Engineering topics
Publications and source records attributed to Carnes, James R..
Fault isolation and sensor placement are vital for monitoring and diagnosis. A sensor conveys information about a system's state that guides troubleshooting if problems arise. We are using machine learning methods to uncover behavioral patterns over snapshots of system simulations that will aid fault isolation and sensor placement, with an eye towards minimality, fault coverage, and noise tolerance.
Specific activities in NASA's environmental control and life support system (ECLSS) advanced automation project that is designed to minimize the crew and ground manpower needed for operations are discussed. Various analyses and the development of intelligent software for the initial and evolutionary Space Station Freedom (SSF) ECLSS are described. The following are also discussed: (1) intelligent monitoring and diagnostics applications under development for the ECLSS domain; (2) integration into the MSFC ECLSS hardware testbed; and (3) an evolutionary path from the baseline ECLSS automation to the more advanced ECLSS automation processes.
Viewgraphs on Environmental Control and Life Support System (ECLSS) advanced automation project evolution and technology assessment are presented. Topics covered include: the ECLSS advanced automation project; automatic fault diagnosis of ECLSS subsystems descriptions; in-line, real-time chemical and microbial fluid analysis; and object-oriented, distributed chemical and microbial modeling of regenerative environmental control systems description.
A Failure Modes and Effects Analysis workstation (FMESassist) was designed for use during the development of the space station. It assists engineers in the complex task of tracking failures and their effects on the system. Engineers experience increased productivity through reduced clerical loads, reduced data inconsistency, and significantly reduced analysis time. System developments benefit from a more thorough analysis than was available using previous methods.
Qualitative modeling is the study of how the physical world behaves. These physical models accept partial descriptions of the world and output the possible changes. Current systems assume that the model is static and that physical entities do not effect change into the world. An approach to planning in physical domains and a working implementation which integrates qualitative models with a temporal interval-based planner are described. The planner constructs plans involving physical qualities and their behavioral descriptions.
One approach to transforming information stored in relational data bases into knowledge based representations and back again is described. This system, called Foundation, allows knowledge bases to take advantage of vast amounts of pre-existing data. A benefit of this approach is inspection, and even population, of data bases through an intelligent knowledge-based front-end.