- Researched about human-robot interaction and how to apply Deep Neural Networks for self-learning process of robot under the initial human demonstration. - Designed an automatic bin-picking system with 6-DOF manipulator, RGB-D Realsense Camera, and pneumatic gripper. The system utilized YOLOv5 DNN model to select the pick-up target objects and Iterative Closest Points algorithm for object’s pose estimation. - Created a multiple-view point cloud scan-matching algorithm for pose estimation of randomly oriented objects, especially occluded objects, which enhances the accuracy in grasping decision.

แสดงเพิ่มเติม

- Researched about applications of the Model-Mediated method, Input-to-State Stability conditions, and Deep Learning in controlling the bilateral teleoperation, which improves the stability and fidelity of system in uncertain environments. - Designed mock surgery experiments to verify the precision of the adaptive model-mediated control method with KUKA 7-DOF manipulator and Omega haptic device. - Proposed an adaptive model reducing error in rendering force to master device from 21% to 10%.

แสดงเพิ่มเติม

- Researched about controlling parallel delta robots and their applications in automation industry. - Designed MATLAB and Simulink models to analyze kinematic, dynamic, and motion planning of company products including industrial delta robots and 4-legged robots.

แสดงเพิ่มเติม

- Researched how to apply machine learning for the optimization of uncertainty hardware design. - Applied Kriging model based on Gaussian interpolation to create surrogate model which optimizes the design for crashworthiness of vehicle side impact. The design minimizes the deformation from impact while maintaining required weight.

แสดงเพิ่มเติม