A Test Bed for Research in Collaborative Robotics
Abstract
A testbed for human-robot collaboration is proposed to facilitate research toward automating manipulation-intensive tasks in unstructured environments. Despite the explosion of low-cost sensor technology over the past decade, robotic perception and cognition are still rudimentary compared to human capabilities and will continue to be so for the foreseeable future. Likewise, the speed, precision, endurance, and safety advantages of robotic manipulation over human manual labor have been well known for over 50 years. Human-robot collaboration is relatively well developed in the area of unmanned vehicles, where human operators are responsible for high-level decisions, but are generally relieved from more mundane tasks such as navigation, takeoff/landing, and basic system monitoring. The same cannot be said for tasks that require manipulation and mobility in highly unstructured environments. Examples include equipment maintenance, explosive ordinance handling, healthcare, and supply/logistics applications. The next big step in military robotics will be systems that partially automate such tasks, hence freeing highly skilled personnel from dull, dirty, and/or dangerous tasks so that they may concentrate on high-level decision making. The same concept extends beyond the Department of Defense to the industrial base it depends upon to support the war fighter. Efficient, agile, and precise manufacturing enabled by collaborative robotic technology is a strategic advantage. The proposed system will integrate three existing robots in the University of ArkansasÕ Manufacturing Automation Laboratory with advanced sensing, grasping, and computing technologies. The testbed is envisioned as a rapidly reconfigurable system capable of supporting multiple research directions. These include collaborative path planning, distributed embedded vision systems for coordination of unmanned ground vehicles, real-time knowledge extraction from unstructured 3D data, 3D activity and object recognition, and defense strategies for IoT attacks. In addition to research, the proposed equipment will allow both graduate and undergraduate engineering students to gain valuable hands-on experience with advanced robotic systems. The equipment will be integrated into the laboratory experiences of multiple courses in Industrial, Mechanical, and Computer Engineering. It will further the extension mission of the University by facilitating collaboration with defense-related manufacturing facilities in the state and region.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 01, 2019
- Source ID
- W911NF1910242
Entities
People
- Harry Pierson
Organizations
- Army Contracting Command
- United States Army
- University of Arkansas