Lasitha Wijayarathne
Robotics & Control Engineer
Building safety-critical robotic systems and intelligent interaction models
Contact-rich manipulation | MPC | Differentiable simulation | Learning-based control
Currently at Intuitive Surgical | PhD Georgia Tech
I am a PhD student at Georgia Tech working on improving dexterterity of redundant robotic systems and trajectory planning with contact awareness in constrained environments. My current research interests are:
Redundant robotics systems
Trajectory optimization
Contact rich physics simulation
Machine learning and model based reinforcement learning
Summary of my research objectives can be found on these slides.
Contact Based Manipulation and sim2real Learning
Currently working (on a internship with MERL) on manipulation primitives and using ML techniques to minimize the sim2real gap. Implications of this work lead towards generalization of manipulation tasks with unknown objects. More details to come...
Simultaneous Trajectory Optimization and Force Control with Soft Contact Mechanics
Force modulation of robotic manipulators has been extensively studied for several decades but is not yet commonly used in safety-critical applications due to a lack of accurate interaction contact modeling and weak performance guarantees - a large proportion of them concerning the modulation of interaction forces. This study presents a high-level framework for simultaneous trajectory optimization and force control of the interaction between manipulator and soft environments. Link (IROS2020)
Identification of Compliant Contact Parameters and Admittance Force Modulation on a Non-stationary Compliant Surface
Although autonomous control of robotic manipulators has been studied for several decades, they are not commonly used in safety-critical applications due to lack of safety and performance guarantees - many of them concerning the modulation of interaction forces. This paper presents a mechanical probing strategy for estimating the environmental impedance parameters of compliant environments, independent of manipulator's controller design, and configuration. Link (ICRA2020)
Force Feedback-Enabled Dexterous Robotic Micromanipulation Platform for Surgical Tasks
This paper describes the design and experimental evaluation of a dexterous, force feedback-enabled robotic micromanipulation system. The system is composed of a dexterous non-backdriveable robotic wrist rigidly attached to a three-DOF, linearly actuated platform, and is instrumented with a six-axis force/torque sensor at the base of the wrist. By leveraging the rigidity and non-backdriveability of the system, a wrench sensed at the tool tip can be estimated through a rigid transformation from the force/torque sensor location to the known position and orientation of the end-effector. Link (ROBIO2018)
Kinetic Measurement Platform for Open Surgical Skill Assessment
Current surgical skill assessment methods are often based on the kinematics of manual surgical instruments during tool- tissue interactions. Though kinematic data are generally regarded as a sufficient basis for skill assessment, the inclusion of kinetic information would allow the assessment of measures such as “respect for tissue” and force control, which are also important aspects of surgical proficiency. Kinetic data would also provide a richer data set upon which automated surgical motion segmentation and classification algorithms can be developed. Link (DMD 2017)
