Project Info
Multi-finger robot hand control
Xiaoli Zhang
xlzhang@mines.edu
Project Goals and Description:
- Model the multi-finger robot hand in the simulation platform
- Develop control algorithms to control the multi-finger robot hand in in-hand manipulation tasks
- Evaluate the developed control algorithms on both simulated and physical robot hands.
More Information:
Grand Challenge: Engineer the tools of scientific discovery.
- Dexpilot: Vision-based teleoperation of dexterous robotic hand-arm system
- Hierarchical fingertip space: A unified framework for grasp planning and in-hand grasp adaptation
- http://wiki.wonikrobotics.com/AllegroHandWiki/index.php/Allegro_Hand
- https://robots.ros.org/allegro-hand/
- https://github.com/simlabrobotics/allegro_hand_ros
Primary Contacts:
Xiaoli Zhang, xlzhang@mines.edu | Lingfeng Tao, tao@mines.edu | Yunsik Jung, yunsikjung@mines.edu
Student Preparation
Qualifications
- Knowledge of kinematics and dynamics
- Knowledge of basic control methods
- Programming ability (Matlab, Python, C++)
TIME COMMITMENT (HRS/WK)
8
SKILLS/TECHNIQUES GAINED
- Understand the control of the multi-finger robot hand
- Modeling and optimization of complex robot structure
- Advanced robot control methods (model predictive control, inverse kinematics, deep reinforcement learning)
- Robot Operating System (ROS)
MENTORING PLAN
- Prepare robot operation tutorials
- Provide necessary resources like literature and open-source code
- One-on-one mentoring in the lab
- Weekly meeting to track progress and answer questions
- Assign Ph.D. students as the co-mentor
PREFERRED STUDENT STATUS
Freshman
Sophomore
Junior
Senior