Abstract
We present a four-arc progressive assessment system with a geospatial common operational picture (COP), developed during the IGNITE '26 hackathon (March 16–19, 2026). Arc 0 establishes a digital engineering ecosystem using OML as the authoritative source of truth. Arc 1 constructs a complete DTE&A digital thread with BFO 2020-aligned ontology (20+ OWL classes, 6 SHACL shapes, 8 SPARQL competency questions), seven Wymore five-tuple subsystem models, and the IDSK five-question framework. Arc 2 diagnoses Conway's Law barriers: 49 sociotechnical obstacles across 6 dimensions, with People and Culture accounting for 47% of the total. Arc 3 operationalizes the WRT-2516 four-dimensional transformation assessment. Arc 4 renders an operational SEAD mission on a CesiumJS 3D globe with 12 force elements, 7 DEVS state machines spanning 34 state variables, threat engagement zones, and a six-phase mission timeline. We describe an adversarial AI validation methodology using 15+ autonomous agent swarms across multiple frontier models (Claude Opus/Sonnet, GPT-5.4 Pro, Codex 5.4, Gemini) that caught four critical fabrications and one geodetic error, achieving 35/35 validation checks. The system demonstrates that AI-orchestrated multi-agent development with formal systems engineering methods can produce defense analysis tools in 72 hours while maintaining rigorous verification through adversarial quality assurance.
PALLC and the work
OUSW(R&E)
DTE&A
TRMC
DOT&EIGNITE—Innovative Group for Next-Generation Integration, Testing, and Engineering—brought multidisciplinary teams together to explore connected digital engineering. The 2026 event involved OUSW(R&E) Engineering and Developmental Test, Evaluation and Assessments (DTE&A), the Test Resource Management Center (TRMC), and the Director, Operational Test and Evaluation (DOT&E).
The event included a limited user test of TRMC’s Golden Forge ecosystem. Teams worked across engineering models, requirements and test evidence to examine how a connected environment could support engineering assessment.
PALLC founder Brad M. Philipbar contributed as simulation lead to a multidisciplinary IGNITE ’26 team. The team linked engineering models, requirements and test evidence with a geospatial common operational picture, supporting progressive assessment and review of SI-assisted outputs.
The team’s paper, Four-Arc Progressive Assessment with Adversarial AI Validation: An IGNITE ’26 Geospatial COP for Digital Engineering, describes that approach. Contact PALLC for more information about the research and its application.
Office of the Under Secretary of War for Research and Engineering