News!

29/6/2019 – We release the source code of the Teach-Repeat-Replan framework, which is a complete and robust autonomous drone system for aggressive flight in complex environments. This project supports fully Autonomous Drone Race.

Check the code here

If you feel this repo useful, please star it or cite our related publications:

• Teach-Repeat-Replan: A Complete and Robust System for Aggressive Flight in Complex Environments, Fei Gao, Luqi Wang, Boyu Zhou, Luxin Han, Jie Pan, shaojie Shen, submitted to IEEE Transactions on Robotics (T-RO).
• Optimal Trajectory Generation for Quadrotor Teach-and-Repeat, Fei Gao, Luqi Wang, Kaixuan Wang, William Wu, Boyu Zhou, Luxin Han, Shaojie Shen, IEEE Robotics and Automation Letters (RA-L), 2019.

This project contains local online mapping, global mapping, local online planning, global spatial-temporal planning, controller, visual-inertial localization, global pose graph optimization, human-robot interface, and a complete simulator.

Our system can be applied to situations where the user has a preferable rough route but isn’t able to pilot the drone ideally. For example, for drone racing or aerial filming, a beginner-level pilot is impossible to control the drone to finish the race safely or take an aerial video smoothly unless months of training. With our system, the human pilot can virtually control the drone with his/her navie operations, then our system automatically generates a very efficient repeating trajectory and autonomously execute it.

Our system can also be used for normal autonomous navigations, like our previous works in video1 and video2. For these applications, the drone can autonomously fly in complex environments using only onboard sensing and planning.

 

23/6/2019 – I have three papers accepted by IROS 2019 (two as the corresponding author), check the publication list for these fresh new works!

Bio

My name is Fei Gao (高飞) and I’m from China. The meaning of my name in Chinese is: flying high in the sky : )

I’m a Ph.D. candidate working on Aerial Robot in the UAV Group, RI, HKUST. I got my Bachelor degree in Control Science & Engineering from Zhejiang University. My research area includes motion planning, control, applied convex optimization, and swarm. I’m also interested in robotics exploration and autonomous aerial videography. You can check the following publication and video lists for details about each project.

Currently, the expected date of my Ph.D. graduation is Jul/Aug. 2019. After that, I’m eager to continue my academic career because of my interests and passions. So HRs please ignore me since I have no such a plan to enter into the job market.

I’m open to all possible faculty or post-doc or other research positions.

Source Code

Teach-Repeat-Replan by Fei Gao and Boyu Zhou

A complete and robust system containing all components for UAV aggressive flight in complex environments. It is built upon on the classical robotics teach-and-repeat framework, which is widely adopted in infrastructure inspection, aerial transportation, and search-and-rescue. Our system can capture users’ intention of a flight mission, convert an arbitrarily jerky teaching trajectory to a guaranteed smooth and safe repeating trajectory, and generate safe local re-plans to avoid unmapped or moving obstacles on the flight.

GTOP: Gradient-Based Trajectory Optimizer by Fei Gao and Boyu Zhou.

A ROS-package for piecewise polynomial trajectory optimization written in C++. Details can be found in the related paper Gradient-Based Online Safe Trajectory Generation for Quadrotor Flight in Complex Environments.

Btraj: Bezier-Curve-Based Trajectory Generator by Fei Gao

Btraj is an online UAV planning framework used to generate safe, dynamically feasible trajectories in previous unknown environments. It can be divided as a front-end path finding module and a back-end trajectory optimization module. In the front-end, we provide two alternates: Fast Marching*(FM*) on a velocity field and A* on a pure grid map. A flight corridor consists of cubes are generated based on the path. In the back-end, we utilize properties of Bezier curve to confine the piecewise Bezier curves entirely within the corridor and dynamical limits. Details can be found in the related paper Online Safe Trajectory Generation For Quadrotors Using Fast Marching Method and Bernstein Basis Polynomial.

TimeOptimizer: Optimal Time Allocator for Quadrotor Trajectory by Fei Gao

TimeOptimizer is a tool to do re-timing (time optimization) of an arbitrary piecewise polynomial-based trajectory (no matter monomial polynomial, Bezier curve, B-spline or others). The objective of this work is to map the original parametrization variable to a new variable (time), with which the trajectory can finish as fast as possible and respect all kinodynamic limits (velocity, acceleration). Details can be found in the related paper Optimal Time Allocation for Quadrotor Trajectory Generation

pointcloudTraj: quadrotor motion planning directly on point clouds by Fei Gao

We present a framework for online generating safe and dynamically feasible trajectories directly on the point cloud, which is the lowest level representation of range measurements and is applicable to different sensor types. We online generate and refine a flight corridor which represents the free space that the trajectory of the quadrotor should lie in. We represent the trajectory as piecewise Bézier curves by using the Bernstein polynomial basis and formulate the trajectory generation problem as a convex program. By using Bézier curves, we can constrain the position and kinodynamics of the trajectory entirely within the flight corridor and given physical limits.

More source code packages for other paper will be released soon.

Publication List

Online Quadrotor Trajectory Generation and Autonomous Navigation on Point Clouds, Fei Gao, Shaojie Shen, 2016 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR 2016), Best Conference Paper Award.

