Welcome to Adaptive Robotic Controls Lab (ArcLab).
ArcLab is located at the Mechanical Engineering (ME) at the University of Hong Kong (HKU).
Our research focuses on various control techniques which can enhance the autonomy of robotics.
Open Positions
Contact Us

05.2024 3D dense mapping in aggressive motions remains challenging. We have used event cameras to improve dense mapping under aggressive motions. The performance is even better than several NERF-based methods.
Please check our Advanced Intelligent Systems paper for more details:
Paper: EVI-SAM: Robust, Real-Time, Tightly-Coupled Event–Visual–Inertial State Estimation and 3D Dense Mapping
Video: EVI-SAM

04.2024 How to search for unknown dynamic objects in unknown environments using multiple robots?
We provide a Multi-Agent Reinforcement Learning framework to solving this issue.
Check our RAL paper and video for more details:
Paper: HMA-SAR: Multi-Agent Search and Rescue for Unknown Located Dynamic Targets in Completely Unknown Environments
Video: HMA-SAR

03.2024 How to control quadruped robots with various configurations (structures, size, weights, etc.) and allow them to walk stably in various terrains?
We provide one solution to that with our MorAL framework. See our RAL paper:
Title: MorAL: Learning Morphologically Adaptive Locomotion Controller for Quadrupedal Robots on Challenging Terrains
Video: MorAL

03.2024 We have just opensourced the software for our event-based visual odometry algorithm: ESVIO.
Code:  Github Link
Paper: P. Chen, W. Guan and P. Lu, "ESVIO: Event-Based Stereo Visual Inertial Odometry," in IEEE Robotics and Automation Letters, vol. 8, no. 6, pp. 3661-3668, June 2023.

02.2024 Our paper on continuum robotic arm has been accepted by IEEE Robotics and Automation Letters.
A Tendon-driven Continuum Manipulator with Robust Shape Estimation by Multiple IMUs

11.2023 Our paper on SLAM has been accepted by IEEE Transactions on Intelligent Vehicles.
ECMD: An Event-Centric Multisensory Driving Dataset for SLAM

10.2023 Our paper on quadruped robot has been accepted by IEEE Robotics and Automation Letters:
FT-Net: Learning Failure Recovery and Fault-tolerant Locomotion For Quadruped Robots

Our FT-net can enable quadruped robots to work even in the presence of partial and total rotor failures.

09.2023 Our paper on event-based visual odometry has been accepted by IEEE Transactions on Automation Science and Engineering:
PL-EVIO: Robust monocular event-based visual inertial odometry with point and line features

Our event-based VO can enable quadrotors to perform aggressive maneuvers by using the onboard estimation without replying on motion capture systems.

07.2023 Our paper on deep reinforcement learning has been accepted by IEEE Robotics and Automation Letters:
Learning Agile Flights through Narrow Gaps with Varying Angles using Onboard Sensing


06.2023 Our paper on dynamic obstacle avoidance has been accepted by IEEE/ASME Transactions on Mechatronics.
Title: Flying in Dynamic Scenes With Multitarget Velocimetry and Perception-Enhanced Planning

05.2023 Our paper on visual odometry has been accepted by IEEE Transactions on Instrumentation and Measurement
Title: Filtering 2D-3D Outliers by Camera Adjustment for Visual Odometry

04.2023 Our event-based VIO has been accepted by IEEE Robotics and Automation Letters:
ESVIO: Event-based Stereo Visual Inertial Odometry
The ESVIO can provide a stable and accurate state estimation in the presence of aggressive motions (aggressive rotational rates) and dark environments.

aggressive rotational rates:

from bright to dark environment:

Link                     VIDEO Youtube           VIDEO bilibili

03.2023 We are excited to share our aerial continuum arm AeCoM.
AeCoM: An Aerial Continuum Manipulator with IMU-based Kinematic Modeling and Tendon-slacking Prevention. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2023, 1-13.


                                       LINK                       VIDEO Youtube                   VIDEO Bilibili   

07.2022 Our paper Perception and Avoidance of Multiple Small Fast Moving Objects for Quadrotors with Only Low-cost RGBD Camera has been accepted by IEEE Robotics and Automation Letters
Our quadrotor can avoid multiple fast-moving tennis balls which are thrown simultaneously or continuously with only a RGBD camera.

07.2022 Our paper on event camera-based VIO has been acceptd by IROS 2022:
Monocular Event Visual Inertial Odometry based on Event-corner using Sliding Windows Graph-based Optimization


06.2022 Our paper has been accepted by Journal of Intelligent & Robotic Systems
Terminal Sliding Mode Control for Quadrotors with Chattering Reduction and Disturbances Estimator: Theory and Application
(the work was finished in 2019)


05.2022 Our papers have been accepted by Franklin Open.
Event-Triggered Integral Sliding Mode Formation Control for Multiple Quadrotor UAVs with Unknown Disturbances

Comparative Study of State and Unknown Input Estimation for Continuous-discrete Stochastic Systems.


12.04.2022 Our paper  Real-time identification and avoidance of simultaneous static and dynamic obstacles on point cloud for UAVs navigation submitted to Robotics and Autonomous Systems has been accepted.

Our drone can avoid both static and dynamic obstacles while navigating in an unknown environment.


