Human-Robot Collaborative Autonomy: From Multi-sensory Perception to Cognition and Execution
Abstract
Short Work Statement:The proposed equipment (robots with arms, hands and fingers that are touch sensitive and can grip hard and soft, large and small objects of various shapes and material composition, computers and peripherals) is to support research in developing dexterous robots that can collaborate with humans in close proximity in performing a variety of tasks. This requires multi-disciplinary research including perception, planning and control, machine learning, etc.Objective:Purchase robots with arms, hands and fingers for research in developing dexterous robots for human-robots collaboration.Approach:The multi-disciplinary research supported by the proposed equipment is for developing robots that can learn from and work with humans performing a variety of tasks that require dexterous manipulation such as, cooking, equipment repair, assembling structures, etc. The underlying research involves learning from demonstrations, visual perception and language understanding, control theory, task and motion planning. All of these are of importance for DOD. The PIsare currently supported by a grant from DARPA. They also have pending proposals with the Army Robotics CTA and the NSF NRI program.ONR Mission/RelevanceThe proposed equipment will support research primarily in the ONR Autonomy. The research is also relevant to ONR Information Dominance focus area. Major challenges for robots are fine-grained manipulation of objects, hand-eye coordination, and motion planning to perform a task. The proposed equipment supports research in these areas, particularly in hand-eye coordination and dexterous manipulation
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 23, 2016
- Source ID
- N000141612926
Entities
People
- John Baras
Organizations
- Office of Naval Research
- United States Navy
- University of Maryland