Smart sensor technology for joint test assembly flights.

Berry, Nina M. ; Sheaffer, Donald A. ; Bierbaum, Rene Lynn ; Dimkoff, Jason L. ; Walsh, Edward J. ; Deyle, Travis Jay (University of Nebraska-Lincoln, (Omaha Campus)) ; Marx, Kenneth D. ; Pancerella, Carmen M. ; Doser, Adele Beatrice (Sandia National Laboratories, Albuquerque, NM) ; Armstrong, Robert C.

Sandia National Laboratories

关键词: Decision Making;    Training;    Artificial Intelligence;    Microelectronics;    Feasibility Studies;   

DOI  :  10.2172/918324
RP-ID  :  SAND2003-8589
RP-ID  :  AC04-94AL85000
RP-ID  :  918324

美国|英语

来源: UNT Digital Library

【 摘 要 】

The world relies on sensors to perform a variety of tasks from the mundane to sophisticated. Currently, processors associated with these sensors are sufficient only to handle rudimentary logic tasks. Though multiple sensors are often present in such devices, there is insufficient processing power for situational understanding. Until recently, no processors that met the electrical power constraints for embedded systems were powerful enough to perform sophisticated computations. Sandia performs many expensive tests using sensor arrays. Improving the efficacy, reliability and information content resulting from these sensor arrays is of critical importance. With the advent of powerful commodity processors for embedded use, a new opportunity to do just that has presented itself. This report describes work completed under Laboratory-Directed Research and Development (LDRD) Project 26514, Task 1. The goal of the project was to demonstrate the feasibility of using embedded processors to increase the amount of useable information derived from sensor arrays while improving the believability of the data. The focus was on a system of importance to Sandia: Joint Test Assemblies for ICBM warheads. Topics discussed include: (1) two electromechanical systems to provide data, (2) sensors used to monitor those systems, (3) the processors that provide decision-making capability and data manipulation, (4) the use of artificial intelligence and other decision-making software, and (5) a computer model for the training of artificial intelligence software.