Abstract:

The global decline of natural pollinators poses a serious threat to agricultural productivity, particularly in crops such as Hass avocado that depend heavily on cross-pollination. Robotic pollination has emerged as a potential alternative, yet little is known about the efficiency and coordination of multi-drone systems under realistic conditions. This paper presents a theoretical framework and simulation-based analysis of two coordination strategies for pollinator drones: a sweep strategy, where drones divide the orchard into strips and follow systematic coverage, and a greedy strategy, where each drone visits the nearest available receptive flower. A stochastic orchard model with varying flower densities and temporal receptivity windows was implemented, and experiments were conducted in MATLAB under factorial combinations of swarm size, density, and strategy. Results show that the sweep approach consistently achieves higher coverage and balanced task distribution, while the greedy approach reduces energy consumption per visit but introduces workload imbalance and partial coverage gaps. The comparison reveals a trade-off between maximizing agricultural robustness and minimizing operational cost. In addition, this study is framed within the context of Hass avocado production in Colombia, a sector of growing international relevance, where pollination challenges directly affect productivity and competitiveness. The study provides a reproducible baseline for evaluating coordination strategies in robotic pollination and highlights future opportunities for hybrid algorithms, field validation, and integration with adaptive multi-agent control. Beyond the technical contribution, these findings support the exploration of sustainable digital agriculture practices and offer insights for strengthening resilience and innovation in tropical crops such as Hass avocado.