Human-Centered Robotic and Automation Laboratory
人中心機器人與自動化實驗室
LATEST PROJECTS
Metal Polishing Robotics
Metal Polishing Robotics
We provide a solution for metal polishing robotics including trajectory planning, ACF, positioning, and quality inspection.
Collaborative Robot Technologies
Collaborative Industrial Robots (CIBOTs)
We provide a collaborative-teaching method on industrial robots. It helps commercial industrial robots upgrade to collaborative robots.
Self-learning robots
Robots are able to adapt to environments to finish tasks. In the video, the robot avoided obstacles to learn the desired trajectory for the specific task.
AI / Machine Learning
Real-time object detection
We focus on condensing deep leaning architectures to reduce inference time. In the video, it only took 5-8 ms to recognize each workpiece on the convey.
Deep learning robotic grasps
We focus on robotic grasps by deep learning approaches. Robots are intelligent to grasp objects without predefined grasps.
Point-cloud 3D object detection
We focus on dealing with imbalance 3D point cloud data and robots learn to grasp objects using 3D point cloud data.
Automatic cancer tumor cell detection system
We focus on developing an automatic cell counting system by machine-learning approaches.
Sensing
and Imaging
Image pattern recognition
We provide an AI solution to recognize patterns on IC wafers. The accuracy is much higher than the traditional methods.
Active force compliant tool
We focus on developing an active force compliant 1-DOF tool for polishing. The tool can respond to external force and adapt its stroke.
Mechatronics and
Designs
Wood industry automation
We focus on developing cutting-edge automation technologies for wood industry including polishing and quality inspection.
Robot intelligent insertion
We focus on developing an intelligent robotic system to learn how to insert objects into plugs. In the video, the system avoided collisions and learned the correct positions to insert.
Robot intelligent insertion by human demonstration
We focus on developing an intelligent robotic system from human demonstrations. In the video, the system learned the optimal trajectory to perform insertion and adapted to the environmental disturbance.
Task learning by human demonstration
We focus on understanding what humans teach to robots. In the video, the robot recognized the person's motion and repeated the same task.
Robot rubik cube
We focus on how to understand tasks and synthesize robot motion to accomplish tasks. In the video, the robot recognized the rubik cube and figured out how to use dual arms to solve the puzzle.