Verification, Validation, Assurance, and Trust of Machine Learning Models and Data for Safety-Critical Applications in Armaments Systems

To address the challenges posed by verification and validation (V&V) of artificial-intelligence (AI) and machine-learning (ML) for high assurance in safety-critical functions of armaments systems, ARiA and Arizona State University (ASU) will develop and demonstrate the feasibility of the Machine Learning Qualification Process (MLQP). The Phase I effort will (1) research and develop fundamental metrics and measures that can be used in the certification and qualification of training data sets and models in armaments systems and (2) develop a Phase II plan for MLQP including an initial design specification and capabilities description to guide development of a prototype solution. The Phase II development plan will address the reduction of technical risk based on assessment of the maturity of metrics and measures developed in Phase I and will provide guidance for risk-reducing activities conducted during Phase II.