6G-Empowered Agentic Multi-Vehicle Digital Twin Coordination for Sustainable Autonomous Farm Operations

Abstract

This paper presents an agentic AI framework for task offloading in emerging 6G-enabled agricultural systems. Distributed decision-making replaces static offloading policies, allowing autonomous computation placement between onboard hardware, peer machines, and edge nodes. The approach integrates predictive Quality-of-Service (QoS) estimation, latency constraints, signal integrity, and energy metrics. Evaluations rely on distributed computation nodes to analyze performance and efficiency under rural coverage conditions, bandwidth-accuracy trade-offs, multi-node scalability, and energy impacts. The work demonstrates how proactive, multi-agent coordination enables resilient, efficient, and sustainable agricultural operations. This contribution focuses on (i) agentic decision-making for enhanced scheduling, (ii) integration of predictive QoS and energy-aware optimization and (iii) a Kubernetes-based validation setup for controlled evaluation under realistic rural connectivity conditions using a decentralized policy based on predicted QoS and resource metrics.