Seeding precision dataset

The dataset used in this study is generated through a physics-based simulation of drone aerial seeding under varying environmental conditions. Each sample represents one mission configuration defined by environmental factors such as wind behavior and turbulence, together with drone operating settings. For every configuration, multiple feasible routes are evaluated using the simulation model, which estimates the likelihood of successful seed placement across the target area. The route that achieves the best overall placement performance is selected as the label for that configuration. The dataset therefore contains input features describing the environmental conditions and flight setup, along with output labels indicating the optimal route and its associated performance score. This structure enables supervised learning, where a surrogate model learns to predict route quality or select the best route directly from environmental and geometric features. The dataset supports the development of a route-aware, stochastic planning approach by capturing variability in wind conditions and their effect on aerial seed placement outcomes.