Grasping

Slides for the next two labs are available here. You will need access to a Google account with Google Drive for the following labs.

Tutorial on the basics of grasping. The following topics are covered to provide students with a basic understanding of how grasp quality estimation works: friction cones, wrench spaces, and grasp quality computation.

Please download the source code from here and upload it to your Google Drive. Some parts of the code are left empty for students to complete, deriving on the concepts taught in class and in the tutorial. Instructions on how to complete the program are provided in the worksheet itself as well.

Solution for the worksheet is given here.

This lab is focused on robot-object grasp pose generation using physics-based simulators.

Useful links:

User Manual for the Pybullet Physics Engine.

Code to visualize Pybullet in Google Colab.

Pybullet-grasp library.

Assignment:

This assignment explores Pybullet, one of the most common physics engines, and applies it to grasp generation for robotics. Students are provided with a notebook with PyBullet installed, a robot hand model, and five object models. The task is to generate the best grasp for each object by evaluating the generated grasps with known grasping metrics.

Please download the source code and upload it to your Google Drive.