Quadrotor trajectory generation in dynamic environments using semi-definite relaxation on nonconvex QCQP,  Fei Gao, Shaojie Shen, 2017 IEEE International Conference on Robotics and Automation (ICRA 2017).

Real-time Monocular Dense Mapping on Aerial Robots Using Visual-Inertial Fusion, Zhenfei Yang, Fei Gao, Shaojie Shen, 2017 IEEE International Conference on Robotics and Automation (ICRA 2017).

Gradient-Based Online Safe Trajectory Generation for Quadrotor Flight in Complex Environments, Fei Gao, Yi Lin, Shaojie Shen, 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017).

Autonomous Aerial Navigation Using Monocular Visual-Inertial Fusion, Yi Lin*, Fei Gao*, Tong Qin*, Wenliang Gao*, Tianbo Liu, William Wu, Zhenfei Yang, Shaojie Shen, ( * for the equal first author ), 2017, Journal of Field Robotics (JFR).

Collaborative Air-Ground Target Searching in Complex Environments, Changsheng Shen, Yuanzhao Zhang, Zimo Li, Fei Gao, Shaojie Shen, 2017 Symposium on Safety, Security, and Rescue Robotics (SSRR 2017).

Online Safe Trajectory Generation For Quadrotors Using Fast Marching Method and Bernstein Basis Polynomial, Fei Gao, William WU, Yi Lin, Shaojie Shen, 2018 IEEE International Conference on Robotics and Automation (ICRA 2018). (Full Text)

ACT: An Autonomous Drone Cinematography System for Action Scenes, Chong Huang, Fei Gao, Jie Pan, Shaojie Shen, Kwang-Ting (Tim) Cheng et al, 2018 IEEE International Conference on Robotics and Automation (ICRA 2018).

A Collaborative Aerial-Ground Robotic System for Fast Exploration, Daqian Cheng*, Luqi Wang*, Fei Gao, Fengyu Cai, Jixin Guo, Mengxiang Lin, Shaojie Shen, 2018 International Symposium on Experimental Robotics (ISER 2018).

Optimal Time Allocation for Quadrotor Trajectory Generation, Fei Gao, William Wu, Jie Pan, Boyu Zhou, Shaojie Shen, 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2018). (Full Text)

Flying on Point Clouds: Online Trajectory Generation and Autonomous Navigation for Quadrotor in Cluttered Environments, Fei Gao, William Wu, Wenliang Gao, Shaojie Shen, 2018, Journal of Field Robotics (JFR).

Optimal Trajectory Generation for Quadrotor Teach-and-Repeat, Fei Gao, Luqi Wang, Kaixuan Wang, William Wu, Boyu Zhou, Luxin Han, Shaojie Shen, 2019,  IEEE Robotics and Automation Letter (RA-L), presented at ICRA 2019

Real-Time Scalable Dense Surfel Mapping, Kaixuan Wang, Fei Gao, Shaojie Shen, 2019 IEEE International Conference on Robotics and Automation (ICRA 2019)

Robust and Efficient Quadrotor Trajectory Generation for Fast Autonomous Flight, Boyu Zhou,  Fei Gao*, Luqi Wang, Chuhao Liu, Shaojie Shen, (* for the corresponding author), 2019, IEEE Robotics and Automation Letter (RA-L), will be presented at IROS 2019. 

FIESTA: A Fast Incremental Euclidean Distance Fields for Online Quadrotor Motion Planning, Luxin Han, Fei Gao*, Boyu Zhou, Shaojie Shen, (* for the corresponding author), 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2019).

Flying Through A Narrow Gap Using Neural Network: An End-to-end Planning And Control Approach, Jiarong Lin, Luqi Wang, Fei Gao, Shaojie Shen, Fu Zhang, 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2019).

Supplement Video

• Quadrotor Motion Planning Directly on Point Clouds (SSRR 2016 Best Paper)

• Gradient-based Online Quadrotor Safe Trajectory Planning in Complex 3D Environments (IROS 2017)

• Autonomous Aerial Navigation Using Monocular Visual-Inertial Fusion (JFR 2017)

• Online Safe Trajectory Generation For Quadrotors Using Fast Marching Method and Bernstein Basis Polynomial (ICRA 2018)

• ACT: An Autonomous Drone Cinematography System for Action Scenes (ICRA 2018,  collaboration with UCSB Learning-Based Multimedia Lab)

• Optimal Time Allocation for Quadrotor Trajectory Generation (IROS 2018)

• Flying on Point Clouds: Online Trajectory Generation and Autonomous Navigation for Quadrotor in Cluttered Environments (JFR, 2018)

• Optimal Trajectory Generation for Quadrotor Teach-and-Repeat (RA-L, 2019)

• Robust and Efficient Quadrotor Trajectory Generation for Fast Autonomous Flight (RA-L, 2019)

• FIESTA: Fast Incremental Euclidean Distance Fields for Online Motion Planning of Aerial Robots (IROS 2019)

• Flying through a narrow gap using neural network: an end to end planning and control approach (IROS 2019)

Teaching

Teaching Assistant at:

• ELEC 1110: Introduction to Electro-Robot Design (Under-Graduate/2016 Spring)
• ELEC 6910P: Introduction to Aerial Robotics (Post-Graduate/2016 Fall)
• ELEC 1110: Introduction to Electro-Robot Design (Under-Graduate/2017 Spring)