Capability-based Synthesis for Intelligent Automation
Abstract
This project defines and develops capability specifications of AI components used in autonomous systems. These specifications permit automatic synthesis of correct-by-construction high-level behaviors to enable intelligent automation of complex systems. This research will develop and apply these capability-based synthesis techniques to define autonomous behaviors using a combination of Hierarchical Finite State Machines and Behavior Trees. The approach will be demonstrated using the DDS-based Robot Operating System and the Flexible Behavior Engine. This provides a demonstration of the capability-based synthesis approach and provides an open-source platform for training new users in the tools of automatic synthesis, validation, and verification of autonomous system behaviors
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 11, 2025
- Source ID
- N001742310018
Entities
People
- David Conner
Organizations
- Christopher Newport University
- United States Navy