Technology Experiment Demonstrations on the EUVE Flight Testbed Involving a Collaboration Between NASA Ames Research Center and the Center for EUV Astrophysics

摘要

In the fall of 1993, CEA faced the prospect that Extrema UltraViolet Explorer (EUVE) might be turned off after its primary mission because of budget cuts. Although cost saving measures were examined and implemented throughout CEA, the project was forced to consider radical, new, low-cost approaches for operating the observatory. A three part strategy was developed which included: (1) converting operations from around-the-clock monitoring to day-shift only monitoring; (2) attempting to reduce the large NASA institutional costs (e.g., TDRSS) of operating EUVE; and (3) operating EUVE as a mission operations testbed to introduce new technology in a systematic and disciplined manner (Malina 1994). The last strategy recognized the need to prototype new technology for reduced cost operations while increasing the value of the mission to NASA as a technology testbed. A concurrent study for Dr. Guenter Reigier of NASA Headquarters Code SZ headed by Dr. Ron Polidan (GSFC Code 681) recommended a transition to one-shift operations for both spacecraft and payload operations to reduce costs on the project. Two key factors in accepting risk were that the primary objectives of the mission had been accomplished and the inherent safety of the EUVE spacecraft and payload had been demonstrated. Both the spacecraft and the instrument contained on-board safing mechanisms that had performed remarkably well. The payload safing mechanisms had been activated over a dozen times with no recovery problems. As of the date of this report, both spacecraft and payload continue to perform very well. Two partial failures in the tape recorders and one redundant transmitter failure occurred in 1994, none of which has prevented or restricted science operations. Over 99 of the science data continues to be returned as originally scheduled. No major failures in any payload component have occurred. The risk associated with one-shift operations was mitigated by the introduction of artificial intelligence (AI) software into the EUVE Science Operations Center (ESOC) to monitor the health of the payload during the unstaffed shifts. Our strategy was to evaluate the potential of several commercial and government developed systems before making a full commitment to implement one-shift operations. Since we intended to introduce an AI system to replace entire shifts, not augment existing operator functions, the system had to work in a fully automated fashion, integrated with our existing software. Some of the key criteria were compatibility with our existing software environment (UNIX, distributed network), extensibility to add our own functions, and good technical support and documentation. User interface capabilities, while important, were not a strong factor in the selection since the software would act in the absence of peo- ple.