"Precision Irrigation 2.0: Intelligent Co-Design of Drone Fleets and Soil Sensor Networks for Water-Efficient Agriculture"

Precision irrigation in center-pivot fields demands a critical yet unsolved challenge: simultaneously deciding where to place soil sensors, what technology to deploy, how many drones to operate, and how to route them—all while ensuring synchronized, real-time field snapshots within tight time windows. This work presents an innovative Decision Support System (DSS) that solves this coupled co-design problem by jointly optimizing sensor placement, multi-UAV routing, and fleet sizing under realistic electromagnetic communication constraints and agronomic synchrony requirements. Unlike sequential workflows that decouple these decisions and incur substantial sub-optimality penalties, our framework integrates three computational layers: (1) a mixed-integer linear program optimizing sensor placement and routing jointly; (2) an Adaptive Large Neighborhood Search heuristic with Hamiltonian lower-bound verification, ensuring fleet-size recommendations are certified within 0.19% of optimality; and (3) an economic layer comparing passive RFID tags (cheap but demanding larger drone fleets) against active LoRaWAN nodes (expensive but requiring fewer drones). Applied to a 241-cell center-pivot field, the DSS reveals that co-design value is regime-dependent—substantial under LoRaWAN—and identifies a critical sensor-density band (76–79 nodes) where the cost-optimal technology switches. The framework accommodates non-circular field geometries and demonstrates that enforcing spatial homogeneity reduces blind-spot radii by up to 9% at zero additional cost. Verified on real agronomic synchrony windows (15–30 minutes), this DSS empowers farm operators to make data-driven infrastructure decisions before deployment, balancing hardware investment against operational drone costs over multi-year farm lifecycles.