27.01.2022 Our paper Stereo Orientation Prior for UAV Robust and Accurate Visual Odometry is accepted by IEEE/ASME Transactions on Mechatronics.
Link  VIDEO 

15.09.2021 GUAN W.P. joins the lab as a PhD student. He will be working on SLAM for ground robots.

23.08.2021 Our paper is accepted by IEEE Transactions on Neural Networks and Learning Systems (TNNLS).
Title: Flying Through a Narrow Gap Using End-to-end Deep Reinforcement Learning Augmented with Curriculum Learning and Sim2Real
We used purely deep reinforcement learning to tackle the task instead of referring to any traditional methods.
PDF Link

20.08.2021 LU Minghao, SUN Jiahao and XIE Yuhan joined the lab as PhD/Mphil students.

01.07.2021 Our paper is accepeted by IROS 2021.
Title: A Motion decoupled Aerial Robotic Manipulator for Better Inspection
We designed a motion decoupling mechanism and a gravity compensation mechanism to facilitate the use of these aerial manipulators for inspection tasks.
Video Link

01.11.2020 XU Juntao and XIE Yuhan joined the lab as research assistant. They work on the visual tracking and deep reinforcement learning for drones.

05.09.2020 PENG Rui joined the lab as a PhD student. He was a Research assistant working on the vision for UAVs.

A preliminary demo of the quadrotor flight despite the loss of two propellers.

This is the demo video for the following paper:
P Lu, E van Kampen, Active fault-tolerant control for quadrotors subjected to a complete rotor failure, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 4698-4703.

ArcLab has moved to department of Mechanical Engineering, the University of Hong Kong.

 NIE Shengyi and PENG Rui joined the group as a research assistant.

 Paper "Computationally Efficient Obstacle Avoidance Trajectory Planner for UAVs Based on Heuristic Angular Search Method" is accepted by IROS 2020. Experimental results coming soon.

 Paper "Nonsingular Terminal Sliding Mode Control for a Quadrotor UAV with a Total Rotor Failure" has been accepted and published online by Aerospace Science and Technology.

 A preliminary video of deep reinforcement learning for aggressive quadrotor flights.
 We purely use deep reinforcement learning to plan the trajectory of a quadrotor to pass through titled narrow gaps.


 Paper "Performance comparison of representative model-based fault reconstruction algorithms for aircraft sensor fault detection and diagnosis"
 has been accepted and published online by Aerospace Science and Technology.

 Paper "Dynamic Obstacle Avoidance for UAVs Using a Fast Trajectory Planning Approach " has been accepted by ROBIO 2019.
 PDF  LINK  VIDEO(initial version)
 Drone is passing through moving objects in real time.

 We have won the thrid place in the IROS Autonomous Drone Racing competition.
 Video for part of our flight during the competition.


 Welcome our new RA- Rui CAO.

  A preview of the work on accurate stereo visual odometry for quadrotors
We developed a stereo-vision Orientation prior visual odometry for UAV navigation.

 A preview of the video regarding terminal sliding mode for quadrotors subjected to rotor failures and external disturbances. Our proposed method is very robust against partial rotor failures and external disturbances.

  Welcome our new PhD Han Chen!

 Video of our work: Object Pose Estimation using end-to-end CNN for UAVs

  Our paper titled "Adaptive Unscented Kalman Filter-based Disturbance Rejection With    Application to High Precision Hydraulic Robotic Control" has been accepted by IROS 2019!
  The video of the paper can be found at: Video
  The preliminary version of the paper is at: PDF


 Collaborative paper titled "Boundary Effect-Aware Visual Tracking for UAV with Online Enhanced Background Learning and Multi-Frame Consensus Verification" with Tong Ji University has been accepted by IROS 2019!

 Dr Lu is giving a Keynote at Global Conference on UAV application and Anti-UAV 2019.

 Keynote speech at ICEDME 2019.


 Welcome our new RA Yanhui Guo!

 Welcome our new Postdoc Zhiwei HOU!

 Welcome our new PhD student Ran DUAN!
 Ran was working in our lab as a Research Assistant before his PhD Study.

 Welcome our new Research Assistant Zhangjie TU!

 Dr Peng Lu is invited to give a talk at South China University of Technology.
 Advanced control techniques for autonomous UAVs and robotics

 14.11-16.11 2018
 Dr Peng Lu is invited to give a two-day lecture at Chengdu Aircraft Design & Research Institute.

 Welcome our new Research Assistant Ran DUAN!

 Our paper titled "Nonlinear Disturbance Attenuation Control of Hydraulic Robotics"
 has been accepted by ROBIO 2018!
 We present an efficient and effective control approach for robotics such as hydraulic robots.
 The link is at: Paper
 The paper is at: PDF
 The video is at: Youtube Link

  Welcome visitors from Hong Kong Institution of Engineers (HKIE).
  We have performed two successful UAV demos.



  Welcome our new visiting PhD student Qizhi He!

  Welcome our new RA Chenxi Xiao!

  Collaborative paper with TU Delft is accepted by Journal of Guidance, Control and Dynamics!
  Incremental Sliding Mode Fault-Tolerant Flight Control

  Collaborative paper with ETH&UZH RPG Group is accepted by IROS 2018!
  PAMPC: Perception-Aware Model Predictive Control for Quadrotors

Dr. Peng Lu is invited to give a talk at Northwestern Polytechnical University.
The topic is: Nonlinear controllers for unmanned aerial vehicles: challenges and applications

Dr. Peng Lu is invited to give a talk at Shanghai Jiao Tong University.
The topic is: Advanced motion planning and control techniques for robotics