Abstract
A major problem in environmental restoration and waste management is the disposition of used fuel assemblies from the many light water reactors in the United States, which present a radiation hazard to those whose job is to dispose of them, with a similar threat to the general environment associated with long-term storage in fuel repositories around the country. Actinides resident in the fuel pins as a result of their use in reactor cores constitute a significant component of this hazard. Recently, the Department of Energy has initiated an Actinide Recycle Program to study the feasibility of using pyrochemical (molten salt) processes to recover actinides from the spent fuel assemblies of commercial reactors. This project concerns the application of robotics technology to the operation and maintenance functions of a plant whose objective is to recover actinides from spent fuel assemblies, and to dispose of the resulting hardware and chemical components from this process. Such a procedure involves a number of safety and human factors issues. The purpose of the project is to explore the use of robotics and artificial intelligence to facilitate accomplishment of the program goals while maintaining the safety of the humans doing the work and the integrity of the environment. This project will result in a graphic simulation on a Silicon Graphics workstation as a proof of principle demonstration of the feasibility of using robotics along with an intelligent operator interface. A major component of the operator-system interface is a hybrid artificial intelligence system developed at Oak Ridge National Laboratory, which combines artificial neural networks and an expert system into a hybrid, self-improving computer-based system interface.
Get full access to this article
View all access options for this article.
