Research

Physical Human-Robot Interaction

We are working on both the hardware (e.g. design of compliant actuators) and the software (development of control schemes) of exoskeleton robots, such that the robot is able to continuously adjust its role and act intelligently according to the human motion intention. Several fundamental issues, such as coupling dynamics between the robot and the actuator, uncertain dynamic parameters, and stochastic noises in interaction forces, will be explored in the theoretical analysis.

We propose a complementary framework for human-robot collaboration that balances the safety of humans and the efficiency of robots. In this framework, the robot carries out given tasks using a vision-based adaptive controller, and the human expert collaborates with the robot in the null space. Such a decoupling drives the robot to deal with existing issues in task space (e.g., uncalibrated camera, limited field of view) and in null space (e.g., joint limits) by itself while allowing the expert to adjust the configuration of the robot body to respond to unforeseen changes (e.g., sudden invasion, change of environment) without affecting the robot’s main task.

Deformable Object Manipulation

Compared to rigid objects, it is very challenging for the robot to manipulate deformable objects, because of the high DOFs, the unknown deformation model, and unforeseen changes during the manipulation. we consider the problem of large deformation control of elastic deformable objects (DLOs); To achieve it, we propose a coupled offline learning and online adaptation method for efficiently learning the global deformation model; This complementary scheme allows more accurate modeling through offline learning and further updating for new DLOs via online adaptation.

We have developed several 3C robots which can manipulate flexible PCBs and USB wires with high speed, high autonomy, and high accuracy.

Robot-Assisted Optical Tweezers

The performance of current manipulation techniques for optical tweezers is commonly limited by several fundamental issues: 1) effectiveness of trapping is only valid locally around the centre of laser beam; 2) limited FOV of microscope; 3) presence of spatially varying trapping stiffness which is difficult to model and identify. We aim to address the aforementioned problems and bridge the gap between traditional robot manipulators and the new generation of robots at micro/nano scales.