ANSR

DARPA’s Assured Neuro Symbolic Learning and Reasoning (ANSR) program explores combining symbolic reasoning with data-driven learning to make autonomous systems more trustworthy. My team’s contribution explores neurosymbolic decision support and how to present information through tactical mapping interfaces.

I work on software engineering and systems integration for my team’s ANSR research. My focus is on connecting technical components with the information an operator needs to inspect and understand.

The system is collaborative work by researchers and engineers. This page describes my contribution to that team’s work; the linked DARPA page explains the broader ANSR program.

Our contribution combines AI-supported interpretation with an operator-facing interface. Integration work has to account for both the information a component produces and how someone will encounter it in the application.

For a decision-support interface, displaying an output is only part of the problem. The person using it needs to understand what it refers to and what remains uncertain. That requirement shapes how I approach the connection between model behavior and the interface.

This contribution remains research and development work. The overview describes my work at a high level, without claiming production deployment or operational acceptance. Detailed architecture, evaluation results, and field material are not included here.