Total quality concepts in the laboratory: The Automated Intelligent Assistant

Conference · Tue Jan 01 00:00:00 EST 1991

OSTI ID:7296975

Filby, E E; Rankin, R A

Minimizing instrument downtime is a significant factor for the Mass Spectrometry Laboratory (MSL) at the Idaho Chemical Processing Plant, where spent fuel is repocessed to recover uranium. MSL measures the amount and isotopic content of the uranium in the dissolver solution and in the final uranium oxide product. High quality, timely data are needed for accountability, and the Automated Intelligent Assistant is designed for that purpose. Its design revolves around a multi-tasking computer system, and the core system would integrate data collection, interrupt handling, data handling, expert system, and user interface. The ultrahigh vacuum components were chosen as a starting point for the initial implementation. The multi-tasking part of the concept was not yet tackled; the current system employs sequential program code. Information from the analog input board goes through an analog-to-digital converter into the data interface part of the Vacuum System Assistant, which collects a tremendous amount of real-time data. The data were used to develop a rule base for the expert system, which also utilized a commercial expert system shell. The current system identifies 12 normal'' events or situations: 8 responses are triggered by data changes, while 4 of the recognizable events require a user request. The stored data will be useful for training new operators. Plans include the addition of more ability to handle off-normal events and (long-term) implementation of the fully integrated design for the Automated Intelligent Assistant. 14 figs. (DLC)

• Research Organization: Westinghouse Idaho Nuclear Co., Inc., Idaho Falls, ID (United States)
• Sponsoring Organization: USDOE; USDOE, Washington, DC (United States)
• DOE Contract Number: AC07-84ID12435
• OSTI ID: 7296975
• Report Number(s): WINCO-11717; CONF-9106346-2; ON: DE92017212
• Resource Relation: Conference: 5. annual artificial intelligence (AI)/neural networks conference, Baltimore, MD (United States), 18-20 Jun 1991
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL PROTECTION
11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS
IDAHO CHEMICAL PROCESSING PLANT
MASS SPECTROSCOPY
NUCLEAR MATERIALS MANAGEMENT
EXPERT SYSTEMS
DATA ACQUISITION
DATA ANALYSIS
DATA PROCESSING
MATERIAL BALANCE
QUALITY CONTROL
REAL TIME SYSTEMS
REPROCESSING
SPENT FUELS
VACUUM SYSTEMS
CONTROL
ENERGY SOURCES
FUEL REPROCESSING PLANTS
FUELS
MANAGEMENT
MATERIALS
NATIONAL ORGANIZATIONS
NUCLEAR FACILITIES
NUCLEAR FUELS
PROCESSING
REACTOR MATERIALS
SEPARATION PROCESSES
SPECTROSCOPY
US AEC
US DOE
US ERDA
US ORGANIZATIONS
990301* - Information Handling- Data Handling- (1992-)
055001 - Nuclear Fuels- Safeguards
Inspection
& Accountability- Technical Aspects
050800 - Nuclear Fuels- Spent Fuels Reprocessing