Abstract: An autonomous drone swarm system and method for subterranean and GPS-denied operations utilizing artificial intelligence for coordinated operations comprises command drones equipped with large language model processors and subordinate drones coordinated through a hierarchical Queen-Worker architecture. The system processes natural language commands, generates autonomous mission plans, and coordinates multi-drone operations through an encrypted self-healing mesh communication network utilizing multiple modalities including radio frequency, optical, acoustic, and visual channels. A Unified Ground Frame with Surface-Referenced Z (UGF-SRZ) coordinate system provides signed vertical coordinates enabling seamless operation across surface and subterranean environments. GPS-denied navigation combines multi-sensor dead reckoning, collaborative positioning, and distributed mapping with fog-layer processing for enhanced accuracy.

Abstract: An autonomous drone swarm system and method utilizing artificial intelligence for coordinated operations comprises command drones equipped with large language model processors and subordinate drones coordinated through a hierarchical Queen-Worker architecture. The system processes natural language commands, generates autonomous mission plans, and coordinates multi-drone operations through an encrypted self-healing mesh communication network utilizing laser, radio frequency, and visual communication channels. Multi-modal sensor integration including electro-optical, infrared, LiDAR and photogrammetry, radio frequency, and chemical detection provides comprehensive environmental awareness while federated learning algorithms enable distributed coordination in signal-denied environments.