Prospective

I am looking for organized, motivated Ph.D candidates or research assistants to work on a variety of cool topics in autonomous robotics. Students with either EE/CS/ME/Mechatronics major are welcome to apply. Candidates are expected to have a solid background in programming and/or system prototying, and the eagerness to learn new techniques/knowledge like myself. For sure, a love for robotics is a must. Candidates with serious experience in Robocon, Robcup or IARC are especially welcomed.

Personally, I have a broad research interest. My main research area is algorithmic robotics, including motion planning, multi-robot system, robotic perception, grasping, manipulation, reinforcement learning, learning from demonstration, so on so forth. But I am also excited about combining intelligent software (e.g., deep learning, vision, and interaction approaches) with intelligent hardware (e.g., soft gripper, dexterous manipulator, tactile sensor array, and cameras). I am also serious about applying my research in real industry, and my main focus is two novel topics: 3C assembly leveraging deformable object manipulation, and intelligent warehouse management. If you are interested in these projects and have cool ideas, you are always welcome to discuss with me.

Note 1: I understand that some students may be talented in research or lab capabilities, but may not have a shiny GPA/English score. If you are of this case, my suggestion is that you can work with me for a short time (half year or one year) and get some serious publications (in robotics SCI is not necessary a sign for serious publications), which will help to compensate for your disadvantages during the application.

Note 2: Students are encouraged to apply for the HK Phd Fellowship (HKPF) . There are two reasons. If you get the HKPF, first you will enjoy a financial benefit over general research postgraduate students, and it is an honor. Second, since you are supported by the fellowship, I can also save the money to support more research students :)

From my personal side, I am looking forward to working with students who can make things happen. In particular, after given an idea about one challenging problem, you are expected to fill in the details and present a complete solution. Indeed, this is NOT easy because you may meet many difficulties: the experiment results may not be as expected, the performance is not as good as planned, or the original idea seems to be not feasible at all -_-||. Some students may give up at this step, or just passively wait for help from professors. However, this is not a qualified researcher should behave. Instead, you should repeat the loop of think, check, and trial, until you find a reasonable solution. During this process, for sure I will discuss with you and provide suggestions, but you should NOT trust in every word/idea I said: you must make your own judgement, which is the main gift from a postgraduate study.

Passion alone is not sufficient to be a good researcher. You need to prepare yourself with necessary backgrounds. Here I only give some suggestions if you want to work in algorithmic robotics. I cannot provide too much suggestion about hardware side, since myself is still learning, but you can teach me :)

To brush up your linear algebra background, I suggest working through Stephen Boyd's EE263: Introduction to Linear Dynamical Systems at your own pace.

Sometimes you need to learn about linear system aspects (Kalman filtering, LQR), I recommend Stephen Boyd's EE363: Linear Dynamical Systems.

If you want go deeper into the theory of linear systems, I recommend Claire Tomlin's EE221a: Linear System Theory.

If you want to learn more about convex optimization, try Stephen Boyd's EE364a: Convex Optimization I and Stephen Boyd's EE364b: Convex Optimization II. Both of them have all course materials, including lecture videos, available online.

For (although draft-status) more about optimal control and motion planning, Russ Tedrake's class: Underactuated Robotics: Learning Planning, and Control for Efficient Agile Machines could give you a somewhat different angle, some complementary ideas, and more examples.

To understand more about traditional control theory, I suggest working through Astrom and Murray, Feedback Systems. A more traditional book on nonlinear control is Slotine and Li, Applied Nonlinear Control.

For reinforcement learning, a great introductory text is Sutton and Barto, Reinforcement Learning, and a more mathematically oriented text is Bertsekas and Tsitsiklis, Neuro-dynamic programming. David Silver's course is also very good.

For optimal control, I suggest Daniel Liberzon, Calculus of variations and optimal control theory.

For motion planning, the traditional book is Steven Lavalle's planning algorithms.