Probabilistic Programming for Analyzing Complex Systems (PPACS)

Development planning is necessary for the US Air Force (USAF) to effectively analyze gaps in its planned missions and capabilities, guide investment in new aircraft and technologies, and plan for future workforce skills and development. A key development planning task is analyzing the impacts of alternative designs on the concept of operations (CONOPs), the lifecycle costs of system of systems (SoS), the ability to meet the mission effectiveness measures (MEMs), and the sensitivity of the system design to various internal and external factors. Performing this analysis is difficult and costly due to the variety and complexity of Air Force systems, as well as uncertainty in the system design as well as uncertainty imposed by external factors. To address this development planning challenge in multi-domain USAF systems and operations, Charles River Analytics developed a method for Probabilistic Programming for Analyzing Complex Systems (PPACS). This comprehensive tool uses advanced probabilistic programming and probabilistic relational models to support a wide variety of development planning needs. It will enable systems engineers to integrate and analyze uncertain, multi-domain systems, thereby improving the accuracy and decreasing the cost of development planning for complex and uncertain problems.