Chen-Lung Eric Lu

Configuration Dependent Kinematic Calibration (ARM 22-01-F-07)

Accurate robot kinematics is critical for precise tool placement using joint control of an articulated robot. Because of non-geomtric factors such as joint/link flexibility, the kinematic model of a robot is configuration dependent. I developed a configuration dependent kinematic model calibration framework for robot arms, leveraging Fourier basis, Neural Network and Autoencoder to achieve sub-millimeter accuracy in TCP position estimation, significantly surpassing baseline performance by sixfold.

In-situ Multi-Robot Multi-Sensor Data-Driven Wire Arc Additive Manufacturing (ARM 22-01-F-07)

Robotic wire arc additive manufacturing (WAAM) is a promising technology for large-scale metal part fabrication. However, the process suffers from geometric inaccuracy with certain materials. I developed an in-situ multi-robot multi-sensor data-driven WAAM framework to achieve high-fidelity part fabrication. A co-located line scanner enables geometry sensing during deposition, while a secondary robot with a thermal camera provides temperature feedback. I further developed data-driven WAAM model with Neural Network and Gaussian Process Regression for WAAM optimization.

Multi-Robot Scan-n-Print for Wire Arc Additive Manufacturing (ARM 22-01-F-07)

To enhance WAAM process geometric accuracy and smoothness, I developed a multi-robot scan-n-print framework for WAAM monitoring and control. I consider a three-robot setup: a 6-DoF welding robot, a 2-DoF trunnion platform, and a 6-DoF sensing robot with awrist-mounted laser line scanner measuring the printed part height profile. I demonstrate significant improvements of the closed loop scan-n-print approach over the current open loop result on both a flat wall and a more complex turbine blade shape.

Robot Motion Optimization (ARM 21-02-F-19)

Developed a full stack robot arm motion optimization framework achieving sub-millimeter accuracy in tracking desired curves in space with the robot arm, surpassing baseline speeds by up to 7.8 times.

Created a user-friendly Python interface for FANUC robots, facilitating seamless programming. Successfully implemented and showcased the developed system on GE's dual arm testbed, highlighting practical applications and versatility.

Emotional Expressive Robot

Designed a control strategy for a 10-degree-of-freedom mobile manipulator to express internal information, focusing on emotions, through motion behaviors. Utilized crowdsourcing to learn the best parameter combinations for expressing different emotions.