Fault detection and reliability of the GASS multiprocessor survey system

The failure detection features of the geophysical airborne survey system (GASS) are investigated. An overview of the system design is given to permit discussion of the failure detection problem in GASS. The types of data collected are ordered by their importance to the survey mission, and the types of failures that may occur are identified. The ability of the system to detect errors is analyzed for each data type and for the communications links between major subsystems. Since GASS is an environmental survey system, model-based failure detection methods are not applicable. Recommendations are made to enhance system reliability by improving various aspects of failure detection within GASS. All major hardware in GASS is specified; thus all modifications suggested are implementable in software or through minor hardware alterations. It is concluded that the GASS failure-detection system provides SDF (software-detectable fault) and extensive HDF (hardware-detectable fault) information for most of its equipment. The only significant weak points noted in the system were in the area of DDF (data detectable fault) failure detection.<
>