One of the goals of the Space Shuttle Program was to lower operating costs by eventually reducing the size and scope of the ground support team required, in all previous programs, to monitor and assess spacecraft and subsystem performance and functionality. @o accomplish this goal, however, meant that major portions of a task, hitherto performed by hundreds of specialized experts, would have to be performed onboard by a relatively small crew already busy with mission operations and experiment support. A major challenge facing the Space Shuttle designers, therefore, was to devise an approach which would accommodate the onboard system management requirement but which would not overwhelm the flightcrew. Further, the design would have to provide initially for full ground support capability with an orderly transition of the function onboard as the capability became validated by flight experience.
It quickly became obvious that the only way to avoid overtaxing the crew would be to automate much of the system monitoring and assessment task. Because the computational requirements could only be grossly estimated initially, the capability of the central computer complex to assimilate the load was questioned. Therefore, a tradeoff study was conducted to determine the relative merits of an integrated approach versus a separate, independent computer dedicated to system management. A corollary issue involved the data acquisition process. On previous programs, only that information required by the crew to operate the spacecraft or to respond to emergencies was made available onboard; the rest was reduced and analyzed on the ground. The traditional approach to onboard instrumentation was to install a network of sensors, transducers, pickoffs, and signal conditioners together with a telemetry processor, which acquired, formatted, and multiplexed data for transmission to the ground. The data set thus acquired contained all of the information required to perform a system assessment, but, because the instrumentation network was overlayed on and essentially independent of the onboard operating systems, many of .the measurements were accessible only on the ground. For the Space Shuttle, provisions had to be made to make all required data accessible onboard as well.
Two computation and data acquisition options were examined. (See fig. 3-6.) In alternate 1, the traditional instrumentation system was augmented with the necessary computational resources to provide an essentially independent capability. In alternate 2, the data were provided to and assimilated in the operating avionics system and its computer complex. The difficulties involved in integrating another computer, the associated controls, and the required displays into the system discouraged consideration of a separate approach, and the problems of providing the necessary additional data to the operating system were also difficult to resolve. The design finally chosen, alternate 2, was to install data buffers in the instrumentation telemetry processors which could be accessed by all the central computers and thereby would provide a source for those measurements not already available in the operating systems. The system management function was initially relegated to the fifth computer in the central complex, the one not included in redundant- set operations. As the system matured, however, many of the system management functions proved to be critical and were transferred to the redundant complex.
Figure 3-6. - System management approaches
Means for assessing and condensing the information into a reasonable set which could be readily assimilated and acted on by the crew remained as another issue. Critical- function management was incorporated into the redundant set, and automatic failure detection and response were mechanized where appropriate.
Overall system monitoring was accomplished by comparing the sensed valued of selected measurements against preset upper and lower limits, and, depending on the urgency, either switching to an alternate path or annunciating the situation to the crew for appropriate action. Cathode-ray-tube (CRT) display pages devised for each subsystem provided quick and concise monitoring capability. Other crew assistance features which were considered included switch monitoring to assure that the correct system mode and configuration -for a given operational situation were established, communications antenna management controllable either from the ground or onboard to ensure optimum coverage with minimum crew involvement, and an extensive caution and warning system to provide alerts for any abnormal situation.
NASA Office of Logic Design
Last Revised: February 03, 2010
Digital Engineering Institute
Web Grunt: Richard Katz