QUASAR: Quantifiable Assurance Cases for Trusted Autonomy
Abstract
The goal of QUASAR was to develop techniques and tool support for the construction of dynamic assurance cases (DACs) that address both static design-time concerns and dynamic operational concerns and, for the latter, to provide quantifiable and executable measures of confidence in appropriate assurance properties of the target platforms. In particular, the project focused on applying these techniques to learning-enabled components (LECs) that comprise key parts of autonomous systems, and to that end, we used these techniques to construct platform-specific assurance cases in collaboration with our TA4 partners, Boeing (air domain) and Northrop Grumman (undersea domain).
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2023
- Accession Number
- AD1209631
Entities
People
- Ewen Denney
- Ganesh J. Pai
- Irfan Sljivo
- Rebecca Lee