Lehrinhalte
This course introduces fundamental algorithmic approaches for creating robot systems that can autonomously manipulate physical objects in unstructured environments such as homes. We will cover basic principles for endowing autonomous robots with planning, perception, and decision-making capabilities, i.e., topics include perception (including approaches based on deep learning and approaches based on 3D geometry), planning (robot kinematics and trajectory generation, collision-free motion planning, task-and-motion planning, and planning under uncertainty), as well as dynamics and control for adaptive and reactive manipulation.
Tentative list of topics:
[list]
[*]Topology in robotics and rigid body motions
[*]Refresher on forward, inverse kinematics and dynamics
[*]Differential kinematics and optimization
[*]Geometric perception and object pose detection
[*]Object pose estimation and tracking and multi-sensor fusion
[*]Grasp generation and grasp evaluation
[*]Trajectory Optimization
[*]Search and Sampling-based motion planning
[*]Force control
[*]Visuomotor policies and intuitive physics
[*]Task and motion planning and belief-space planning
[/list]
Practical exercises will guide understanding fundamental mathematical and algorithmic principles for enabling robotic manipulators to perceive their environment, estimate the current state of the robot itself and the robots or humans in their surroundings, and create a strategy for executing various tasks that involve autonomously manipulating objects in cluttered scenes.
Voraussetzungen
Recommended:
The students should have a fundamental knowledge of robotics and linear algebra. Furthermore, Fundamentals of Robotics (20-00-0735-iv Grundlagen der Robotik) is recommended. Experience in Robot Learning (20-00-0629-vl Lernende Roboter) is also a plus.
Combining the course with the seminar and project lab will equip the students with a greater understanding and in-depth knowledge of the necessary components and principles to enable robotic autonomous manipulation
Online-Angebote
moodle
Course Contents
This course introduces fundamental algorithmic approaches for creating robot systems that can autonomously manipulate physical objects in unstructured environments such as homes. We will cover basic principles for endowing autonomous robots with planning, perception, and decision-making capabilities, i.e., topics include perception (including approaches based on deep learning and approaches based on 3D geometry), planning (robot kinematics and trajectory generation, collision-free motion planning, task-and-motion planning, and planning under uncertainty), as well as dynamics and control for adaptive and reactive manipulation.
Tentative list of topics:
[list]
[*]Topology in robotics and rigid body motions
[*]Refresher on forward, inverse kinematics and dynamics
[*]Differential kinematics and optimization
[*]Geometric perception and object pose detection
[*]Object pose estimation and tracking and multi-sensor fusion
[*]Grasp generation and grasp evaluation
[*]Trajectory Optimization
[*]Search and Sampling-based motion planning
[*]Force control
[*]Visuomotor policies and intuitive physics
[*]Task and motion planning and belief-space planning
[/list]
Practical exercises will guide understanding fundamental mathematical and algorithmic principles for enabling robotic manipulators to perceive their environment, estimate the current state of the robot itself and the robots or humans in their surroundings, and create a strategy for executing various tasks that involve autonomously manipulating objects in cluttered scenes.
Preconditions
Recommended:
The students should have a fundamental knowledge of robotics and linear algebra. Furthermore, Fundamentals of Robotics (20-00-0735-iv Grundlagen der Robotik) is recommended. Experience in Robot Learning (20-00-0629-vl Lernende Roboter) is also a plus.
Combining the course with the seminar and project lab will equip the students with a greater understanding and in-depth knowledge of the necessary components and principles to enable robotic autonomous manipulation
Online Offerings
moodle
- Lehrende: Georgia Chalvatzaki