SBIR/STTR Award attributes
To address the MDAs need for intelligent dynamic data logging software to reduce run time simulation processor usage, memory usage, and network data bandwidth, and to manage analytical data capture and storage, Physical Optics Corporation (POC) proposes to develop a new Runtime Intelligent Dynamic Data Logging Executive (RIDDLE) program. It is based on a new design incorporating intelligent log data analysis to explore data, detect anomalous occurrences, select analysis schemes, and assess results. Specifically, RIDDLE will use decision trees, discriminant analysis, logistic regression, nearest neighbors, and ensemble classification to perform both root cause crash analysis and data selection for logging with sufficient fidelity to diagnose problems and understand complex simulation events. As a result, RIDDLE directly addresses the MDAs requirements for efficiently logging simulation data based on occurrence/predicted occurrence of the simulator and the simulated events of interest. In Phase I, POC will design and develop the RIDDLE concept and demonstrate the feasibility of the approach, reaching an anticipated TRL of 3, and an incremental development plan for Phases II/III. In Phase II, POC will develop a prototype for intelligent data logging capable of being tested by simulation engineers in a run-for record reaching an anticipated TRL of 5.Approved for Public Release | 17-MDA-9395(24 Oct 17)
