Rapidly Refreshable Synthetic Ranges: AI-Generated Terrain and Agentic Adversaries for Counter-Uncrewed Training

13 May 2027
Adaptive Edge Theatre
Outpacing Autonomous and Uncrewed System Threats

Recent conflicts show uncrewed systems changing in tactics, autonomy and electronic-warfare resilience faster than ranges and training scenarios can be updated. Physical ranges are costly, fixed and constrained by safety and spectrum rules. Conventional synthetic environments often depend on terrain databases and scripted threats that take months to build. Forces need synthetic ranges that can be regenerated as quickly as the threat changes. This session presents an approach from Yasin Science & Technology that combines physics-grounded digital twins, AI-generated terrain and agentic AI adversaries. It draws on lunar digital twin development for NASA, where the problem is extreme: elevation data is incomplete, much of the surface is permanently shadowed, there is no GPS and communications are delayed. Our methodology builds terrain in layers. It starts from authoritative elevation data, fills gaps with procedural and AI-generated terrain, and textures from available imagery, so a usable, physics-consistent environment can be produced from sparse and imperfect sources. We then populate the environment with agentic uncrewed adversaries whose autonomy, swarming behavior and response to jamming and communications loss are configurable parameters, not fixed scripts. Friendly uncrewed and robotic systems are informed by YST's work on robotics for confined spaces and mine and defect detection. Scenarios are versioned and reproducible, so an operational lesson can be turned into a new threat profile and rehearsed without rebuilding the environment. The work shows that the main bottleneck in counter-uncrewed training is not rendering fidelity but how fast credible terrain and adaptive threat behavior can be generated. It also shows that methods built for denied, data-poor environments transfer directly to contested terrestrial operations. The session outlines how this pipeline supports mission rehearsal, counter-autonomy experimentation and human-machine teaming in communications-denied conditions.

Chairperson
Enrico Raue
Enrico Raue, New Markets Development and Sales Manager - MASA
Speakers
Alhassan Yasin
Alhassan Yasin, Professor and CEO - Johns Hopkins University