Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Automated Glovebox Work-cell Design - 18370

Conference ·
OSTI ID:22977679
; ; ; ;  [1]
  1. The University of Texas at Austin, Nuclear and Applied Robotics Group (United States)

Los Alamos National Laboratory (LANL) is the primary plutonium science and manufacturing facility in the United States supporting multiple national security programs. It is responsible for completing a high variety of small-batch manufacturing tasks requiring the handling and precision machining of Special Nuclear Materials (SNM). Given the myriad of tasks that must be completed in a limited workspace, LANL intends to utilize flexible, multi-use robotic work-cells capable of completing a variety of mission critical tasks in a single glovebox with no re-tooling needed between batch jobs. This approach would allow LANL more flexibility in its production schedule and to meet its future production and research objectives. In this paper, we present a system that integrates multiple sensors and other manufacturing peripherals into existing glovebox infrastructure. Currently, a technician can place an object in the work area where vision sensors and supervisory software identify the object and its pose and automatically determine which manufacturing task to complete. Pose uncertainty is addressed in real-time by the task planner, eliminating the need for fixtures. Grasps are validated using pressure sensors, payload sensors, and grasp quality metrics to ensure that objects are not dropped or damaged during manipulation. The robot, sensors, and peripherals are controlled using open-source software built utilizing ROS (Robot Operating System), which minimizes redevelopment and allows for re-use and extensibility. Established capabilities support vision, collision-free path planning, and more, minimizing development time. For this effort, particular attention was given to the integration of general peripheral devices (drill, chuck, etc.), which is critical to efficiently utilizing ROS in manufacturing environments. This integration is achieved with an in-development framework that allows for quick integration of peripherals, including I/O devices and motors. Contact tasks utilize the Generalized Contact Control Framework (GCCF) previously developed at UT Austin. GCCF allows application developers to easily define multiple contact control tasks for a single robot. This paper summarizes the key software components necessary to develop the multi-use work-cell including existing ROS components. UT-developed software for grasp validation, peripheral integration, and contact control is presented in more depth. Finally the glovebox system is demonstrated completing a set of tasks with no fixturing necessary. Primary tasks include surface preparation/ablation, drilling, press-fitting, and sorting, as well as supporting material transfer and inspection tasks. The glovebox work-cell reduces cost, operator dose, ergonomic issues, and time loss due to material transfer between gloveboxes that only support one task. If multiple glovebox work-cells are in use, they have the potential to eliminate bottlenecks and scale production capabilities beyond current limits. (authors)

Research Organization:
WM Symposia, Inc., PO Box 27646, 85285-7646 Tempe, AZ (United States)
OSTI ID:
22977679
Report Number(s):
INIS-US--20-WM-18370
Country of Publication:
United States
Language:
English

Similar Records

Automated tracking and grasping of a moving object with a robotic hand-eye system
Journal Article · Wed Mar 31 23:00:00 EST 1993 · IEEE Transactions on Robotics and Automation (Institute of Electrical and Electronics Engineers); (United States) · OSTI ID:5711153

1999 IEEE international conference on robotics and automation
Book · Thu Jul 01 00:00:00 EDT 1999 · OSTI ID:20013273

Towards Cooperative Control of Humanoid Robots for Handling High-Consequence Materials in Gloveboxes - 17291
Conference · Sat Jul 01 00:00:00 EDT 2017 · OSTI ID:22802328