Architectures for independent test data review on NPOESS VIIRS

To ensure timely and accurate execution of thermal vacuum testing on the first VIIRS flight unit, the Integrated Program Office (IPO) requested near real-time, independent government assessment of test execution. Several organizations shared this responsibility, including: The Aerospace Corporation, NASA, MIT Lincoln Laboratory, and others. In the previous phases of the test program, these organizations performed test data analysis on a two week time-line. To deliver independent assessments of test progress in time for consent to proceed meetings would require a dramatic reduction in response time. To achieve the required response times, changes were needed in ways the program distributed and reviewed data. Two major tasks were essential to the success of this effort. First, the time necessary to distribute daily test data to analysts around the country would need to be less than 24 hours. This was no small task, as VIIRS generates data at a considerable rate[l, 2]. Typical daily data sets were on the order of 100 GB. To facilitate this, the Data Products Division at the Integrated Program Office designed and executed a unique, adaptable, effective, and low-cost data distribution operation. Second, a suite of tools for rapid inspection of commonly reviewed statistics was developed and shared with the community to optimize the efficiency of government analysts. The VIIRS Data Analysis and Decision Support cluster (DADS) was developed at Lincoln Laboratory to fill this need. This paper will describe the system and programmatic architectures necessary to deliver these results, highlighting key design decisions that could be carried forward for use on other instruments and programs.