Sihao Sun

Sihao Sun

Researcher in Robotics

Delft University of Technology

Biography

I am a researcher with a focus on planning, estimation, and control of aerial robotic systems. My work has been recognized with the prestigious Best Paper Award of IEEE Robotics and Automation Letters. My research on aerial robot fault-tolerant control and perception has been featured in well-regarded media outlets like IEEE Spectrum. I’m the winner of the Veni grant as part of the Dutch Research Council (NWO) talent program.

I’m currently a researcher at the deparment of Cognitive Robotics (CoR) of Delft University of Technology. From 2022 to 2023, I have been a postdoctoral researcher at the Robotics and Mechatronics (RaM) group at the University of Twente, collaborating with Prof. Antonio Franchi. Prior to this, from 2020 to 2021, I worked as a postdoctoral researcher at the Robotics and Perception Group (RPG) at the University of Zurich under the guidance of Prof. Davide Scaramuzza.

In December 2020, I earned my PhD in Aerospace Engineering from the Control and Simulation Group (C&S) at Delft University of Technology, supervised by Dr. Coen de Visser and Prof. Guido de Croon. My doctoral thesis has addressed the fault-tolerant control problem of aerial robots subject to significant dynamical uncertainties.

Interests

  • Aerial Robotics
  • Robotics Perception
  • Incremental Nonlinear Control
  • Multi-robot Systems

Education

  • PhD in Aerospace Engineering, 2020

    Delft University of Technology

News

  • Feb 15, 2024. I have been apointed an Associate Editor of the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024).

  • January 15, 2024. I have joined the department of Cognitive Robotics as a researcher funded by the Veni Grant. A new journey now begins!πŸ¦ΎπŸ€–πŸ¦Ώ

  • October 17, 2023. I have been appointed an Associate Editor of the IEEE Robotics and Automation Letters (RA-L).

  • August 3, 2023. I have won the prestigious Veni grant from Dutch Research Council.πŸ†

  • April 20, 2023. My MSc student Fang Nan (Now a PhD student with Prof. Marco Hutter at RSL-ETH) has won the prestigious ETH Medal (Outstanding Master’s thesis), congratulations!!!πŸŽ‰πŸŽ‰πŸŽ‰

  • April 1, 2023. A new paper has been accepted by ICUAS on Collaborative Aerial Manipulation, check it out!

  • May 20, 2022. I have received a Best Paper Award from Robotics and Automation Letters!!! πŸŽ‰πŸŽ‰πŸŽ‰

  • May 2, 2022. A new paper accepted by IEEE Transactions on Robotics about comparing NMPC and flatness-based controller for quadrotor agile flights.

Research Projects

Accurate Aerial Manipulation under Uncertainties

This project performs mechatronics design and developes novel algorithms to increase the accuracy of aerial robotic manipulators under dynamical uncertainties.

Cooperative Aerial Manipulation with Cables

This project aims at developing algorithms for aerial robots to collaboratively control the pose of an object using cables

Agile Flight Control for Drones

This project aims at designing flight controllers to achieve agile flight for drones, with the aim of improving its performance in time-critical missions.

Quadrotor Fault Tolerant Flight Control

This project leads to a set of fault-tolerant control algorithms for a quadrotor with motor failures in realistic scenarios, such as withstanding strong winds up to 10m/s, and recovery from extreme conditions.

Aerodynamic Modeling Identification for Multi-Rotor Drones

This project aims at obtaining a set of aerodynamic model for drones using system identification and machine learning technique.

High Efficiency Air Cargo Design

A student project with the aim of desining an unmanned air-cargo with high aerodynamic and structral efficiency.

Publication list

Uncertainty Modeling Enabled Meta Adaptive Control for Aerial Manipulators

Meta Adaptive Control for Aerial Manipulators.

Nonlinear MPC for Full-Pose Manipulation of a Cable-Suspended Load using Multiple UAVs

MPC-based framework for multi-lifting systems.

Nonlinear MPC for Quadrotor Fault-Tolerant Control

We study the performance of using nonlinear MPC on quadrotor fail-safe problem. The result is amazing.

A Comparative Study of Nonlinear MPC and Differential-Flatness-Based Control for Quadrotor Agile Flight

We systematically compare two state-of-the-art flight controllers for quadcopter agile flights, nonlinear MPC and differential-flatness-based controller.

Model Predictive Contouring Control for Near-Time-Optimal Quadrotor Flight

This work proposes a novel control methods for quadcopter agile flights using model-predictive-contouring control.

Performance, Precision, and Payloads: Adaptive Nonlinear MPC for Quadrotors

We propose L1-MPC, an adaptive-model-predictive control framework showing impressive performance against model uncertainties and disturbances.

Autonomous Quadrotor Flight despite Rotor Failure with Onboard Vision Sensors Frames vs. Events

This algorithm uses only onboard vision sensors to control a quadrotor after complete faliure of a rotor, without the aid of GPS, UWB, or other external sensors.

Incremental Nonlinear Fault-Tolerant Control of a Quadrotor With Complete Loss of Two Opposing Rotors

This work, for the first time, applies Incremental Nonlinear Dynamic Inversion controller on an under-actuated control system, namely a quadrotor with complete loss of two opposing rotors. A high-speed wind-tunnel flight test demonstrates the robustness of this method.

High-speed flight of quadrotor despite loss of single rotor

We propose a flight controller to achieve high-speed flights of a quadrotor with one rotor entirely off. Flight tests in the wind tunnel show robustness of our controller in the presence of significant aerodynamic effects.

Upset Recovery Control for Quadrotors Subjected to a Complete Rotor Failure from Large Initial Disturbances

We introduce a flight controller that can recover a quadrotor with one rotor complete off from arbitrary initial orientations and body rates.

Quadrotor gray-box model identification from high-speed flight data

In this work, we presented a gray-box aerodynamic model for a quadrotor identified from windtunnel high-speed flight data, using a step-wise regression algorithm.

Quadrotor safe flight envelope prediction in the high-speed regime: A Monte-Carlo approach

A sampling based approach for estimating the reachable set of a quadrotor in high-speed regime to predict its (dynamic) flight envelope.

Quadrotor fault tolerant incremental sliding mode control driven by sliding mode disturbance observers

This work introduces a novel control method named Incremental Sliding Mode Control, which greatly reduces model dependency and improves the robustness against uncertainties. The algorithm has been validated in a quadrotor fault-tolerant control problem in real flight.