Autonomous Robotic Manipulation - What’s Within Reach? Jeannette Bohg

Explore the principles and challenges of autonomous robotic manipulation in this Stanford webinar featuring Assistant Professor Jeannette Bohg. Gain insights into decision-making under uncertainty, the importance of multimodal information, and choosing the right robot action representation. Discover how failures and limitations of existing approaches have shaped current research. Learn about feature representation, unigrasp, continuous feedback systems, and environmental exploitation in robotic manipulation. Watch demonstrations and examples that illustrate key concepts, and understand the research questions driving advancements in this field. Engage with topics such as noisy sensing, inaccurate models, hard-to-predict dynamics, and the development of manipulation skills for autonomous robots.

Introduction
My Research
Feature Representation
Examples
Insights
Learnings
Unigrasp
UDF File
Visualization
Toy example
Demonstrations
Summary
Continuous Feedback
System Architecture
Example
Recap
Exploit the Environment
Manipulation Skills
Approach
Results
Demonstration
Results from my first project
Research questions
Q A