Efficient coordination of multiple robots is essential for precision agriculture tasks such as planting, watering, and soil monitoring. This paper presents a decentralized, behavior-based planning framework for agents operating in a constrained farmlike grid. Each agent performs tasks independently, prioritizing unexplored cells based on local proximity and column availability. The system supports dynamic rerouting without global coordination and promotes collaboration by allowing idle agents to assist peers, thereby reducing idle time and improving coverage efficiency. We implement this framework in a simulated farm environment and compare it against a rigid preassigned-column baseline. Across diverse grid scenarios, our approach achieves complete task coverage with lower average simulation time, robust recovery from agent failures, and consistent performance under varying deployment conditions. The system is scalable, fault-tolerant, and well-suited for autonomous multi-robot operations in dynamic agricultural settings.
Related links
Details
Title
Multi-Agent Coverage Planning for Agricultural Soil and Crop Monitoring
Publication Details
Conference Proceedings: International Mechanical Engineering Congree & Exposition, Vol.5: Dynamics, Vibration, and Control, IMECE2025-166209
Resource Type
Conference proceeding
Conference
ASME 2025 International Mechanical Engineering Congress and Exposition (IMECE2025) (Memphis, Tennessee, USA, 11/15/2025–11/19/2025)