Chengnan Jimmy Shentu

Chengnan (Jimmy) Shentu

I am a PhD student with the Department of Computer Science at the University of Toronto, advised by Prof. Jessica Burgner-Kahrs. I hold a B.A.Sc in Engineering Science at the University of Toronto, majoring in Robotics Engineering with a minor in Artificial Intelligence.

In the past, I've been fortunate to collaborate with Xin Yi at Tsinghua University, and Peter Grant at University of Toronto Institute for Aerospace Studies. I have also completed a one-year SerDes Application and Test Internship at Huawei Canada in the industry.

Currently I am working on developing more capable and accessible continuum robots.

Email / CV / Google Scholar / LinkedIn

Education
	Ph.D in Computer Science Department of Computer Science, University of Toronto Sep 2022 - Aug 2027 \| Toronto, ON, Canada Supervisor: Prof. Jessica Burgner-Kahrs
	B.A.Sc in Engineering Science, Robotics Faculty of Applied Science and Engineering, University of Toronto Sep 2017 - Apr 2022 \| Toronto, ON, Canada The Engineering Science Research Opportunities Program Scholarship, 2019 Dean's Honour List - 2017-2021

Research

I am interested in bridging concepts from robotics, machine learning, and control theory to build robotic systems capable of safe and efficient interactions with humans and the environment. I also do some mechanical designs occasionally for fun.

	Universal-jointed Tendon-driven Continuum Robot: Design, Kinematic Modeling, and Locomotion in Narrow Tubes (Extended Abstract) Chengnan Shentu, Jessica Burgner-Kahrs ICRA@40, 2024 website / PDF / arXiv / We propose a novel synovial universal-jointed TDCR design which restricts robot motion to achieve torsional rigidity and efficient state representation (through a set of discrete joint angels). The modules' geometry is parametrically determined by tube dimensions and joint angle limits.
	A Non-Linear Model Predictive Task-Space Controller Satisfying Shape Constraints for Tendon-Driven Continuum Robots Maximillian Hachen, Chengnan Shentu, Sven Lilge, Jessica Burgner-Kahrs Under Review website / PDF / arXiv / We propose a MPC to enable tendon-driven continuum robots (TDCRs) to navigate between target end-effector positions while avoiding collision with a user-defined safe zone, which is crucial for safe teleoperation.
	MoSS: Monocular Shape Sensing for Continuum Robots Chengnan Shentu, Enxu Li, Chaojun Chen, Puspita Triana Dewi, David B. Lindell, Jessica Burgner-Kahrs IEEE Robotics and Automation Letters (RA-L), 2023 *Equal contribution PDF / arXiv / IEEE Xplore / GitHub We propose a novel monocular shape sensing method for continuum robots, called MoSSNet. It comprises an encoder and three parallel decoders to uncover spatial, length, and contour information from a single RGB image, and then obtains the 3D shape through curve fitting. MoSSNet outperforms existing stereo-vision-based shape sensing methods in terms of real-time capability and has much lower hardware complexity compared to embedded sensing methods.
	Open Continuum Robotics - One Actuation Module to Create them All Reinhard M. Grassmann, Chengnan Shentu, Taqieldin Hamoda, Puspita Triana Dewi, Jessica Burgner-Kahrs Frontiers in Robotics and AI, 2024 PDF / arXiv(html) / Frontiers / Youtube / OpenCR Project To democratize continuum robots research, we propose an actuation module to build torque controlled continuum robots and provide open-source software and hardware with our initiative called the Open Continuum Robotics Project.
	Proprioceptive Impedance Control of a Planar Tendon-Driven Continuum Robot Undergraduate Thesis, supervised by Jessica Burgner-Kahrs Sep 2021 - Apr 2022 In this work, I investigated and implemented an impedance controller for a planar continuum robot in simulation, to achieve variable dynamic behavior for safe interactions with the environment.
	Risk of Side Channel Attack on Head Mounted Devices Supervised by Prof. Xin Yi, Tsinghua University May - Sep, 2021 This work seeks to investigate the risk of side channel attack on head mounted consumer devices, such as VR headsets and smart-glasses, through inertial measurement unit (IMU) by recovering speech or motion information. I identified several machine learning models that could eavesdrop on device userswithout microphone access and discussed how to defend such attacks effectively.
	Mixed Parameter Estimationto of Full Stall Aircraft Model Supervised by Prof. Peter Grant, University of Toronto Institute for Aerospace Studies May - Aug, 2019 This project is motivatedby aircraft stall being one of the leading causes of commercial aircraft accidents. The outcome is an improved aircraft flight model suitable for professional simulation training.

Projects
	Autonomous Vehicle Control University of Toronto Self-Driving Car Team 1st Place among 8 Universities, AutoDrive Challenge, 2021 Development of velocity scheduler and model predictive controller with dynamic vehicle model using C++ in ROS
	Pacbot University of Toronto Robotics Association Competition Organized and led weekly meetings with subteam members focusing on the design and implementation of mechatronics for the autonomous robot.
	BALL BALL U: An Autonmous Ball Dispensing Prototype Engineering Design Competition 2nd Place, University of Toronto, Apr 2019 Technical Report / Glance / Competition Designed, fabricated and programmed a proof-of-concept robot prototype that autonomously detect anddeploy objects to canisters.