Collaborative Dexterous Manipulation: Mechanisms and Interfaces
Abstract
Despite renewed focus over the last decade and numerous advances, robotic manipulation is not yet versatile enough for general, unscripted tasks in unstructured environments. Human-robot collaborative teams are poised for deployment in complex scenarios sooner than autonomous systems, as the humans cognitive abilities can deal with corner conditions and decisions that are difficult for autonomy. Still, a number of roadblocks remain for developing side-by-side, or eyes-on" collaboration and control mechanisms for manipulation. Supervisory control is difficult due to the intrinsic high dimensionality of the output space, whereas low-level teleoperation relies on dedicated hardware which can be obtrusive in the #field. The end-effector must be versatile enough for complex tasks, but still intuitive to control. No current robotic manipulation system meets all these goals. This project aims to advance the state of the art in Human-in-the-Loop robotic manipulation by jointly addressing end-effector design and both supervisory and low-level teleoperation through unobtrusive interfaces.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 10, 2016
- Source ID
- N000141612026
Entities
People
- Matei Ciocarlie
Organizations
- Office of Naval Research
- Trustees of Columbia University in the City of New York
- United States Navy