saved/run_<n>/epoch<k>.pth
tensorboard --logdir saved
TensorRT acceleration
cd OmniRisk
python trt_transfer.py --trial <n> --epoch <k>   # run on the deployment device
python omnirisk_node.py --use_tensorrt 1
Real-world deployment
# In every terminal
source ros_ws/devel/setup.bash

# T1 MAVROS
sudo chmod 666 /dev/ttyACM0
roslaunch mavros px4.launch fcu_url:=/dev/ttyACM0:57600

# After MAVROS connects: IMU / attitude at 200Hz
rosrun mavros mavcmd long 511 105 5000 0 0 0 0 0  # HIGHRES_IMU
rosrun mavros mavcmd long 511 31  5000 0 0 0 0 0  # ATTITUDE_QUATERNION
rosrun mavros mavcmd long 511 83  5000 0 0 0 0 0  # ATTITUDE_TARGET

# T2 MID360 + FAST-LIO + EKF
roslaunch fast_lio fastlio_ekf.launch

# T3 px4ctrl
roslaunch px4ctrl run_ctrl.launch

# T4 M-detector
roslaunch m_detector detector_mid360.launch

# T5 OmniRisk
source omnirisk_env/bin/activate
cd OmniRisk && python omnirisk_node.py --use_tensorrt 1

# T6 RViz
rviz -d OmniRisk/omnirisk.rviz

# T7 Joystick goal (_input_mode:=usb | mavros)
cd OmniRisk && python3 rc_goal_teleop.py _input_mode:=usb

# Takeoff / land
rostopic pub -1 /px4ctrl/takeoff_land quadrotor_msgs/TakeoffLand "takeoff_land_cmd: 1"
rostopic pub -1 /px4ctrl/takeoff_land quadrotor_msgs/TakeoffLand "takeoff_land_cmd: 2"

# Checks
rostopic hz /ekf_quat/ekf_odom              # > 50Hz
rostopic hz /mavros/setpoint_raw/attitude   # > 100Hz
rostopic hz /so3_control/pos_cmd

Citation

@article{he2026omnirisk,
  title   = {OmniRisk: Omnidirectional Trajectory-Risk Learning for Agile Quadrotor Dynamic Avoidance},
  author  = {He, Yifan and Liu, Yang and Zhao, Wenhao and Lin, Hai and Zhang, Deping and Ma, Mingze and Zhou, Xin and Gao, Fei and Yu, Huan and Dai, Zipeng and Ding, Ziming},
  journal = {arXiv preprint arXiv:2609.18191},
  year    = {2026}
}

Acknowledgements

This work was carried out at Differential Robotics; we are especially grateful to Zipeng Dai for his invaluable mentorship.

License

Released under the MIT License.