Flight critical systems research (fcsr) task order statement of work (SOW) 1.0 task order title: software anomaly taxonomy validation 2.0 background: NASA s aviation safety (avsafe) program s…

Flight critical systems research (fcsr) task order statement of work (SOW) 1.0 task order title: software anomaly taxonomy validation 2.0 background: NASA s aviation safety (avsafe) program s integrated vehicle health management (ivhm) project is pursuing foundational research in the development of technologies for automated detection, diagnosis, prognostics, and mitigation of adverse events due to aircraft software. This effort is being conducted under the sub-element entitled software health management in version 2.0 of the ivhm technical plan ( http://WWW.aeronautics.NASA.GOV/NRA_PDF/ivhm_tech_plan_C1.PDF ), which seeks to explore software health management in the context of system level dependability cases -- a central recommendation of a recent national academies report (jackson, d.; ET AL. Software for dependable systems: sufficient evidence? National academies press, 2007, http://WWW7.nationalacademies.ORG/cstb/PUB_dependable.html ). Under previous flight critical systems research (fcsr) tasking (NNL07AB06T concept development for software health management ) a catalog of historical aircraft software anomalies, uncovered during verification and validation activities conducted throughout the software development lifecycle, was compiled, and a comprehensive failure taxonomy for software failures in aircraft systems was developed. The compilation of the catalog and taxonomy was derived through inspection, by subject matter experts, of text-based report archives of selected advanced flight-critical software development programs. 3.0 objective: the objective of this task is to investigate the use of auto-classification algorithms for purposes of replicating and validating the software failure taxonomy developed under task NNL07AB06T. The investigation will include the application of the classification parameters and toolset to large datasets that would otherwise be unsuitable for inspection by subject matter experts. The resulting taxonomy of anomalies will serve as a candidate for software health management frameworks. 4.0 description of task requirements: the contractor shall perform the following tasks: task 1: classification tool a) configure, assess and select an auto-classification algorithm, such as the mariana auto-classifier algorithm (https://dashlink.ARC.NASA.GOV/algorithm/mariana/), for application to software anomaly detection. Task 2: taxonomy validation a) apply classification tool selected in task 1 to datasets used for deriving the taxonomy generated under task NNL07AB06T. B) apply the classification tool selected in task 1 to representative datasets comprised of a minimum of 1,000 anomaly records. 5.0 period of performance: 11 months from task order award date. 6.0 travel: IT is anticipated that the following trips will be required: description where persons no. Of trips no. Of days technical interchange meeting NASA larc hampton, VA and/or NASA ARC moffett field, CA 2 2 1 7.0 reporting requirements: in addition to the contract documentation requirements listed in contract exhibit a, the contractor shall submit the following reports. Task REF report description QTY. Due date delivery instructions 1 & 2 reporting for this task shall be included in the monthly technical letter progress report. Via email or electronic media (e.g., CD or DVD) in MS word, excel, powerpoint, or adobe PDF in accordance with contractor exhibit a distribution. 8.0 deliverables: in addition to the contract documentation requirements listed in contract exhibit a, the contractor shall deliver the following items: item no. Task REF deliverable description QTY. Due date delivery instructions 1 1A interim report: at a minimum this document shall describe the selection and configuration of an auto-classification algorithm as described in task 1A. 1 EA 5 months from task order award date. Via email or electronic me