Intelligent Three Level Learning Architecture for Autonomous UAV Swarms in Search and Rescue

Short summary

This paper proposes a three-level hierarchical learning architecture for autonomous UAV swarms in search and rescue, combining Hebbian neuroplasticity, multi-agent reinforcement learning with GNNs, and meta-learning with BDI reasoning. The architecture is formalized through 22 architectural contracts providing guarantees for safety, optimality, and cognitive resilience. It introduces Swarm Meta-Cognition as an emergent property enabling swarms to monitor and switch cognitive strategies during dynamic operations.

• •Three-level hybrid architecture integrates reflexes, skills, and reasoning for UAV swarms
• •22 architectural contracts provide formal guarantees including safety and cognitive resilience
• •Swarm Meta-Cognition enables self-monitoring and strategy switching across hierarchy levels

Generated with AI, which can make mistakes.