This lecture introduces various approaches for autonomous grasp and manipulation planning and execution - from classical approaches to deep reinforcement learning methods (end-to-end learning). The usual grasping pipeline covers scene segmentation, object fitting, grasp selection, grasp evaluation, planning, and execution. If time admits we will also study tactile sensing and tactile-based manipulation - an important research topic nowadays.
As a theoretical foundation to tackle these topics, we will first study classical grasp evaluation based on wrench-space analysis: We will consider how to formally describe grasps, how to model contacts and friction, as well as learn about grasp evaluation measures like grasp stability and manipulability.
Tentative Schedule:
The lecture will be given in English if desired by the audience. The slides will be in English.
required prerequisites: Robotic Manipulators
(Robotics Basics: Transforms, Forward/Inverse Kinematics, Velocities/Twists, Jacobian IK)
Frequency | Weekday | Time | Format / Place | Period |
---|
Module | Course | Requirements | |
---|---|---|---|
39-M-Inf-MI Manuelle Intelligenz | Autonomes Greifen | Ungraded examination
Graded examination |
Student information |
The binding module descriptions contain further information, including specifications on the "types of assignments" students need to complete. In cases where a module description mentions more than one kind of assignment, the respective member of the teaching staff will decide which task(s) they assign the students.
At least 50% of the exercises need to be solved.
The (oral) exam will cover topics from both, the lecture and exercises.
A corresponding course offer for this course already exists in the e-learning system. Teaching staff can store materials relating to teaching courses there: