ABSTRACT

Robotics and Computer Vision Program

ABSTRACT


Proposal #: 0328782
Title: Study Contact States and Compliant Motion between General Objects Critical to Real and Virtual World Applications
PI: Xiao, Jing
U of NC Charlotte

For real-world robotic operations involving contacts or compliant motion as well as for certain virtual-world applications such as dynamic simulation and haptic interaction, two important and related problems regarding information of contacts among general objects (including robots) remain largely unsolved:
(1) how to obtain knowledge of contact states effectively and efficiently for handling motion involving complex contacts, and
(2) how to enable automatic on-line (real-time) identification of geometrically valid contact states in the presence of real-world uncertainties or digital approximation errors in virtual environments to ensure valid subsequent action or response.
Problem (1) is particularly crucial and challenging for tasks involving a large number of complex contact situations and high-dimensional motion, and Problem (2) is especially critical to tasks with high accuracy requirement, including robotic assembly, manipulation, dynamic simulation, virtual prototyping, and certain haptics applications. This project seeks to find principled solutions to these problems through systematic investigation of a number of fundamental issues related to contact states and compliant motion involving general objects, which include objects with curved surfaces, articulated objects, and certain deformable objects. As an integrated part or derivative of such basic investigation, the project will further study: physically accurate haptic display of contact states in virtual environments, and general compliant motion planning and execution in real world, as well as their related applications.

The general problems addressed in this project (as introduced above) are crucial to advancing the state of the art in several related areas and their many applications, including robotics, haptics, and dynamic simulation, and yet there has been little principled and systematic research towards these problems. The PI has studied aspects of the problems restricted to contacts between only polyhedral rigid bodies. Hence, the project represents a very significant, multi-dimensional extension with new and substantial challenges, but not without the considerable foundation built upon the PI's related previous work.

The project research activities, aside from its expected contributions to both basic research and a wide range of related applications, will also have a significant impact on student education and research training. The interdisciplinary breadth of the research activities spans many areas in robotics and control, AI and intelligent systems, computational geometry, geometrical and physical modeling and simulation, haptics, computer graphics and virtual environments, as well as human-machine interaction and virtual collaboration. The project is also comprehensive with both theoretical and algorithmic components and experimental and system integration components. It will thus provide a rich and balanced environment that can accommodate students from all levels (from Ph.D. to undergraduate) to obtain knowledge and research training in the various related areas.

The research project will also provide an excellent opportunity for (a) cross-discipline research collaboration between the PI and her colleagues at the newly established College of Information Technology (COIT) at UNC Charlotte, and (b) close international collaboration between the PI's group and a leading research group on compliant motion control in Europe. Such collaborations will be instrumental to enhancing research capabilities as well as research infrastructure at the new COIT. The project will clearly enrich student education and contribute significantly to the special goal of COIT of providing students an interdisciplinary learning and research experience in the broad spectrum of Information Technology.

Please report errors in award information by writing to: awardsearch@nsf.gov.