Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Pick and place with a legged manipulator
As part of the Cybathlon challenge, this project consists of solving a challenging manipulation task consisting of picking up a bottle from the ground and placing it vertically on a table (without toppling it).
This project is part of the Cybathlon challenge where we are competing with ANYmal (with an arm) to help people with disabilities in their everyday life. In the project, the student will be able to work on several components especially image processing, grasping and planning for a legged manipulator.
Several algorithmic components which can be used to solve the task have already been developed in the lab (image recognition and tracking, grasping) and it will be up to the student to decide which components to use and especially solve the main challenge which consists of placing the bottle vertically on the table (includes estimating the pose of the bottle in the hand, figuring the placing position, placing it smoothly and ensuring that it is stable before releasing the grip). A force sensor on the end effector could potentially be used for the accurate placing.
This project is part of the Cybathlon challenge where we are competing with ANYmal (with an arm) to help people with disabilities in their everyday life. In the project, the student will be able to work on several components especially image processing, grasping and planning for a legged manipulator.
Several algorithmic components which can be used to solve the task have already been developed in the lab (image recognition and tracking, grasping) and it will be up to the student to decide which components to use and especially solve the main challenge which consists of placing the bottle vertically on the table (includes estimating the pose of the bottle in the hand, figuring the placing position, placing it smoothly and ensuring that it is stable before releasing the grip). A force sensor on the end effector could potentially be used for the accurate placing.
- Testing and adapting robust detection and grasping modules for grasping a bottle (20%)
- Developing a perception (vision/depth) module for estimating the bottle pose after grasping (30%).
- Developing a closed loop planning and control module for placing the object (30%).
- Integrating and testing on hardware (30%).
- Testing and adapting robust detection and grasping modules for grasping a bottle (20%) - Developing a perception (vision/depth) module for estimating the bottle pose after grasping (30%). - Developing a closed loop planning and control module for placing the object (30%). - Integrating and testing on hardware (30%).
- Strong coding skills
- Experience in Robotics
- Experience in Perception is optional but considered a plus
- Experience with ROS is optional but considered a plus
- Strong coding skills - Experience in Robotics - Experience in Perception is optional but considered a plus - Experience with ROS is optional but considered a plus