Integrating Multi-Agent Systems in AI: A Framework Inspired by Physiology for Complex System Design

Abstract This study explores the integration of artificial intelligence (AI) through a Multi-Agent System (MAS), utilizing autonomous networks to implement a novel framework demonstrated in a navigation system for the visually impaired. It capitalizes on MAS's scalability and robustness, employing a holonic, recursive agent structure inspired by physiological systems for efficient modeling. The framework facilitates intelligent agent integration and utilizes MQTT and DDS for critical data exchange. Structured in five sections, the paper discusses design, experimental validation, and results, highlighting MAS's potential to enhance distributed intelligent systems across various sectors.